Marvel Rivals, a new video game from NetEase Inc. and Walt Disney Co., is climbing PC and console charts this holiday season, making the free-to-play superhero shooter title a fresh hit for the two companies.

‘Marvel Rivals’ Video Game Climbs Charts in Win for Disney, NetEase - Bloomberg

Disney debuted an ESPN tab on Disney+ on Wednesday for ESPN+ subscribers and will make about 100 live games available on Disney+ to nonsubscribers.

ESPN hopes to reach more casual sports fans with Disney+ integration

The new Disney+ documentary 'Elton John: Never Too Late' details the pop star's career from his meteoric rise to fame to his farewell tour — and he says that lyrics written by Brandi Carlile heavily inspired the movie.

Brandi Carlile on Writing Elton John's 'Never Too Late' Doc Theme: 'He Doesn't Like to Say Nice Stuff About Himself'

This story is available exclusively to Business Insider subscribers. Become an Insider and start reading now. Have an account?

Disney Isn't Talking Politics, but Live Sports Is a New Battleground - Business Insider

A hacktivist group called NullBulge leaked data from almost 10,000 of Disney’s Slack channels in July.

Disney says most of its business will stop using Slack by the end of the year

Financial details of Disney's new deal with DirecTV weren't disclosed as part of Saturday's announcement.

ESPN and other channels return to DirecTV with a new Disney deal after a nearly 2-week blackout

The dispute lasted 13 days, impacting two weekends of programming.

Disney, ESPN networks restored for DirecTV customers ahead of college football weekend

After digitally recreating the late actor Peter Cushing's likeness for 2016's "Star Wars: Rogue One," Disney is being sued for doing so.

Disney Sued for Bringing Actor Back From the Dead for "Star Wars" Movie

Disney

For the first time, the FIFA Women’s World Cup will be hosted exclusively by a streaming service starting in 2027. And not just any streaming service. Netflix. The leader of New Hollywood. This is a significant development that demonstrates how live sports, especially women’s sports, is converging with content on connected TV. From a Sports Outsider to a Streaming Power Player For years, live sports was a weakness for Netflix. The company did not make a serious stab at sports until 2023 with the airing of the Netflix Cup, a novelty golf event featuring a crossover between Formula One drivers and PGA Tour golfers. But Netflix didn’t become a leader by watching the world pass it by. Netflix has expanded its reach into live sports in some significant ways, including hosting two NFL games on Christmas day, striking a major deal with World Wrestling Entertainment, and, of course, hosting the popular, if widely derided, boxing match between Jake Paul and Mike Tyson. These early ventures reflect Netflix’s strategy of testing the waters with unconventional yet buzzworthy sports spectacles alongside more mainstream sport events. The Netflix Cup and Jake Paul/Mike Tyson match may not have been a traditional sports events, but they demonstrated how the platform could create a blend of entertainment and athleticism. Netflix’s deals with WWE and the NFL indicate an effort to establish itself as a platform for marquee events that integrate live action with storytelling. The Perfect Timing for Women’s Sports A deal with the FIFA Women’s World Cup is well timed, as interest in women’s sports continues to grow. To wit: The FIFA Women’s World Cup in 2023 was the most-viewed women’s sports event in history, attracting nearly 2 billion global viewers.
The Paris 2024 Olympics saw a significant focus on women’s sports, with women’s events drawing large audiences (the women’s gymnastics team final drew 12.7 million U.S. viewers, while the women’s soccer gold medal match attracted nine million viewers).
In 2024, the WNBA achieved its most-watched season, with average viewership for its championship series reaching 1.6 million, a 25-year high.
Per Netflix, the Katie Taylor vs. Amanda Serrano boxing match (technically serving as an undercard for the Jake Paul/Mike Tyson match) was the No. 1 most-watched professional women’s sports event in U.S. history with 74 million live global viewers. The surge in viewership reflects a cultural shift in how audiences perceive and engage with women’s sports. Factors such as increased investment in women’s leagues, better marketing strategies, and the amplified visibility offered by digital platforms have all contributed to this growth. The FIFA Women’s World Cup on Netflix has the potential to further elevate women’s sports to new heights, especially given Netflix’s ability to craft compelling narratives that resonate globally. Netflix’s Edge in Sports Streaming The Jake Paul/Mike Tyson event, while proving to have questionable sports value, shows how Netflix can surround an event with engaging entertainment, as demonstrated by the pre-fight docuseries that highlighted Tyson’s redemption arc and Paul’s brash rise as an influencer-turned-boxer. Netflix also possesses enormous entertainment cachet that crosses over into sports: it’s no coincidence that Beyoncé agreed to perform at halftime during an NFL Christmas Day game after Netflix joined forces with the NFL. What sets Netflix apart is its ability to amplify the entertainment aspects of sports. Netflix can tap into its vast library of content and production expertise to create immersive, multi-layered experiences. Documentaries, behind-the-scenes footage, and athlete-focused mini-series transform simple matches into full-fledged cultural events. Consider what this could look like in practice: exclusive behind-the-scenes documentaries chronicling teams’ journeys to the tournament; halftime performances by global music superstars; interactive features allowing fans to vote on player awards or participate in trivia games during matches, all integrated into Netflix’s platform. All of this could redefine how fans engage with soccer. Strengthening Advertiser Appeal Through Women’s Sports Women’s sports also offer advertisers a unique chance to associate their brands with themes of empowerment, inclusion, and progress—values that resonate strongly with younger audiences, particularly Millennials and Gen Z. Hosting the FIFA Women’s World Cup allows Netflix to position itself as a champion of these ideals while also attracting brands eager to align with this growing cultural movement. For Netflix, this alignment boosts advertising revenue and positions Netflix well. With its tiered ad model already in place, Netflix is well-equipped to capitalize on the advertising opportunities presented by live sports. In effect, the FIFA Women’s World Cup is proving ground for Netflix’s ability to drive revenue from both content and advertising. A Powerful Partnership A partnership between Netflix and the FIFA Women’s World Cup will forge a powerful partnership, blending sports, entertainment, and connected TV for a global audience. The move underlines a broader trend: live sports are becoming a linchpin for connected TV platforms seeking to lock in subscriptions and dominate global markets. Unlike traditional cable or network television, connected TV offers interactivity, data-driven insights, and a chance to engage younger, digitally savvy audiences who demand more than just live coverage. Netflix’s foray into sports capitalizes on these advantages, positioning the platform as a hub where sports meet storytelling. For the first time, the FIFA Women’s World Cup will be hosted exclusively by a streaming service starting in 2027 . And not just any streaming service. Netflix. The leader of New Hollywood. This is a significant development that demonstrates how live sports, especially women’s sports, is converging with content on connected TV. will be hosted exclusively by a streaming service starting in 2027 will be hosted exclusively by a streaming service starting in 2027 From a Sports Outsider to a Streaming Power Player For years, live sports was a weakness for Netflix. The company did not make a serious stab at sports until 2023 with the airing of the Netflix Cup, a novelty golf event featuring a crossover between Formula One drivers and PGA Tour golfers. But Netflix didn’t become a leader by watching the world pass it by. Netflix has expanded its reach into live sports in some significant ways, including hosting two NFL games on Christmas day, striking a major deal with World Wrestling Entertainment, and, of course, hosting the popular, if widely derided, boxing match between Jake Paul and Mike Tyson. These early ventures reflect Netflix’s strategy of testing the waters with unconventional yet buzzworthy sports spectacles alongside more mainstream sport events. The Netflix Cup and Jake Paul/Mike Tyson match may not have been a traditional sports events, but they demonstrated how the platform could create a blend of entertainment and athleticism. Netflix’s deals with WWE and the NFL indicate an effort to establish itself as a platform for marquee events that integrate live action with storytelling. The Perfect Timing for Women’s Sports A deal with the FIFA Women’s World Cup is well timed, as interest in women’s sports continues to grow. To wit: The FIFA Women’s World Cup in 2023 was the most-viewed women’s sports event in history, attracting nearly 2 billion global viewers. The Paris 2024 Olympics saw a significant focus on women’s sports, with women’s events drawing large audiences (the women’s gymnastics team final drew 12.7 million U.S. viewers, while the women’s soccer gold medal match attracted nine million viewers). In 2024, the WNBA achieved its most-watched season, with average viewership for its championship series reaching 1.6 million, a 25-year high. Per Netflix, the Katie Taylor vs. Amanda Serrano boxing match (technically serving as an undercard for the Jake Paul/Mike Tyson match) was the No. 1 most-watched professional women’s sports event in U.S. history with 74 million live global viewers. The FIFA Women’s World Cup in 2023 was the most-viewed women’s sports event in history, attracting nearly 2 billion global viewers. The Paris 2024 Olympics saw a significant focus on women’s sports, with women’s events drawing large audiences (the women’s gymnastics team final drew 12.7 million U.S. viewers, while the women’s soccer gold medal match attracted nine million viewers). In 2024, the WNBA achieved its most-watched season, with average viewership for its championship series reaching 1.6 million, a 25-year high. Per Netflix, the Katie Taylor vs. Amanda Serrano boxing match (technically serving as an undercard for the Jake Paul/Mike Tyson match) was the No. 1 most-watched professional women’s sports event in U.S. history with 74 million live global viewers. The surge in viewership reflects a cultural shift in how audiences perceive and engage with women’s sports. Factors such as increased investment in women’s leagues, better marketing strategies, and the amplified visibility offered by digital platforms have all contributed to this growth. The FIFA Women’s World Cup on Netflix has the potential to further elevate women’s sports to new heights, especially given Netflix’s ability to craft compelling narratives that resonate globally. Netflix’s Edge in Sports Streaming The Jake Paul/Mike Tyson event, while proving to have questionable sports value, shows how Netflix can surround an event with engaging entertainment, as demonstrated by the pre-fight docuseries that highlighted Tyson’s redemption arc and Paul’s brash rise as an influencer-turned-boxer. Netflix also possesses enormous entertainment cachet that crosses over into sports: it’s no coincidence that Beyoncé agreed to perform at halftime during an NFL Christmas Day game after Netflix joined forces with the NFL. What sets Netflix apart is its ability to amplify the entertainment aspects of sports. Netflix can tap into its vast library of content and production expertise to create immersive, multi-layered experiences. Documentaries, behind-the-scenes footage, and athlete-focused mini-series transform simple matches into full-fledged cultural events. Consider what this could look like in practice: exclusive behind-the-scenes documentaries chronicling teams’ journeys to the tournament; halftime performances by global music superstars; interactive features allowing fans to vote on player awards or participate in trivia games during matches, all integrated into Netflix’s platform. All of this could redefine how fans engage with soccer. Strengthening Advertiser Appeal Through Women’s Sports Women’s sports also offer advertisers a unique chance to associate their brands with themes of empowerment, inclusion, and progress—values that resonate strongly with younger audiences, particularly Millennials and Gen Z. Hosting the FIFA Women’s World Cup allows Netflix to position itself as a champion of these ideals while also attracting brands eager to align with this growing cultural movement. For Netflix, this alignment boosts advertising revenue and positions Netflix well. With its tiered ad model already in place, Netflix is well-equipped to capitalize on the advertising opportunities presented by live sports. In effect, the FIFA Women’s World Cup is proving ground for Netflix’s ability to drive revenue from both content and advertising. A Powerful Partnership A partnership between Netflix and the FIFA Women’s World Cup will forge a powerful partnership, blending sports, entertainment, and connected TV for a global audience. The move underlines a broader trend: live sports are becoming a linchpin for connected TV platforms seeking to lock in subscriptions and dominate global markets. Unlike traditional cable or network television, connected TV offers interactivity, data-driven insights, and a chance to engage younger, digitally savvy audiences who demand more than just live coverage. Netflix’s foray into sports capitalizes on these advantages, positioning the platform as a hub where sports meet storytelling.

Netflix Finally Comes For Sports Streaming

A rising number of businesses have adopted the so-called BTC standard, which means that they plan to regularly add Bitcoin to their balance sheet. However, none of these companies are big enough to change the world’s view of Bitcoin as a reserve asset. The only notable Bitcoin holder among big businesses is Tesla, but Elon Musk’s company does not have an active Bitcoin purchase strategy. Microsoft had a chance to become a major corporate player in the crypto market. Its shareholders have recently voted on the NCPPR (National Center for Public Policy Research) proposal to add BTC to the company’s reserves. To be precise, NCPPR urged Microsoft’s board to conduct a study on whether the potential inclusion of BTC as a balance sheet asset would serve the long-term interests of the company’s shareholders. The board of directors, which issues its recommendations before the vote, recommended rejecting the NCPPR proposal. The shareholders voted in line with recommendations and rejected the idea of adding Bitcoin to Microsoft’s balance sheet. Just 0.55% of votes were cast for the proposal to consider BTC as a balance sheet asset. Microsoft’s main shareholders are institutional investors. The largest investors in Microsoft are Vanguard and BlackRock (the one that launched the most successful spot ETF on Bitcoin). They voted against the proposal. Microsoft’s board of directors argued that it had already conducted a thorough analysis of BTC and came to the conclusion that it was not suitable for the company’s balance sheet. According to the board, corporate reserves should include stable and predictable investments to ensure liquidity and operational funding. It’s hard to argue that Bitcoin is a stable investment, like cash. In fact, no one would need BTC if it behaved like cash, which means constantly losing value in real terms, month after month and year after year. Investors buy BTC to preserve and grow their capital in real terms in order to buy more goods, services, and assets in the future. Bitcoin is not suited to guarantee the availability of a predefined amount of money in dollar equivalent at a certain point in time. Apparently, similar proposals would lose in the upcoming shareholder votes as institutional investors would follow the recommendations of the companies’ boards. However, BTC activists will soon learn from the mistakes they made during the first attempt to push big companies to adopt a Bitcoin strategy. Recommending Bitcoin as a substitute for traditional liquidity sources is a clear mistake as Bitcoin is hardly suitable for such purposes. Instead, activists must insist that Bitcoin should find its place in long-term investments of big companies together with their factories, equipment, and intellectual property. Such an approach would eliminate the need to evaluate the current volatility of Bitcoin, therefore depriving the companies’ boards of their key argument against the world’s main cryptocurrency. As shareholders of big companies include investment giants that have already issued spot ETFs on crypto, the chances to break the resistance of traditional finance will grow significantly. A rising number of businesses have adopted the so-called BTC standard, which means that they plan to regularly add Bitcoin to their balance sheet. However, none of these companies are big enough to change the world’s view of Bitcoin as a reserve asset. The only notable Bitcoin holder among big businesses is Tesla, but Elon Musk’s company does not have an active Bitcoin purchase strategy. Microsoft had a chance to become a major corporate player in the crypto market. Its shareholders have recently voted on the NCPPR (National Center for Public Policy Research) proposal to add BTC to the company’s reserves. To be precise, NCPPR urged Microsoft’s board to conduct a study on whether the potential inclusion of BTC as a balance sheet asset would serve the long-term interests of the company’s shareholders. The board of directors, which issues its recommendations before the vote, recommended rejecting the NCPPR proposal. The shareholders voted in line with recommendations and rejected the idea of adding Bitcoin to Microsoft’s balance sheet. Just 0.55% of votes were cast for the proposal to consider BTC as a balance sheet asset. Microsoft’s main shareholders are institutional investors. The largest investors in Microsoft are Vanguard and BlackRock (the one that launched the most successful spot ETF on Bitcoin). They voted against the proposal. Microsoft’s board of directors argued that it had already conducted a thorough analysis of BTC and came to the conclusion that it was not suitable for the company’s balance sheet. According to the board, corporate reserves should include stable and predictable investments to ensure liquidity and operational funding. It’s hard to argue that Bitcoin is a stable investment, like cash. In fact, no one would need BTC if it behaved like cash, which means constantly losing value in real terms, month after month and year after year. Investors buy BTC to preserve and grow their capital in real terms in order to buy more goods, services, and assets in the future. Bitcoin is not suited to guarantee the availability of a predefined amount of money in dollar equivalent at a certain point in time. Apparently, similar proposals would lose in the upcoming shareholder votes as institutional investors would follow the recommendations of the companies’ boards. However, BTC activists will soon learn from the mistakes they made during the first attempt to push big companies to adopt a Bitcoin strategy. Recommending Bitcoin as a substitute for traditional liquidity sources is a clear mistake as Bitcoin is hardly suitable for such purposes. Instead, activists must insist that Bitcoin should find its place in long-term investments of big companies together with their factories, equipment, and intellectual property. Such an approach would eliminate the need to evaluate the current volatility of Bitcoin, therefore depriving the companies’ boards of their key argument against the world’s main cryptocurrency. As shareholders of big companies include investment giants that have already issued spot ETFs on crypto, the chances to break the resistance of traditional finance will grow significantly.

Microsoft Defeats Proposal Asking Company to Add Bitcoin to Its Balance Sheet

Microsoft ukax Propuesta ukaruw atipt’i, Empresarux Bitcoin ukar Balance ukar yapxatañapatak mayi

মাইক্রোসফ্ট তার ব্যালেন্স শীটে বিটকয়েন যুক্ত করার জন্য কোম্পানিকে অনুরোধ করা প্রস্তাবকে পরাজিত করেছে

Microsoft rechaza propuesta que pedía a la empresa añadir Bitcoin a su balance

Microsoft-ek proposamena gainditzen du konpainiari Bitcoin bere balantzean gehitzeko eskatuz

מיקרוסופט מביסה את ההצעה המבקשת מחברה להוסיף ביטקוין למאזנה שלה

マイクロソフト、ビットコインをバランスシートに加えるよう求める提案を否決

Microsoft alongi Proposition oyo esengaki na entreprise ebakisa Bitcoin na bilan na yango

„Microsoft“ atmetė pasiūlymą, prašydama įmonės įtraukti Bitcoin į savo balansą

Microsoft отклонила предложение, призывающее компанию добавить биткоин в свой баланс

Microsoft besegrar förslag som ber företaget att lägga till Bitcoin till sin balansräkning

Microsoft пешниҳоди аз ширкат хоҳиши илова кардани Bitcoin ба варақаи тавозунашро рад кард

Microsoft відхиляє пропозицію про додавання біткойна до свого балансу

Table of Links Abstract and Introduction


Domain and Task
2.1. Data sources and complexity
2.2. Task definition


Related Work
3.1. Text mining and NLP research overview
3.2. Text mining and NLP in industry use
3.3. Text mining and NLP for procurement
3.4. Conclusion from literature review


Proposed Methodology
4.1. Domain knowledge
4.2. Content extraction
4.3. Lot zoning
4.4. Lot item detection
4.5. Lot parsing
4.6. XML parsing, data joining, and risk indices development


Experiment and Demonstration
5.1. Component evaluation
5.2. System demonstration


Discussion
6.1. The ‘industry’ focus of the project
6.2. Data heterogeneity, multilingual and multi-task nature
6.3. The dilemma of algorithmic choices
6.4. The cost of training data


Conclusion, Acknowledgements, and References 4.1. Domain knowledge With the semi-structured data available from TED tender XMLs and award XMLs, Vamstar developed XML parsers that extract different fields and store their unique values in a temporary data store (to be called the ‘tender and award fields’ or just ‘fields’ in short). As an example, given the XML in Figure 1, a record would be created with fields such as: buyer name and addresses, lots (e.g., ‘Verapamil solution for inj. Lot No: 62’ but with multiple values as the notice contains more than 90 lots), award criteria, etc. This allows generation of useful domain knowledge in different ways. First, training data for different fields can be created by simply taking the unique values from each field. These values usually take the form of short phrases or sentences, and therefore, can be used to train text classifiers at phase/sentence level. Note that this data is multilingual. Second, domain lexicons for different fields can be built by extracting ‘representative’ words from each field. Given a field that we are interested in, we merge all values from different languages, and use machine translation tools to translate them into English. Instead of drawing a boundary between what is and is not a ‘representative’ word for that field, here we explain a process to calculate the ‘specificities’ of words and we refer to this as ‘domain specificity lexicon’, which is used in other components of our method. Given a word 𝑤, the basic idea is to measure how unique 𝑤 is to lot/item descriptions compared to other text content in a tender. The basic process is as follows: Identify the set of ‘fields’ that are most relevant to a tender. For this we selected 5 fields - this is rather a subjective interpretation by the domain experts: ‘Name and addresses’ of buyers (Section I.1.1 of the notice in Figure 1), ‘notice title’ (II.1.1), ‘short description’ (II.1.4), ‘lot and item descriptions’ (a concatenation of values from fields like II.2.1), ‘contract criteria’ (a concatenation of values from fields like II.2.5). Following the process mentioned above, all values are then translated into English;


For each field, create a bag-of-words (‘BOW’) representation by concatenating the values of that field across all records in the database, applying stop words removal and lowercasing;


For 𝑤, calculate 4 metrics as follows: 𝑛𝑡𝑓(𝑤) normalised frequency of 𝑤 in the BOW representation of the ‘lot and item descriptions’ field, calculated as the ratio between the frequency of 𝑤 in the BOW over the total words in the BOW; 𝑛𝑑𝑓(𝑤) the normalised ‘document frequency’ of 𝑤, calculated as the ratio between the number of fields in which 𝑤 is found over the number of all fields (i.e., 5); 𝑛𝑡𝑓ᐧ𝑛𝑑𝑓(𝑤) , inspired by the tf-idf measure used in document retrieval, this is the product − of 𝑛𝑡𝑓(𝑤) and the inverse of 𝑛𝑡𝑓(𝑤); 𝑤𝑒𝑖𝑟𝑑𝑛𝑒𝑠𝑠(𝑤), a score calculated by comparing 𝑛𝑡𝑓(𝑤) against that calculated in a reference, general purpose corpus (in this case, the Brown corpus). Thus our domain specificity lexicon contains unique words found within the ‘lot and item descriptions’ field, and each word has four associated scores. In addition to this, we also use another two dictionaries compiled by Vamstar. One is a list of words often used as measurement units (e.g., mg, ml), and the other is a list of words often used to describe the ‘form’ of required items (e.g., pack, box, bottle). Both are very short and contain only a few dozens of entries. These lexicons and dictionaries are also translated to other languages using Google Translate. Authors:
(1) Ziqi Zhang*, Information School, the University of Sheffield, Regent Court, Sheffield, UKS1 4DP (Ziqi.Zhang@sheffield.ac.uk);
(2) Tomas Jasaitis, Vamstar Ltd., London (Tomas.Jasaitis@vamstar.io);
(3) Richard Freeman, Vamstar Ltd., London (Richard.Freeman@vamstar.io);
(4) Rowida Alfrjani, Information School, the University of Sheffield, Regent Court, Sheffield, UKS1 4DP (Rowida.Alfrjani@sheffield.ac.uk);
(5) Adam Funk, Information School, the University of Sheffield, Regent Court, Sheffield, UKS1 4DP (Adam.Funk@sheffield.ac.uk). This paper is available on arxiv under CC BY 4.0 license. Table of Links Abstract and Introduction Domain and Task
2.1. Data sources and complexity
2.2. Task definition Related Work
3.1. Text mining and NLP research overview
3.2. Text mining and NLP in industry use
3.3. Text mining and NLP for procurement
3.4. Conclusion from literature review Proposed Methodology
4.1. Domain knowledge
4.2. Content extraction
4.3. Lot zoning
4.4. Lot item detection
4.5. Lot parsing
4.6. XML parsing, data joining, and risk indices development Experiment and Demonstration
5.1. Component evaluation
5.2. System demonstration Discussion
6.1. The ‘industry’ focus of the project
6.2. Data heterogeneity, multilingual and multi-task nature
6.3. The dilemma of algorithmic choices
6.4. The cost of training data Conclusion, Acknowledgements, and References Abstract and Introduction Abstract and Introduction Abstract and Introduction Abstract and Introduction Domain and Task 2.1. Data sources and complexity 2.2. Task definition Domain and Task Domain and Task 2.1. Data sources and complexity 2.1. Data sources and complexity 2.2. Task definition 2.2. Task definition Related Work 3.1. Text mining and NLP research overview 3.2. Text mining and NLP in industry use 3.3. Text mining and NLP for procurement 3.4. Conclusion from literature review Related Work Related Work 3.1. Text mining and NLP research overview 3.1. Text mining and NLP research overview 3.2. Text mining and NLP in industry use 3.2. Text mining and NLP in industry use 3.3. Text mining and NLP for procurement 3.3. Text mining and NLP for procurement 3.4. Conclusion from literature review 3.4. Conclusion from literature review Proposed Methodology 4.1. Domain knowledge 4.2. Content extraction 4.3. Lot zoning 4.4. Lot item detection 4.5. Lot parsing 4.6. XML parsing, data joining, and risk indices development Proposed Methodology Proposed Methodology Proposed Methodology 4.1. Domain knowledge 4.1. Domain knowledge 4.2. Content extraction 4.2. Content extraction 4.3. Lot zoning 4.3. Lot zoning 4.4. Lot item detection 4.4. Lot item detection 4.5. Lot parsing 4.5. Lot parsing 4.6. XML parsing, data joining, and risk indices development 4.6. XML parsing, data joining, and risk indices development Experiment and Demonstration 5.1. Component evaluation 5.2. System demonstration Experiment and Demonstration Experiment and Demonstration 5.1. Component evaluation 5.1. Component evaluation 5.2. System demonstration 5.2. System demonstration Discussion 6.1. The ‘industry’ focus of the project 6.2. Data heterogeneity, multilingual and multi-task nature 6.3. The dilemma of algorithmic choices 6.4. The cost of training data Discussion Discussion 6.1. The ‘industry’ focus of the project 6.1. The ‘industry’ focus of the project 6.2. Data heterogeneity, multilingual and multi-task nature 6.2. Data heterogeneity, multilingual and multi-task nature 6.3. The dilemma of algorithmic choices 6.3. The dilemma of algorithmic choices 6.4. The cost of training data 6.4. The cost of training data Conclusion, Acknowledgements, and References Conclusion, Acknowledgements, and References Conclusion, Acknowledgements, and References Conclusion, Acknowledgements, and References 4.1. Domain knowledge With the semi-structured data available from TED tender XMLs and award XMLs, Vamstar developed XML parsers that extract different fields and store their unique values in a temporary data store (to be called the ‘tender and award fields’ or just ‘fields’ in short). As an example, given the XML in Figure 1, a record would be created with fields such as: buyer name and addresses, lots (e.g., ‘Verapamil solution for inj. Lot No: 62’ but with multiple values as the notice contains more than 90 lots), award criteria, etc. This allows generation of useful domain knowledge in different ways. First, training data for different fields can be created by simply taking the unique values from each field. These values usually take the form of short phrases or sentences, and therefore, can be used to train text classifiers at phase/sentence level. Note that this data is multilingual. training data Second, domain lexicons for different fields can be built by extracting ‘representative’ words from each field. Given a field that we are interested in, we merge all values from different languages, and use machine translation tools to translate them into English. Instead of drawing a boundary between what is and is not a ‘representative’ word for that field, here we explain a process to calculate the ‘specificities’ of words and we refer to this as ‘domain specificity lexicon’, which is used in other components of our method. Given a word 𝑤, the basic idea is to measure how unique 𝑤 is to lot/item descriptions compared to other text content in a tender. The basic process is as follows: domain lexicons Identify the set of ‘fields’ that are most relevant to a tender. For this we selected 5 fields - this is rather a subjective interpretation by the domain experts: ‘Name and addresses’ of buyers (Section I.1.1 of the notice in Figure 1), ‘notice title’ (II.1.1), ‘short description’ (II.1.4), ‘lot and item descriptions’ (a concatenation of values from fields like II.2.1), ‘contract criteria’ (a concatenation of values from fields like II.2.5). Following the process mentioned above, all values are then translated into English; For each field, create a bag-of-words (‘BOW’) representation by concatenating the values of that field across all records in the database, applying stop words removal and lowercasing; For 𝑤, calculate 4 metrics as follows: 𝑛𝑡𝑓(𝑤) normalised frequency of 𝑤 in the BOW representation of the ‘lot and item descriptions’ field, calculated as the ratio between the frequency of 𝑤 in the BOW over the total words in the BOW; 𝑛𝑑𝑓(𝑤) the normalised ‘document frequency’ of 𝑤, calculated as the ratio between the number of fields in which 𝑤 is found over the number of all fields (i.e., 5); 𝑛𝑡𝑓ᐧ𝑛𝑑𝑓(𝑤) , inspired by the tf-idf measure used in document retrieval, this is the product − of 𝑛𝑡𝑓(𝑤) and the inverse of 𝑛𝑡𝑓(𝑤); 𝑤𝑒𝑖𝑟𝑑𝑛𝑒𝑠𝑠(𝑤), a score calculated by comparing 𝑛𝑡𝑓(𝑤) against that calculated in a reference, general purpose corpus (in this case, the Brown corpus). Identify the set of ‘fields’ that are most relevant to a tender. For this we selected 5 fields - this is rather a subjective interpretation by the domain experts: ‘Name and addresses’ of buyers (Section I.1.1 of the notice in Figure 1), ‘notice title’ (II.1.1), ‘short description’ (II.1.4), ‘lot and item descriptions’ (a concatenation of values from fields like II.2.1), ‘contract criteria’ (a concatenation of values from fields like II.2.5). Following the process mentioned above, all values are then translated into English; Identify the set of ‘fields’ that are most relevant to a tender. For this we selected 5 fields - this is rather a subjective interpretation by the domain experts: ‘Name and addresses’ of buyers (Section I.1.1 of the notice in Figure 1), ‘notice title’ (II.1.1), ‘short description’ (II.1.4), ‘lot and item descriptions’ (a concatenation of values from fields like II.2.1), ‘contract criteria’ (a concatenation of values from fields like II.2.5). Following the process mentioned above, all values are then translated into English; For each field, create a bag-of-words (‘BOW’) representation by concatenating the values of that field across all records in the database, applying stop words removal and lowercasing; For each field, create a bag-of-words (‘BOW’) representation by concatenating the values of that field across all records in the database, applying stop words removal and lowercasing; For 𝑤, calculate 4 metrics as follows: 𝑛𝑡𝑓(𝑤) normalised frequency of 𝑤 in the BOW representation of the ‘lot and item descriptions’ field, calculated as the ratio between the frequency of 𝑤 in the BOW over the total words in the BOW; 𝑛𝑑𝑓(𝑤) the normalised ‘document frequency’ of 𝑤, calculated as the ratio between the number of fields in which 𝑤 is found over the number of all fields (i.e., 5); 𝑛𝑡𝑓ᐧ𝑛𝑑𝑓(𝑤) , inspired by the tf-idf measure used in document retrieval, this is the product − of 𝑛𝑡𝑓(𝑤) and the inverse of 𝑛𝑡𝑓(𝑤); 𝑤𝑒𝑖𝑟𝑑𝑛𝑒𝑠𝑠(𝑤), a score calculated by comparing 𝑛𝑡𝑓(𝑤) against that calculated in a reference, general purpose corpus (in this case, the Brown corpus). For 𝑤, calculate 4 metrics as follows: 𝑛𝑡𝑓(𝑤) normalised frequency of 𝑤 in the BOW representation of the ‘lot and item descriptions’ field, calculated as the ratio between the frequency of 𝑤 in the BOW over the total words in the BOW; 𝑛𝑑𝑓(𝑤) the normalised ‘document frequency’ of 𝑤, calculated as the ratio between the number of fields in which 𝑤 is found over the number of all fields (i.e., 5); 𝑛𝑡𝑓ᐧ𝑛𝑑𝑓(𝑤) , inspired by the tf-idf measure used in document retrieval, this is the product − of 𝑛𝑡𝑓(𝑤) and the inverse of 𝑛𝑡𝑓(𝑤); 𝑤𝑒𝑖𝑟𝑑𝑛𝑒𝑠𝑠(𝑤), a score calculated by comparing 𝑛𝑡𝑓(𝑤) against that calculated in a reference, general purpose corpus (in this case, the Brown corpus). Thus our domain specificity lexicon contains unique words found within the ‘lot and item descriptions’ field, and each word has four associated scores. In addition to this, we also use another two dictionaries compiled by Vamstar. One is a list of words often used as measurement units (e.g., mg, ml), and the other is a list of words often used to describe the ‘form’ of required items (e.g., pack, box, bottle). Both are very short and contain only a few dozens of entries. These lexicons and dictionaries are also translated to other languages using Google Translate. Authors: (1) Ziqi Zhang*, Information School, the University of Sheffield, Regent Court, Sheffield, UKS1 4DP (Ziqi.Zhang@sheffield.ac.uk); (2) Tomas Jasaitis, Vamstar Ltd., London (Tomas.Jasaitis@vamstar.io); (3) Richard Freeman, Vamstar Ltd., London (Richard.Freeman@vamstar.io); (4) Rowida Alfrjani, Information School, the University of Sheffield, Regent Court, Sheffield, UKS1 4DP (Rowida.Alfrjani@sheffield.ac.uk); (5) Adam Funk, Information School, the University of Sheffield, Regent Court, Sheffield, UKS1 4DP (Adam.Funk@sheffield.ac.uk). Authors: Authors: (1) Ziqi Zhang*, Information School, the University of Sheffield, Regent Court, Sheffield, UKS1 4DP (Ziqi.Zhang@sheffield.ac.uk); (2) Tomas Jasaitis, Vamstar Ltd., London (Tomas.Jasaitis@vamstar.io); (3) Richard Freeman, Vamstar Ltd., London (Richard.Freeman@vamstar.io); (4) Rowida Alfrjani, Information School, the University of Sheffield, Regent Court, Sheffield, UKS1 4DP (Rowida.Alfrjani@sheffield.ac.uk); (5) Adam Funk, Information School, the University of Sheffield, Regent Court, Sheffield, UKS1 4DP (Adam.Funk@sheffield.ac.uk). This paper is available on arxiv under CC BY 4.0 license. This paper is available on arxiv under CC BY 4.0 license. available on arxiv available on arxiv

How Vamstar Makes Sense of Complex Tender Data

Curaçao, Fergusonweg, December 24th, 2024, Chainwire/--BetFury transferred $750,000 of liquidity to PancakeSwap, making the leading DEX the primary liquidity holder for BFG tokens. This collaboration introduces the BFG-CAKE Syrup Pool, a time-limited Trading Competition, CAKE integration in BetFury Futures, and more. Additionally, BetFury has launched a special CAKE Staking Pool exclusively for new users, accessible with a unique promo code provided in this article. BetFury & PancakeSwap collaboration Two projects from diverse niches have united with one goal: to offer new crypto-earning opportunities. The projects have a shared vision: bridging the gap between gaming enthusiasts and crypto traders. BetFury is connected to DeFi: it offers a wide range of crypto features like Crypto Staking with up to 60% APR and Futures, which has expanded with the CAKE token integration. $20,000 BFG-CAKE Syrup Pool | New way to earn BFG Cakes are even sweeter with syrup, especially when served generously. That’s why the BFG Syrup Pool offers a whopping $20,000 total capacity. PancakeSwap users can stake CAKE for 90 days to earn BFG tokens, with over 20% APR based on the number of participants. The only limit is up to 100 CAKE per wallet. The next step: with access to over 8,000 slots, 20 exclusive in-house games, and bets across 50+ sports categories. By holding just 100 BFG, users can unlock BFG Staking rewards in BTC, ETH, USDT, BNB, and TRX—maximizing crypto-earning potential. $50,000 Trading Competition | Leaderboard & Extra rewards The Trading Competition marks the second phase of the collaboration. With a $50,000 prize pool in USDT and BFG tokens, BetFury invites all traders of the BFG-CAKE pair to compete. A leaderboard featuring the top 15 winners will be revealed at the end, while 15-20 randomly chosen participants will receive bonus rewards. The competition runs for 20 days only, from December 24 to January 12. Exclusive CAKE Staking Pool for New Users on BetFury BetFury is excited to launch the exclusive CAKE Staking Pool, offering up to 40% APR. This pool is tailored for new users who can access it using the BFCAKE code. The CAKE Staking Pool has already started and is available for 7 days on the BetFury platform. During this period, participants can earn BFG on PancakeSwap and CAKE on BetFury, boosting their crypto holdings and gaining an edge in the Trading Competition. About BetFury BetFury is a trusted crypto and entertainment ecosystem that has been at the forefront of iGaming, crypto-earning features, and sports betting for over five years. With a global community of over 3 million registered users, BetFury offers a wide array of competitive products, including original games with up to 99.28% RTP, BFG Staking, Crypto Staking, NFT Lootboxes, an Affiliate Program, and more. To enhance user experience, BetFury also features a Telegram Bot that boosts crypto rewards and ensures quick, seamless access to the platform. Adding to its appeal is a lucrative Referral Program, where referrers can earn up to $1,500 in bonuses and a 30% commission from their referrals' activity, with the option to share rewards. Contact PR Manager Alisia Preston BetFury pr@betfury.com This story was distributed as a release by Chainwire under HackerNoon’s Business Blogging Program. Learn more about the program here Curaçao, Fergusonweg, December 24th, 2024, Chainwire/--BetFury transferred $750,000 of liquidity to PancakeSwap, making the leading DEX the primary liquidity holder for BFG tokens. This collaboration introduces the BFG-CAKE Syrup Pool, a time-limited Trading Competition, CAKE integration in BetFury Futures, and more. Additionally, BetFury has launched a special CAKE Staking Pool exclusively for new users, accessible with a unique promo code provided in this article. BetFury & PancakeSwap collaboration Two projects from diverse niches have united with one goal: to offer new crypto-earning opportunities. The projects have a shared vision: bridging the gap between gaming enthusiasts and crypto traders. BetFury is connected to DeFi: it offers a wide range of crypto features like Crypto Staking with up to 60% APR and Futures , which has expanded with the CAKE token integration. Futures Futures $20,000 BFG-CAKE Syrup Pool | New way to earn BFG Cakes are even sweeter with syrup, especially when served generously. That’s why the BFG Syrup Pool offers a whopping $20,000 total capacity. PancakeSwap users can stake CAKE for 90 days to earn BFG tokens, with over 20% APR based on the number of participants. The only limit is up to 100 CAKE per wallet. BFG Syrup Pool BFG Syrup Pool The next step: with access to over 8,000 slots, 20 exclusive in-house games, and bets across 50+ sports categories. By holding just 100 BFG, users can unlock BFG Staking rewards in BTC, ETH, USDT, BNB, and TRX—maximizing crypto-earning potential. $50,000 Trading Competition | Leaderboard & Extra rewards The Trading Competition marks the second phase of the collaboration. With a $50,000 prize pool in USDT and BFG tokens, BetFury invites all traders of the BFG-CAKE pair to compete. Trading Competition Trading Competition A leaderboard featuring the top 15 winners will be revealed at the end, while 15-20 randomly chosen participants will receive bonus rewards. The competition runs for 20 days only, from December 24 to January 12. Exclusive CAKE Staking Pool for New Users on BetFury BetFury is excited to launch the exclusive CAKE Staking Pool, offering up to 40% APR. This pool is tailored for new users who can access it using the BFCAKE code . The CAKE Staking Pool has already started and is available for 7 days on the BetFury platform. BFCAKE code BFCAKE code During this period, participants can earn BFG on PancakeSwap and CAKE on BetFury, boosting their crypto holdings and gaining an edge in the Trading Competition. About BetFury BetFury is a trusted crypto and entertainment ecosystem that has been at the forefront of iGaming, crypto-earning features, and sports betting for over five years. BetFury BetFury With a global community of over 3 million registered users, BetFury offers a wide array of competitive products, including original games with up to 99.28% RTP, BFG Staking, Crypto Staking, NFT Lootboxes, an Affiliate Program, and more. To enhance user experience, BetFury also features a Telegram Bot that boosts crypto rewards and ensures quick, seamless access to the platform. Adding to its appeal is a lucrative Referral Program, where referrers can earn up to $1,500 in bonuses and a 30% commission from their referrals' activity, with the option to share rewards. Contact PR Manager Alisia Preston BetFury pr@betfury.com This story was distributed as a release by Chainwire under HackerNoon’s Business Blogging Program. Learn more about the program here This story was distributed as a release by Chainwire under HackerNoon’s Business Blogging Program. Learn more about the program here here here

BetFury x PancakeSwap Partnership: $20K BFG Syrup Pool, $50K Trading Competition And More

Asociación entre BetFury y PancakeSwap: fondo de $20 000 en BFG Syrup, competencia de trading de $50 000 y más

BetFury x PancakeSwap -kumppanuus: 20 000 dollarin BFG-siirappiallas, 50 000 dollarin kauppakilpailu ja paljon muuta

שותפות BetFury x PancakeSwap: מאגר סירופ BFG בסך 20K$, תחרות מסחר של 50K$ ועוד

BetFury x PancakeSwap パートナーシップ: 2 万ドルの BFG シロップ プール、5 万ドルのトレーディング コンテストなど

BetFury x PancakeSwap серіктестігі: $20K BFG сироп бассейні, $50K сауда бәсекесі және т.б.

Партнерство BetFury x PancakeSwap: 20.000 $ BFG сируп базен, 50.000 $ тргување конкуренција и повеќе

Perkongsian BetFury x PancakeSwap: Kolam Sirap BFG $20K, Pertandingan Dagangan $50K Dan Banyak Lagi

BetFury x PancakeSwap Asociación: $20K BFG Jarabe Piscina, $50K Comercio Competencia Y Astawan

Parteneriat BetFury x PancakeSwap: 20.000 USD BFG Syrup Pool, 50.000 USD Competiție de tranzacționare și multe altele

BetFury x PancakeSwap හවුල්කාරිත්වය: $20K BFG සිරප් තටාකය, $50K වෙළඳ තරඟය සහ තවත් දේ

BetFury x PancakeSwap Partnership: $20K BFG Syrup Pool, $50k Tartanka Ganacsiga iyo in ka badan

Partneriteti BetFury x PancakeSwap: 20K $ BFG Shurup Pool, $50K Konkurrencë Tregtare dhe më shumë

Table of Links Abstract and 1. Introduction 2. Preliminaries and 2.1. Blind deconvolution 2.2. Quadratic neural networks 3. Methodology 3.1. Time domain quadratic convolutional filter 3.2. Superiority of cyclic features extraction by QCNN 3.3. Frequency domain linear filter with envelope spectrum objective function 3.4. Integral optimization with uncertainty-aware weighing scheme 4. Computational experiments 4.1. Experimental configurations 4.2. Case study 1: PU dataset 4.3. Case study 2: JNU dataset 4.4. Case study 3: HIT dataset 5. Computational experiments 5.1. Comparison of BD methods 5.2. Classification results on various noise conditions 5.3. Employing ClassBD to deep learning classifiers 5.4. Employing ClassBD to machine learning classifiers 5.5. Feature extraction ability of quadratic and conventional networks 5.6. Comparison of ClassBD filters 6. Conclusions Appendix and References 4.2. Case study 1: PU dataset 4.2.1. Dataset description The PU dataset was collected by Paderborn University (PU) Bearing Data Center [79]. This dataset encompasses multiple faults for 32 bearings, including vibration and current signals. The bearings were categorized into three groups: i) Six healthy bearings; ii) Twelve bearings with manually-induced damage (seven with outer race faults, five with inner race faults); iii) Fourteen bearings with real damage, induced by accelerated lifetime tests (five with outer race faults, six with inner race faults, and three with multiple faults). Four operating conditions were implemented, varying the rotational speed (N = 1500rpm or N = 900rpm), load torque (M = 0.7Nm or M = 0.1Nm), and radial force (F = 1000N or F = 400N). The sampling frequency was set at 64KHz. This dataset is one of the more difficult datasets in the field of bearing fault diagnosis [71]. In this study, we exclusively utilize the real damaged bearings for classification to validate the methods in real-world scenarios. We classify the fourteen faulty bearings and six healthy bearings into fourteen categories, as illustrated in Table 3. All the four operating conditions are tested in the experiments, with the codes assigned as N09M07F10, N15M01F10, N15M07F4, and N15M07F10. Given the large volume of data in the PU dataset, each bearing has 20 segments, so we do not use overlapping to construct the datasets. We slice 20 × 2048 sub-segments with a stride of 2048, resulting in 400 data per bearing. The total number of datasets is 7600 × 2048. The data collected in the first 19 segments are randomly divided into training and validation sets at a ratio of 0.8:0.2, while the data from the 20th segment are allocated to the test set. Consequently, the dataset includes 5776 training data, 1444 validation data, and 380 test data. Lastly, we established four SNR levels (-4dB, -2dB, 0dB, 2dB) to evaluate the diagnosis performance. 4.2.2. Classification results The classification results are presented in Table 4. We highlight key findings from the analysis. Notably, the ClassBD model demonstrates superior performance compared to its competitors across various noise levels and operating conditions. Overall, the average F1 scores of ClassBD on the four conditions are higher than 94%. Specifically, under high noise conditions (at -4 dB), the ClassBD exhibits a substantial performance gap relative to other methods. For instance, on the N15M01F10 dataset with -4dB noise, the ClassBD achieves an impressive 95% F1 score, while the second-best method, EWSNet, only attains 70% F1. These results underscore the efficacy of ClassBD as a robust anti-noise model, consistently delivering competitive performance across diverse high-noise scenarios. 4.2.3. Classification under small sample conditions In this study, we rigorously assess the performance of various methods under the challenging scenario of extremely limited sample availability. While our approach is not explicitly tailored for small sample issues, we conduct straightforward tests to gain insights into their behavior. Specifically, we employ the N09M07F10 dataset, comprising 20 signal segments per category. In the most stringent case, we retain only one sample from each signal segment, reserving the last segment of the signal as the test set. Consequently, we collect one to three samples from each signal segment and finally compose 15, 30, and 45 samples per class for training sets. Notably, we exclude noise from this experiment. The results, depicted in Figure 3, reveal that even under these constrained conditions, the ClassBD, EWSNet, and DRSN models exhibit commendable performance. Remarkably, all three methods achieve over 90% F1 scores with a training dataset of just 15 samples per class. Notably, the performance of ClassBD and EWSNet closely align. Their average F1 scores stand at 97.70% and 97.47%, respectively, positioning them as the top-performing approaches. In summary, our findings underscore the promising potential of ClassBD in addressing the challenges posed by small sample sizes Authors:
(1) Jing-Xiao Liao, Department of Industrial and Systems Engineering, The Hong Kong Polytechnic University, Hong Kong, Special Administrative Region of China and School of Instrumentation Science and Engineering, Harbin Institute of Technology, Harbin, China;
(2) Chao He, School of Mechanical, Electronic and Control Engineering, Beijing Jiaotong University, Beijing, China;
(3) Jipu Li, Department of Industrial and Systems Engineering, The Hong Kong Polytechnic University, Hong Kong, Special Administrative Region of China;
(4) Jinwei Sun, School of Instrumentation Science and Engineering, Harbin Institute of Technology, Harbin, China;
(5) Shiping Zhang (Corresponding author), School of Instrumentation Science and Engineering, Harbin Institute of Technology, Harbin, China;
(6) Xiaoge Zhang (Corresponding author), Department of Industrial and Systems Engineering, The Hong Kong Polytechnic University, Hong Kong, Special Administrative Region of China. This paper is available on arxiv under CC by 4.0 Deed (Attribution 4.0 International) license. Table of Links Abstract and 1. Introduction Abstract and 1. Introduction 2. Preliminaries and 2.1. Blind deconvolution 2. Preliminaries and 2.1. Blind deconvolution 2.2. Quadratic neural networks 2.2. Quadratic neural networks 3. Methodology 3. Methodology 3.1. Time domain quadratic convolutional filter 3.1. Time domain quadratic convolutional filter 3.2. Superiority of cyclic features extraction by QCNN 3.2. Superiority of cyclic features extraction by QCNN 3.3. Frequency domain linear filter with envelope spectrum objective function 3.3. Frequency domain linear filter with envelope spectrum objective function 3.4. Integral optimization with uncertainty-aware weighing scheme 3.4. Integral optimization with uncertainty-aware weighing scheme 4. Computational experiments 4.1. Experimental configurations 4.1. Experimental configurations 4.2. Case study 1: PU dataset 4.2. Case study 1: PU dataset 4.3. Case study 2: JNU dataset 4.3. Case study 2: JNU dataset 4.4. Case study 3: HIT dataset 4.4. Case study 3: HIT dataset 5. Computational experiments 5.1. Comparison of BD methods 5.1. Comparison of BD methods 5.2. Classification results on various noise conditions 5.2. Classification results on various noise conditions 5.3. Employing ClassBD to deep learning classifiers 5.3. Employing ClassBD to deep learning classifiers 5.4. Employing ClassBD to machine learning classifiers 5.4. Employing ClassBD to machine learning classifiers 5.5. Feature extraction ability of quadratic and conventional networks 5.5. Feature extraction ability of quadratic and conventional networks 5.6. Comparison of ClassBD filters 5.6. Comparison of ClassBD filters 6. Conclusions 6. Conclusions Appendix and References Appendix and References 4.2. Case study 1: PU dataset 4.2.1. Dataset description 4.2.1. Dataset description 4.2.1. Dataset description The PU dataset was collected by Paderborn University (PU) Bearing Data Center [79]. This dataset encompasses multiple faults for 32 bearings, including vibration and current signals. The bearings were categorized into three groups: i) Six healthy bearings; ii) Twelve bearings with manually-induced damage (seven with outer race faults, five with inner race faults); iii) Fourteen bearings with real damage, induced by accelerated lifetime tests (five with outer race faults, six with inner race faults, and three with multiple faults). Four operating conditions were implemented, varying the rotational speed (N = 1500rpm or N = 900rpm), load torque (M = 0.7Nm or M = 0.1Nm), and radial force (F = 1000N or F = 400N). The sampling frequency was set at 64KHz. This dataset is one of the more difficult datasets in the field of bearing fault diagnosis [71]. In this study, we exclusively utilize the real damaged bearings for classification to validate the methods in real-world scenarios. We classify the fourteen faulty bearings and six healthy bearings into fourteen categories, as illustrated in Table 3. All the four operating conditions are tested in the experiments, with the codes assigned as N09M07F10, N15M01F10, N15M07F4, and N15M07F10. Given the large volume of data in the PU dataset, each bearing has 20 segments, so we do not use overlapping to construct the datasets. We slice 20 × 2048 sub-segments with a stride of 2048, resulting in 400 data per bearing. The total number of datasets is 7600 × 2048. The data collected in the first 19 segments are randomly divided into training and validation sets at a ratio of 0.8:0.2, while the data from the 20th segment are allocated to the test set. Consequently, the dataset includes 5776 training data, 1444 validation data, and 380 test data. Lastly, we established four SNR levels (-4dB, -2dB, 0dB, 2dB) to evaluate the diagnosis performance. 4.2.2. Classification results 4.2.2. Classification results 4.2.2. Classification results The classification results are presented in Table 4. We highlight key findings from the analysis. Notably, the ClassBD model demonstrates superior performance compared to its competitors across various noise levels and operating conditions. Overall, the average F1 scores of ClassBD on the four conditions are higher than 94%. Specifically, under high noise conditions (at -4 dB), the ClassBD exhibits a substantial performance gap relative to other methods. For instance, on the N15M01F10 dataset with -4dB noise, the ClassBD achieves an impressive 95% F1 score, while the second-best method, EWSNet, only attains 70% F1. These results underscore the efficacy of ClassBD as a robust anti-noise model, consistently delivering competitive performance across diverse high-noise scenarios. 4.2.3. Classification under small sample conditions 4.2.3. Classification under small sample conditions 4.2.3. Classification under small sample conditions In this study, we rigorously assess the performance of various methods under the challenging scenario of extremely limited sample availability. While our approach is not explicitly tailored for small sample issues, we conduct straightforward tests to gain insights into their behavior. Specifically, we employ the N09M07F10 dataset, comprising 20 signal segments per category. In the most stringent case, we retain only one sample from each signal segment, reserving the last segment of the signal as the test set. Consequently, we collect one to three samples from each signal segment and finally compose 15, 30, and 45 samples per class for training sets. Notably, we exclude noise from this experiment. The results, depicted in Figure 3, reveal that even under these constrained conditions, the ClassBD, EWSNet, and DRSN models exhibit commendable performance. Remarkably, all three methods achieve over 90% F1 scores with a training dataset of just 15 samples per class. Notably, the performance of ClassBD and EWSNet closely align. Their average F1 scores stand at 97.70% and 97.47%, respectively, positioning them as the top-performing approaches. In summary, our findings underscore the promising potential of ClassBD in addressing the challenges posed by small sample sizes Authors: (1) Jing-Xiao Liao, Department of Industrial and Systems Engineering, The Hong Kong Polytechnic University, Hong Kong, Special Administrative Region of China and School of Instrumentation Science and Engineering, Harbin Institute of Technology, Harbin, China; (2) Chao He, School of Mechanical, Electronic and Control Engineering, Beijing Jiaotong University, Beijing, China; (3) Jipu Li, Department of Industrial and Systems Engineering, The Hong Kong Polytechnic University, Hong Kong, Special Administrative Region of China; (4) Jinwei Sun, School of Instrumentation Science and Engineering, Harbin Institute of Technology, Harbin, China; (5) Shiping Zhang (Corresponding author), School of Instrumentation Science and Engineering, Harbin Institute of Technology, Harbin, China; (6) Xiaoge Zhang (Corresponding author), Department of Industrial and Systems Engineering, The Hong Kong Polytechnic University, Hong Kong, Special Administrative Region of China. Authors: Authors: (1) Jing-Xiao Liao, Department of Industrial and Systems Engineering, The Hong Kong Polytechnic University, Hong Kong, Special Administrative Region of China and School of Instrumentation Science and Engineering, Harbin Institute of Technology, Harbin, China; (2) Chao He, School of Mechanical, Electronic and Control Engineering, Beijing Jiaotong University, Beijing, China; (3) Jipu Li, Department of Industrial and Systems Engineering, The Hong Kong Polytechnic University, Hong Kong, Special Administrative Region of China; (4) Jinwei Sun, School of Instrumentation Science and Engineering, Harbin Institute of Technology, Harbin, China; (5) Shiping Zhang (Corresponding author), School of Instrumentation Science and Engineering, Harbin Institute of Technology, Harbin, China; (6) Xiaoge Zhang (Corresponding author), Department of Industrial and Systems Engineering, The Hong Kong Polytechnic University, Hong Kong, Special Administrative Region of China. This paper is available on arxiv under CC by 4.0 Deed (Attribution 4.0 International) license. This paper is available on arxiv under CC by 4.0 Deed (Attribution 4.0 International) license. available on arxiv

ClassBD Outperforms Competitors in Real-World Bearing Fault Diagnosis Using PU Dataset

Table of Links Abstract and 1. Introduction 2. Preliminaries and 2.1. Blind deconvolution 2.2. Quadratic neural networks 3. Methodology 3.1. Time domain quadratic convolutional filter 3.2. Superiority of cyclic features extraction by QCNN 3.3. Frequency domain linear filter with envelope spectrum objective function 3.4. Integral optimization with uncertainty-aware weighing scheme 4. Computational experiments 4.1. Experimental configurations 4.2. Case study 1: PU dataset 4.3. Case study 2: JNU dataset 4.4. Case study 3: HIT dataset 5. Computational experiments 5.1. Comparison of BD methods 5.2. Classification results on various noise conditions 5.3. Employing ClassBD to deep learning classifiers 5.4. Employing ClassBD to machine learning classifiers 5.5. Feature extraction ability of quadratic and conventional networks 5.6. Comparison of ClassBD filters 6. Conclusions Appendix and References 3. Methodology The proposed framework, as illustrated in Figure 1, primarily consists of two BD filters, namely a time domain quadratic convolutional filter and a frequency domain linear filter. These filters serve as a plug-and-play denoising module, and they are designed to perform the same function as conventional BD methods to ensure that the output is in the same dimension as the input. The time domain filter is characterized by two symmetric quadratic convolutional neural network (QCNN) layers. A 16-channel QCNN is employed to filter the input signal (1 × 2048), and an inverse QCNN layer is used to fuse the 16 channels into one for recovering the input signal.


The frequency domain filter, on the other hand, starts with the fast Fourier transform (FFT) with an emphasis on highlighting the discrete frequency components. Subsequently, a linear neural layer filters the frequency domain of the signals, and the inverse FFT (IFFT) is conducted to recover the time domain signal. Moreover, an objective function in the envelope spectrum (ES) is designed for optimization. Authors:
(1) Jing-Xiao Liao, Department of Industrial and Systems Engineering, The Hong Kong Polytechnic University, Hong Kong, Special Administrative Region of China and School of Instrumentation Science and Engineering, Harbin Institute of Technology, Harbin, China;
(2) Chao He, School of Mechanical, Electronic and Control Engineering, Beijing Jiaotong University, Beijing, China;
(3) Jipu Li, Department of Industrial and Systems Engineering, The Hong Kong Polytechnic University, Hong Kong, Special Administrative Region of China;
(4) Jinwei Sun, School of Instrumentation Science and Engineering, Harbin Institute of Technology, Harbin, China;
(5) Shiping Zhang (Corresponding author), School of Instrumentation Science and Engineering, Harbin Institute of Technology, Harbin, China;
(6) Xiaoge Zhang (Corresponding author), Department of Industrial and Systems Engineering, The Hong Kong Polytechnic University, Hong Kong, Special Administrative Region of China. This paper is available on arxiv under CC by 4.0 Deed (Attribution 4.0 International) license. Table of Links Abstract and 1. Introduction Abstract and 1. Introduction 2. Preliminaries and 2.1. Blind deconvolution 2. Preliminaries and 2.1. Blind deconvolution 2.2. Quadratic neural networks 2.2. Quadratic neural networks 3. Methodology 3. Methodology 3.1. Time domain quadratic convolutional filter 3.1. Time domain quadratic convolutional filter 3.2. Superiority of cyclic features extraction by QCNN 3.2. Superiority of cyclic features extraction by QCNN 3.3. Frequency domain linear filter with envelope spectrum objective function 3.3. Frequency domain linear filter with envelope spectrum objective function 3.4. Integral optimization with uncertainty-aware weighing scheme 3.4. Integral optimization with uncertainty-aware weighing scheme 4. Computational experiments 4.1. Experimental configurations 4.1. Experimental configurations 4.2. Case study 1: PU dataset 4.2. Case study 1: PU dataset 4.3. Case study 2: JNU dataset 4.3. Case study 2: JNU dataset 4.4. Case study 3: HIT dataset 4.4. Case study 3: HIT dataset 5. Computational experiments 5.1. Comparison of BD methods 5.1. Comparison of BD methods 5.2. Classification results on various noise conditions 5.2. Classification results on various noise conditions 5.3. Employing ClassBD to deep learning classifiers 5.3. Employing ClassBD to deep learning classifiers 5.4. Employing ClassBD to machine learning classifiers 5.4. Employing ClassBD to machine learning classifiers 5.5. Feature extraction ability of quadratic and conventional networks 5.5. Feature extraction ability of quadratic and conventional networks 5.6. Comparison of ClassBD filters 5.6. Comparison of ClassBD filters 6. Conclusions 6. Conclusions Appendix and References Appendix and References 3. Methodology The proposed framework, as illustrated in Figure 1, primarily consists of two BD filters, namely a time domain quadratic convolutional filter and a frequency domain linear filter. These filters serve as a plug-and-play denoising module, and they are designed to perform the same function as conventional BD methods to ensure that the output is in the same dimension as the input. The time domain filter is characterized by two symmetric quadratic convolutional neural network (QCNN) layers. A 16-channel QCNN is employed to filter the input signal (1 × 2048), and an inverse QCNN layer is used to fuse the 16 channels into one for recovering the input signal. The frequency domain filter, on the other hand, starts with the fast Fourier transform (FFT) with an emphasis on highlighting the discrete frequency components. Subsequently, a linear neural layer filters the frequency domain of the signals, and the inverse FFT (IFFT) is conducted to recover the time domain signal. Moreover, an objective function in the envelope spectrum (ES) is designed for optimization. The time domain filter is characterized by two symmetric quadratic convolutional neural network (QCNN) layers. A 16-channel QCNN is employed to filter the input signal (1 × 2048), and an inverse QCNN layer is used to fuse the 16 channels into one for recovering the input signal. The time domain filter is characterized by two symmetric quadratic convolutional neural network (QCNN) layers. A 16-channel QCNN is employed to filter the input signal (1 × 2048), and an inverse QCNN layer is used to fuse the 16 channels into one for recovering the input signal. The frequency domain filter, on the other hand, starts with the fast Fourier transform (FFT) with an emphasis on highlighting the discrete frequency components. Subsequently, a linear neural layer filters the frequency domain of the signals, and the inverse FFT (IFFT) is conducted to recover the time domain signal. Moreover, an objective function in the envelope spectrum (ES) is designed for optimization. The frequency domain filter, on the other hand, starts with the fast Fourier transform (FFT) with an emphasis on highlighting the discrete frequency components. Subsequently, a linear neural layer filters the frequency domain of the signals, and the inverse FFT (IFFT) is conducted to recover the time domain signal. Moreover, an objective function in the envelope spectrum (ES) is designed for optimization. Authors: (1) Jing-Xiao Liao, Department of Industrial and Systems Engineering, The Hong Kong Polytechnic University, Hong Kong, Special Administrative Region of China and School of Instrumentation Science and Engineering, Harbin Institute of Technology, Harbin, China; (2) Chao He, School of Mechanical, Electronic and Control Engineering, Beijing Jiaotong University, Beijing, China; (3) Jipu Li, Department of Industrial and Systems Engineering, The Hong Kong Polytechnic University, Hong Kong, Special Administrative Region of China; (4) Jinwei Sun, School of Instrumentation Science and Engineering, Harbin Institute of Technology, Harbin, China; (5) Shiping Zhang (Corresponding author), School of Instrumentation Science and Engineering, Harbin Institute of Technology, Harbin, China; (6) Xiaoge Zhang (Corresponding author), Department of Industrial and Systems Engineering, The Hong Kong Polytechnic University, Hong Kong, Special Administrative Region of China. Authors: Authors: (1) Jing-Xiao Liao, Department of Industrial and Systems Engineering, The Hong Kong Polytechnic University, Hong Kong, Special Administrative Region of China and School of Instrumentation Science and Engineering, Harbin Institute of Technology, Harbin, China; (2) Chao He, School of Mechanical, Electronic and Control Engineering, Beijing Jiaotong University, Beijing, China; (3) Jipu Li, Department of Industrial and Systems Engineering, The Hong Kong Polytechnic University, Hong Kong, Special Administrative Region of China; (4) Jinwei Sun, School of Instrumentation Science and Engineering, Harbin Institute of Technology, Harbin, China; (5) Shiping Zhang (Corresponding author), School of Instrumentation Science and Engineering, Harbin Institute of Technology, Harbin, China; (6) Xiaoge Zhang (Corresponding author), Department of Industrial and Systems Engineering, The Hong Kong Polytechnic University, Hong Kong, Special Administrative Region of China. This paper is available on arxiv under CC by 4.0 Deed (Attribution 4.0 International) license. This paper is available on arxiv under CC by 4.0 Deed (Attribution 4.0 International) license. available on arxiv

Researchers Propose Novel Framework Combining Time and Frequency Domain Filters

SAN JOSE, Costa Rica, December 23rd, 2024/Chainwire/--Winna.com,a crypto-focused casino gaming platform launched in the summer of 2024, has successfully raised $15 million in a seed funding round, as first reported by Crunchbase. The investment will support Winna.com in enhancing its product offerings and accelerating its growth. The platform already boasts a thriving community of over 10,000 active players, signaling strong early traction in the competitive online gambling space. Transforming Online Gambling Through Crypto Paul Martens, CEO and Co-Founder of Winna.com, highlighted the growing appeal of crypto casinos. "More and more players are recognizing the significant advantages that crypto and Bitcoin casinos offer over traditional online platforms. Privacy and the ability to gamble anonymously are key drivers for this shift," he said. Martens also praised the platform's commitment to transparency through its provably fair gaming system. "Provably fair games give players a unique opportunity to gamble with confidence, knowing the outcomes are not only fair but also verifiable by anyone," he added. Winna currently features a range of provably fair crypto games, including Mines, Plinko, Roulette, Blackjack, Keno, and Dice. Martens revealed plans to leverage the recent funding to expand their portfolio, particularly in player versus player (PvP) games. "We aim to develop more innovative crypto games that continue to deliver an industry-low house edge of just 1%," he said. Diverse Offerings for Every Player In addition to its original crypto games, Winna.com offers over 4,000 slots, live casino games, and game shows from leading providers such as Pragmatic Play, Hacksaw Gaming, Evolution, and more. The platform also operates a crypto-first sportsbook, featuring live coverage of over 10,000 events across 100+ tournaments, including major leagues like the NFL, NBA, UFC, MLB, and the Premier League. Building Trust and Speed in Online Gambling When asked why players choose Winna.com, Martens emphasized the platform's commitment to trust and user satisfaction. "Trust is at the core of our brand. Players know they’ll receive their winnings within seconds, and every game round is fair and verifiable. We prioritize our users’ privacy and avoid unnecessary procedures that are common with some of our competitors," he stated. VIP Excellence and Industry-Leading Rewards Winna.com’s team includes four VIP hosts with extensive experience in physical gaming establishments like MGM in Las Vegas. Elias Platzer, Head of Player Development, explained the importance of their VIP program. "Our VIP program is a cornerstone of our identity. We’re committed to offering the most rewarding experience, with unmatched bonuses and rewards through our Rakeback program," Platzer said. A Vision for the Future Martens concluded by sharing his vision for Winna.com’s future. "We are set to transform the online gaming experience by developing unique games, delivering a one-of-a-kind VIP program, and setting new standards for trust and transparency in the industry," he said. With its innovative approach, commitment to fairness, and focus on user satisfaction, Winna.com aims to contribute meaningfully to the crypto gambling industry. Winna.com is the source of this content. This Press Release is for informational purposes only. The information does not constitute investment advice or an offer to invest. About Winna.com Launched in the summer of 2024, Winna.com is a cutting-edge crypto casino and sportsbook with offices in Costa Rica and Switzerland. Backed by a team of seasoned professionals from the traditional iGaming and crypto industries, Winna.com offers provably fair crypto games, thousands of slots and live casino options, and a crypto-first sportsbook. Focused on trust, fast payouts, and exceptional VIP experiences, Winna.com is redefining the online gaming experience for crypto enthusiasts worldwide. Contact Paul Mertens Winna paul@winna.com This story was distributed as a release by Chainwire under HackerNoon’s Business Blogging Program. Learn more about the program here SAN JOSE, Costa Rica, December 23rd, 2024/Chainwire/--Winna.com,a crypto-focused casino gaming platform launched in the summer of 2024, has successfully raised $15 million in a seed funding round, as first reported by Crunchbase . Crunchbase Crunchbase The investment will support Winna.com in enhancing its product offerings and accelerating its growth. The platform already boasts a thriving community of over 10,000 active players, signaling strong early traction in the competitive online gambling space. Transforming Online Gambling Through Crypto Paul Martens, CEO and Co-Founder of Winna.com, highlighted the growing appeal of crypto casinos. "More and more players are recognizing the significant advantages that crypto and Bitcoin casinos offer over traditional online platforms. Privacy and the ability to gamble anonymously are key drivers for this shift," he said. "More and more players are recognizing the significant advantages that crypto and Bitcoin casinos offer over traditional online platforms. Privacy and the ability to gamble anonymously are key drivers for this shift," he said. Martens also praised the platform's commitment to transparency through its provably fair gaming system. "Provably fair games give players a unique opportunity to gamble with confidence, knowing the outcomes are not only fair but also verifiable by anyone," he added. "Provably fair games give players a unique opportunity to gamble with confidence, knowing the outcomes are not only fair but also verifiable by anyone," he added. Winna currently features a range of provably fair crypto games, including Mines, Plinko, Roulette, Blackjack, Keno, and Dice. Martens revealed plans to leverage the recent funding to expand their portfolio, particularly in player versus player (PvP) games. "We aim to develop more innovative crypto games that continue to deliver an industry-low house edge of just 1%," he said. "We aim to develop more innovative crypto games that continue to deliver an industry-low house edge of just 1%," he said. Diverse Offerings for Every Player In addition to its original crypto games, Winna.com offers over 4,000 slots, live casino games, and game shows from leading providers such as Pragmatic Play, Hacksaw Gaming, Evolution, and more. The platform also operates a crypto-first sportsbook , featuring live coverage of over 10,000 events across 100+ tournaments, including major leagues like the NFL, NBA, UFC, MLB, and the Premier League. sportsbook sportsbook Building Trust and Speed in Online Gambling When asked why players choose Winna.com, Martens emphasized the platform's commitment to trust and user satisfaction. "Trust is at the core of our brand. Players know they’ll receive their winnings within seconds, and every game round is fair and verifiable. We prioritize our users’ privacy and avoid unnecessary procedures that are common with some of our competitors," he stated. "Trust is at the core of our brand. Players know they’ll receive their winnings within seconds, and every game round is fair and verifiable. We prioritize our users’ privacy and avoid unnecessary procedures that are common with some of our competitors," he stated. VIP Excellence and Industry-Leading Rewards Winna.com’s team includes four VIP hosts with extensive experience in physical gaming establishments like MGM in Las Vegas. Elias Platzer, Head of Player Development, explained the importance of their VIP program. "Our VIP program is a cornerstone of our identity. We’re committed to offering the most rewarding experience, with unmatched bonuses and rewards through our Rakeback program," Platzer said. "Our VIP program is a cornerstone of our identity. We’re committed to offering the most rewarding experience, with unmatched bonuses and rewards through our Rakeback program," Platzer said. A Vision for the Future Martens concluded by sharing his vision for Winna.com’s future. "We are set to transform the online gaming experience by developing unique games, delivering a one-of-a-kind VIP program, and setting new standards for trust and transparency in the industry," he said. "We are set to transform the online gaming experience by developing unique games, delivering a one-of-a-kind VIP program, and setting new standards for trust and transparency in the industry," he said. With its innovative approach, commitment to fairness, and focus on user satisfaction, Winna.com aims to contribute meaningfully to the crypto gambling industry. Winna.com is the source of this content. This Press Release is for informational purposes only. The information does not constitute investment advice or an offer to invest. Winna.com Winna.com About Winna.com Launched in the summer of 2024, Winna.com is a cutting-edge crypto casino and sportsbook with offices in Costa Rica and Switzerland. Winna.com Winna.com Backed by a team of seasoned professionals from the traditional iGaming and crypto industries, Winna.com offers provably fair crypto games, thousands of slots and live casino options, and a crypto-first sportsbook. Focused on trust, fast payouts, and exceptional VIP experiences, Winna.com is redefining the online gaming experience for crypto enthusiasts worldwide. Contact Paul Mertens Winna paul@winna.com This story was distributed as a release by Chainwire under HackerNoon’s Business Blogging Program. Learn more about the program here This story was distributed as a release by Chainwire under HackerNoon’s Business Blogging Program. Learn more about the program here here here

New Crypto Casino Platform Winna.com Secures $15 Million In Seed Funding

Neue Krypto-Casino-Plattform Winna.com sichert sich 15 Millionen US-Dollar an Startkapital

La nueva plataforma de casino criptográfico Winna.com obtiene $15 millones en financiación inicial

La nouvelle plateforme de crypto-casino Winna.com obtient un financement initial de 15 millions de dollars

Nova kripto kasino platforma Winna.com osigurava 15 milijuna dolara početnog financiranja

新しい暗号カジノプラットフォーム Winna.com がシード資金として 1,500 万ドルを確保

Winna.com жаңа Crypto Casino платформасы 15 миллион доллар тұқымдық қаржыландыруды қамтамасыз етеді

ເວທີຄາສິໂນ Crypto ໃໝ່ Winna.com ຮັບປະກັນເງິນທຶນແກ່ນພັນ $15 ລ້ານ

Naujoji kriptovaliutų kazino platforma Winna.com užsitikrina 15 mln. USD pradinį finansavimą

नयाँ क्रिप्टो क्यासिनो प्लेटफर्म Winna.com बीउ कोषमा $ 15 मिलियन सुरक्षित गर्दछ

نوی کریپټو کیسینو پلیټ فارم Winna.com د تخمونو تمویل کې 15 ملیون ډالر خوندي کوي

Platform Casino cusub ee loo yaqaan 'Crypto Casino Platform' Winna.com wuxuu damaanad qaadayaa $ 15 Milyan ee Maalgelinta Abuurka

新的加密赌场平台 Winna.com 获得 1500 万美元种子资金

Table of Links Abstract and Introduction


Domain and Task
2.1. Data sources and complexity
2.2. Task definition


Related Work
3.1. Text mining and NLP research overview
3.2. Text mining and NLP in industry use
3.3. Text mining and NLP for procurement
3.4. Conclusion from literature review


Proposed Methodology
4.1. Domain knowledge
4.2. Content extraction
4.3. Lot zoning
4.4. Lot item detection
4.5. Lot parsing
4.6. XML parsing, data joining, and risk indices development


Experiment and Demonstration
5.1. Component evaluation
5.2. System demonstration


Discussion
6.1. The ‘industry’ focus of the project
6.2. Data heterogeneity, multilingual and multi-task nature
6.3. The dilemma of algorithmic choices
6.4. The cost of training data


Conclusion, Acknowledgements, and References 3.2. Text mining and NLP in industry use Text mining and NLP already have wide use in a number of industry contexts. Here, due to limited space, we only look at a few domains briefly and focus only on work for real applications instead of general purpose tasks such as building domain corpora or language models, or fundamental NLP research adapted to domain specific data. In the legal domain, Zhong et al. (2020) summarised three main application areas: judicial decision prediction, similar case matching and question answering. Judicial decision prediction studies the problem of determining the verdict of a court on the basis of textual information about a court case before the verdict was made. This is typically treated as a text classification task, where court decisions are categorised and they are predicted based on features extracted from the case texts. For details, we refer readers to a survey by Francia et al. (2022). Similar case matching is an important application where judicial decisions are made according to similar and representative cases in the past. The goal is to find pairs of similar cases. This is often cast as a retrieval task, and is being extensively studied in Legal IR (Xiao et al., 2019). Similar to judicial decision prediction, existing methods mainly focus on comparing case texts at word or semantic level. In both areas, recent studies have argued for extracting more fine-grained ‘case element’ information (e.g., through using NER) that may better represent legal cases. For example, Hu et al. (2018) extracted ‘legal attributes’ that define the nature of cases in judicial decision prediction. Question answering is a rather complicated NLP task that builds on many low-level tasks, such as finding relevant text passages (PR and text classification), locating case elements (NER and relation extraction), and text similarity matching. QA is not related to our work. For both reasons, we do not go into detail about legal QA. But generally, questions typically concern explanation of legal concepts or case analysis. Legal documents are often lengthy, but well-structured into different parts and follow standard structures that make them easier to process. Over the years, the research community in this domain has defined granular tasks, standards, and created rich resources including training data. In contrast, our work deals with much more heterogeneous data and inconsistent structures, where creating training data for some tasks is expensive. Most established methods in this domain are therefore not directly transferable to our task. With the growth of e-commerce and social media, mining Web content related to service/product provisions has been extensively studied and applied to develop competitive advantage for businesses. A major area of application is analysing reviews from e-commerce or social media platforms to acquire business intelligence that inform various practices. For example, through analysing customer reviews about their own services or products, one can gain insights on customer satisfaction, to inform public relation management (Nave et al., 2018) and product development (Yang et al., 2019). Analysing such data about a sector in general including multiple competitors allows discovery of market trends and even forecast of demands (Chatterjee, 2019; Sharma et al,. 2020). Further, with the increase of platforms of user-generated content, the diffusion of fake information is made easier and becoming an increasing concern. Therefore, work has been done to automatically detect and filter such misinformation (Fang et al., 2020). A thorough review of work in the above areas can be found in Kuma et al. (2021). Work in these areas heavily rely on sentiment analysis, which is a type of text classification task aiming to analyse the sentiment of a text. It is also very useful to extract key elements related to service/product provision (e.g., product features, service processes) for more fine-grained analysis, and this can benefit from NER methods. Many studies also use topic modelling and social network analysis, which are beyond the scope of this work. Compared to the domain in our work, data studied in these areas are rather homogeneous - they are typically free-form review texts that are independent and self-contained, making it easy to adapt state-of-the-art methods. Partially for this reason, there is an abundance of tools developed in this area. However, as explained before, the data we have to deal with is much more complex. Text mining and NLP are also widely used in system requirements analysis and quality control in construction. To name a few, Li et al. (2015) processed software user manuals (PDFs) to extract functional and non-functional requirements. A dictionary of keywords combined with rules are used to extract sentences and topic modelling is applied to group similar sentences into groups, which may represent the same/similar requirements. The work also needed to deal with filtering irrelevant text passages in the extracted texts from PDFs, but this was done manually. Our work is similar in that some of our methods make use of keywords and rules, and also process PDFs. However, we need to deal with a much larger and more heterogeneous collection and the filtering of content must be done automatically. Tiun et al. (2020) developed supervised classifiers of functional and nonfunctional system requirements using well-curated training sentences obtained from Siemens Logistics and Automotive Organization. Khan et al. (2020) analysed Reddit posts about Google Maps, and trained a text classifier to discover posts related to the software feature requirements or issues using a set of manually annotated posts. Both deal with higher quality of data and have access to a large amount of high quality training data, while our project only managed to create limited training data for some but not all tasks due to resource constraints. Lee et al. (2014) applied keywords extraction and co-occurrence analysis in marine structure quality inspection reports. The co-occurrence map is used to aid humans in identifying which quality aspects are most frequently mentioned and in what ways. Similar methods are used in Tian et al. (2021), who processed construction project reports to extract keywords and concepts related to several key aspects of project management (e.g., quality control, safety management). Zhang et al. (2019) developed a text classifier to classify accident reports from construction projects, and implemented a rule-based extractor to identify causes from accidents. The data used in these studies are also rather simple compared to ours. For example, report cases used by Zhang et al. (2019) are composed of short paragraphs and use a consistent structure. Authors:
(1) Ziqi Zhang*, Information School, the University of Sheffield, Regent Court, Sheffield, UKS1 4DP (Ziqi.Zhang@sheffield.ac.uk);
(2) Tomas Jasaitis, Vamstar Ltd., London (Tomas.Jasaitis@vamstar.io);
(3) Richard Freeman, Vamstar Ltd., London (Richard.Freeman@vamstar.io);
(4) Rowida Alfrjani, Information School, the University of Sheffield, Regent Court, Sheffield, UKS1 4DP (Rowida.Alfrjani@sheffield.ac.uk);
(5) Adam Funk, Information School, the University of Sheffield, Regent Court, Sheffield, UKS1 4DP (Adam.Funk@sheffield.ac.uk). This paper is available on arxiv under CC BY 4.0 license. Table of Links Abstract and Introduction Domain and Task
2.1. Data sources and complexity
2.2. Task definition Related Work
3.1. Text mining and NLP research overview
3.2. Text mining and NLP in industry use
3.3. Text mining and NLP for procurement
3.4. Conclusion from literature review Proposed Methodology
4.1. Domain knowledge
4.2. Content extraction
4.3. Lot zoning
4.4. Lot item detection
4.5. Lot parsing
4.6. XML parsing, data joining, and risk indices development Experiment and Demonstration
5.1. Component evaluation
5.2. System demonstration Discussion
6.1. The ‘industry’ focus of the project
6.2. Data heterogeneity, multilingual and multi-task nature
6.3. The dilemma of algorithmic choices
6.4. The cost of training data Conclusion, Acknowledgements, and References Abstract and Introduction Abstract and Introduction Abstract and Introduction Abstract and Introduction Domain and Task 2.1. Data sources and complexity 2.2. Task definition Domain and Task Domain and Task 2.1. Data sources and complexity 2.1. Data sources and complexity 2.2. Task definition 2.2. Task definition Related Work 3.1. Text mining and NLP research overview 3.2. Text mining and NLP in industry use 3.3. Text mining and NLP for procurement 3.4. Conclusion from literature review Related Work Related Work 3.1. Text mining and NLP research overview 3.1. Text mining and NLP research overview 3.2. Text mining and NLP in industry use 3.2. Text mining and NLP in industry use 3.3. Text mining and NLP for procurement 3.3. Text mining and NLP for procurement 3.4. Conclusion from literature review 3.4. Conclusion from literature review Proposed Methodology 4.1. Domain knowledge 4.2. Content extraction 4.3. Lot zoning 4.4. Lot item detection 4.5. Lot parsing 4.6. XML parsing, data joining, and risk indices development Proposed Methodology Proposed Methodology Proposed Methodology 4.1. Domain knowledge 4.1. Domain knowledge 4.2. Content extraction 4.2. Content extraction 4.3. Lot zoning 4.3. Lot zoning 4.4. Lot item detection 4.4. Lot item detection 4.5. Lot parsing 4.5. Lot parsing 4.6. XML parsing, data joining, and risk indices development 4.6. XML parsing, data joining, and risk indices development Experiment and Demonstration 5.1. Component evaluation 5.2. System demonstration Experiment and Demonstration Experiment and Demonstration 5.1. Component evaluation 5.1. Component evaluation 5.2. System demonstration 5.2. System demonstration Discussion 6.1. The ‘industry’ focus of the project 6.2. Data heterogeneity, multilingual and multi-task nature 6.3. The dilemma of algorithmic choices 6.4. The cost of training data Discussion Discussion 6.1. The ‘industry’ focus of the project 6.1. The ‘industry’ focus of the project 6.2. Data heterogeneity, multilingual and multi-task nature 6.2. Data heterogeneity, multilingual and multi-task nature 6.3. The dilemma of algorithmic choices 6.3. The dilemma of algorithmic choices 6.4. The cost of training data 6.4. The cost of training data Conclusion, Acknowledgements, and References Conclusion, Acknowledgements, and References Conclusion, Acknowledgements, and References Conclusion, Acknowledgements, and References 3.2. Text mining and NLP in industry use Text mining and NLP already have wide use in a number of industry contexts. Here, due to limited space, we only look at a few domains briefly and focus only on work for real applications instead of general purpose tasks such as building domain corpora or language models, or fundamental NLP research adapted to domain specific data. In the legal domain , Zhong et al. (2020) summarised three main application areas: judicial decision prediction, similar case matching and question answering. Judicial decision prediction studies the problem of determining the verdict of a court on the basis of textual information about a court case before the verdict was made. This is typically treated as a text classification task, where court decisions are categorised and they are predicted based on features extracted from the case texts. For details, we refer readers to a survey by Francia et al. (2022). Similar case matching is an important application where judicial decisions are made according to similar and representative cases in the past. The goal is to find pairs of similar cases. This is often cast as a retrieval task, and is being extensively studied in Legal IR (Xiao et al., 2019). Similar to judicial decision prediction, existing methods mainly focus on comparing case texts at word or semantic level. In both areas, recent studies have argued for extracting more fine-grained ‘case element’ information (e.g., through using NER) that may better represent legal cases. For example, Hu et al. (2018) extracted ‘legal attributes’ that define the nature of cases in judicial decision prediction. Question answering is a rather complicated NLP task that builds on many low-level tasks, such as finding relevant text passages (PR and text classification), locating case elements (NER and relation extraction), and text similarity matching. QA is not related to our work. For both reasons, we do not go into detail about legal QA. But generally, questions typically concern explanation of legal concepts or case analysis. legal domain Legal documents are often lengthy, but well-structured into different parts and follow standard structures that make them easier to process. Over the years, the research community in this domain has defined granular tasks, standards, and created rich resources including training data. In contrast, our work deals with much more heterogeneous data and inconsistent structures, where creating training data for some tasks is expensive. Most established methods in this domain are therefore not directly transferable to our task. With the growth of e-commerce and social media, mining Web content related to service/product provisions has been extensively studied and applied to develop competitive advantage for businesses. A major area of application is analysing reviews from e-commerce or social media platforms to acquire business intelligence that inform various practices. For example, through analysing customer reviews about their own services or products, one can gain insights on customer satisfaction, to inform public relation management (Nave et al., 2018) and product development (Yang et al., 2019). Analysing such data about a sector in general including multiple competitors allows discovery of market trends and even forecast of demands (Chatterjee, 2019; Sharma et al,. 2020). Further, with the increase of platforms of user-generated content, the diffusion of fake information is made easier and becoming an increasing concern. Therefore, work has been done to automatically detect and filter such misinformation (Fang et al., 2020). A thorough review of work in the above areas can be found in Kuma et al. (2021). service/product provisions Work in these areas heavily rely on sentiment analysis, which is a type of text classification task aiming to analyse the sentiment of a text. It is also very useful to extract key elements related to service/product provision (e.g., product features, service processes) for more fine-grained analysis, and this can benefit from NER methods. Many studies also use topic modelling and social network analysis, which are beyond the scope of this work. Compared to the domain in our work, data studied in these areas are rather homogeneous - they are typically free-form review texts that are independent and self-contained, making it easy to adapt state-of-the-art methods. Partially for this reason, there is an abundance of tools developed in this area. However, as explained before, the data we have to deal with is much more complex. Text mining and NLP are also widely used in system requirements analysis and quality control in construction . To name a few, Li et al. (2015) processed software user manuals (PDFs) to extract functional and non-functional requirements. A dictionary of keywords combined with rules are used to extract sentences and topic modelling is applied to group similar sentences into groups, which may represent the same/similar requirements. The work also needed to deal with filtering irrelevant text passages in the extracted texts from PDFs, but this was done manually. Our work is similar in that some of our methods make use of keywords and rules, and also process PDFs. However, we need to deal with a much larger and more heterogeneous collection and the filtering of content must be done automatically. Tiun et al. (2020) developed supervised classifiers of functional and nonfunctional system requirements using well-curated training sentences obtained from Siemens Logistics and Automotive Organization. Khan et al. (2020) analysed Reddit posts about Google Maps, and trained a text classifier to discover posts related to the software feature requirements or issues using a set of manually annotated posts. Both deal with higher quality of data and have access to a large amount of high quality training data, while our project only managed to create limited training data for some but not all tasks due to resource constraints. system requirements analysis quality control in construction Lee et al. (2014) applied keywords extraction and co-occurrence analysis in marine structure quality inspection reports. The co-occurrence map is used to aid humans in identifying which quality aspects are most frequently mentioned and in what ways. Similar methods are used in Tian et al. (2021), who processed construction project reports to extract keywords and concepts related to several key aspects of project management (e.g., quality control, safety management). Zhang et al. (2019) developed a text classifier to classify accident reports from construction projects, and implemented a rule-based extractor to identify causes from accidents. The data used in these studies are also rather simple compared to ours. For example, report cases used by Zhang et al. (2019) are composed of short paragraphs and use a consistent structure. Authors: (1) Ziqi Zhang*, Information School, the University of Sheffield, Regent Court, Sheffield, UKS1 4DP (Ziqi.Zhang@sheffield.ac.uk); (2) Tomas Jasaitis, Vamstar Ltd., London (Tomas.Jasaitis@vamstar.io); (3) Richard Freeman, Vamstar Ltd., London (Richard.Freeman@vamstar.io); (4) Rowida Alfrjani, Information School, the University of Sheffield, Regent Court, Sheffield, UKS1 4DP (Rowida.Alfrjani@sheffield.ac.uk); (5) Adam Funk, Information School, the University of Sheffield, Regent Court, Sheffield, UKS1 4DP (Adam.Funk@sheffield.ac.uk). Authors: Authors: (1) Ziqi Zhang*, Information School, the University of Sheffield, Regent Court, Sheffield, UKS1 4DP (Ziqi.Zhang@sheffield.ac.uk); (2) Tomas Jasaitis, Vamstar Ltd., London (Tomas.Jasaitis@vamstar.io); (3) Richard Freeman, Vamstar Ltd., London (Richard.Freeman@vamstar.io); (4) Rowida Alfrjani, Information School, the University of Sheffield, Regent Court, Sheffield, UKS1 4DP (Rowida.Alfrjani@sheffield.ac.uk); (5) Adam Funk, Information School, the University of Sheffield, Regent Court, Sheffield, UKS1 4DP (Adam.Funk@sheffield.ac.uk). This paper is available on arxiv under CC BY 4.0 license. This paper is available on arxiv under CC BY 4.0 license. available on arxiv available on arxiv

New Study Shows How Text Mining and NLP Transform Legal, E-commerce, and Construction Industries

Hey Hackers! Welcome to another addition to our Company of the Week feature! Every week, we share an awesome tech brand from our tech company database, making their evergreen mark on the internet. This unique HackerNoon database ranks S&P 500 companies and top startups of the year alike. This week, we’re proud to shine a spotlight on Moonlock, a cybersecurity division that protects you every time you use your everyday devices. Specifically made for Mac, Moonlock is on a quest to make the Apple ecosystem as safe as possible by stopping cyber threats from ever reaching you. Want to be featured on HackerNoon’s Company of the Week? Request Your Tech Company Page on HackerNoon! Meet Moonlock: #FunFact Moonlock was made by MacPaw, a company with a long history of creating amazing apps, first starting with CleanMyMac in 2008. Since then, MacPaw hasn’t taken its foot off the gas. On their website, they rightfully boast: Over the years, we’ve made other successful apps. You may have heard of them — after all, there’s a MacPaw app on every 5th Mac in the world. And its success and its long-standing history eventually led MacPaw to Moonlock. With that in mind, let’s learn more about Moonlock itself. Although it’s made up of a small team, they have a combined 47 years of experience in cybersecurity together. Plus, the team includes a white-hat hacker and a former cybercrime investigator. So, it’s no wonder Moonlock is as efficient as it is and that they were able to create the Moonlock Engine, the tech behind Malware Remove. Moonlock Engine quickly scans your devices, detects malware, and then gets rid of it. It’s just that simple. Here are 3 more important things you need to know about Moonlock Engine, taken straight from the horse’s mouth (or in this case, its website): Moonlock Engine inspects mail attachments, DMG and ZIP archives, USB drives, browser extensions, and launchers to ensure that no malware slips through the cracks.


Moonlock Engine has a lightning-fast scan: now 2x faster on M1-powered Macs and 1.5x faster on Intel-based computers.


Moonlock Engine is independently verified by AV-TEST and was able to detect 93.3% of malicious samples during a private test. Share Your Company's Story via HackerNoon Moonlock 🤝HackerNoon Moonlock has now turned its sights on HackerNoon and has partnered with us for our business blogging program. So far, Moonlock’s stories have garnered over 9.5k reads and counting! Check them out to learn more about how you can stay safe on the internet: Crypto Tool or Data Thief? How Meme-Token-Hunter-Bot and Its Clones Steal from macOS Users
Here’s How We Made a Real-time Phishing Website Detector for MacOS
Tracking Atomic Stealer on macOS: Sophisticated Malware Replacing LedgerLive App Like 4000+ other brands, you can also benefit from publishing on HackerNoon:Learn more about HackerNoon’s Business Blogging Program
Learn about our Technology Press Release Program. That’s it for this week! Until next time, The HackerNoon team Hey Hackers! Welcome to another addition to our Company of the Week feature! Every week, we share an awesome tech brand from our tech company database , making their evergreen mark on the internet. This unique HackerNoon database ranks S&P 500 companies and top startups of the year alike. Company of the Week tech company database This week, we’re proud to shine a spotlight on Moonlock , a cybersecurity division that protects you every time you use your everyday devices. Specifically made for Mac, Moonlock is on a quest to make the Apple ecosystem as safe as possible by stopping cyber threats from ever reaching you. Moonlock Want to be featured on HackerNoon’s Company of the Week? Request Your Tech Company Page on HackerNoon! Want to be featured on HackerNoon’s Company of the Week? Request Your Tech Company Page on HackerNoon ! Want to be featured on HackerNoon’s Company of the Week? Request Your Tech Company Page on HackerNoon Want to be featured on HackerNoon’s Company of the Week? Request Your Tech Company Page on HackerNoon Meet Moonlock: #FunFact Moonlock was made by MacPaw, a company with a long history of creating amazing apps, first starting with CleanMyMac in 2008. Since then, MacPaw hasn’t taken its foot off the gas. On their website , they rightfully boast: Moonlock website Over the years, we’ve made other successful apps. You may have heard of them — after all, there’s a MacPaw app on every 5th Mac in the world. Over the years, we’ve made other successful apps. You may have heard of them — after all, there’s a MacPaw app on every 5th Mac in the world. And its success and its long-standing history eventually led MacPaw to Moonlock. With that in mind, let’s learn more about Moonlock itself. Although it’s made up of a small team , they have a combined 47 years of experience in cybersecurity together. Plus, the team includes a white-hat hacker and a former cybercrime investigator. small team So, it’s no wonder Moonlock is as efficient as it is and that they were able to create the Moonlock Engine, the tech behind Malware Remove. Moonlock Engine quickly scans your devices, detects malware, and then gets rid of it. It’s just that simple. Here are 3 more important things you need to know about Moonlock Engine, taken straight from the horse’s mouth (or in this case, its website): Here are 3 more important things you need to know about Moonlock Engine, taken straight from the horse’s mouth (or in this case, its website ): website Moonlock Engine inspects mail attachments, DMG and ZIP archives, USB drives, browser extensions, and launchers to ensure that no malware slips through the cracks. Moonlock Engine has a lightning-fast scan: now 2x faster on M1-powered Macs and 1.5x faster on Intel-based computers. Moonlock Engine is independently verified by AV-TEST and was able to detect 93.3% of malicious samples during a private test. Moonlock Engine inspects mail attachments, DMG and ZIP archives, USB drives, browser extensions, and launchers to ensure that no malware slips through the cracks. Moonlock Engine inspects mail attachments, DMG and ZIP archives, USB drives, browser extensions, and launchers to ensure that no malware slips through the cracks. Moonlock Engine has a lightning-fast scan: now 2x faster on M1-powered Macs and 1.5x faster on Intel-based computers. Moonlock Engine has a lightning-fast scan: now 2x faster on M1-powered Macs and 1.5x faster on Intel-based computers. Moonlock Engine is independently verified by AV-TEST and was able to detect 93.3% of malicious samples during a private test. Moonlock Engine is independently verified by AV-TEST and was able to detect 93.3% of malicious samples during a private test. Share Your Company's Story via HackerNoon Share Your Company's Story via HackerNoon Share Your Company's Story via HackerNoon Share Your Company's Story via HackerNoon Moonlock 🤝HackerNoon Moonlock has now turned its sights on HackerNoon and has partnered with us for our business blogging program . So far, Moonlock’s stories have garnered over 9.5k reads and counting! Moonlock business blogging program Check them out to learn more about how you can stay safe on the internet: Crypto Tool or Data Thief? How Meme-Token-Hunter-Bot and Its Clones Steal from macOS Users Here’s How We Made a Real-time Phishing Website Detector for MacOS Tracking Atomic Stealer on macOS: Sophisticated Malware Replacing LedgerLive App Crypto Tool or Data Thief? How Meme-Token-Hunter-Bot and Its Clones Steal from macOS Users Crypto Tool or Data Thief? How Meme-Token-Hunter-Bot and Its Clones Steal from macOS Users Crypto Tool or Data Thief? How Meme-Token-Hunter-Bot and Its Clones Steal from macOS Users Here’s How We Made a Real-time Phishing Website Detector for MacOS Here’s How We Made a Real-time Phishing Website Detector for MacOS Here’s How We Made a Real-time Phishing Website Detector for MacOS Tracking Atomic Stealer on macOS: Sophisticated Malware Replacing LedgerLive App Tracking Atomic Stealer on macOS: Sophisticated Malware Replacing LedgerLive App Tracking Atomic Stealer on macOS: Sophisticated Malware Replacing LedgerLive App Like 4000+ other brands, you can also benefit from publishing on HackerNoon:Learn more about HackerNoon’s Business Blogging Program Learn about our Technology Press Release Program. Like 4000+ other brands, you can also benefit from publishing on HackerNoon:Learn more about HackerNoon’s Business Blogging Program Like 4000+ other brands, you can also benefit from publishing on HackerNoon: Learn more about HackerNoon’s Business Blogging Program Learn more about HackerNoon’s Business Blogging Program Learn more about HackerNoon’s Business Blogging Program Learn about our Technology Press Release Program . Learn about our Technology Press Release Program Learn about our Technology Press Release Program Learn about our Technology Press Release Program . . That’s it for this week! Until next time, The HackerNoon team

Stories, News & Mentions about COM

Com’s Stories on HackerNoon

Com's News & Mentions