Pub Date : 2018-07-03DOI: 10.1080/13614568.2018.1527114
E. Herder, M. Bieliková, F. Cena, M. Desmarais
ACM UMAP is an annual conference on user modeling, adaptation and personalization. User modeling concerns the process of understanding the user’s needs, preferences, interests, knowledge and other aspects. This is achieved by reasoning about and extracting knowledge from user data, which includes both data that is explicitly provided by the user—such as profile data—and implicitly gathered usage data—such as browsing data. Adaptation and personalization techniques exploit the user models in order to better tailor a software system, such as a website, to the user needs. Recommender systems are the best known type of personalized systems, but the field is much wider and includes among others personalized search, adaptive user interfaces, personalized advice, and personalized technology-enhanced learning. This special issue contains extended versions of selected papers from UMAP 2017, the 25th edition of the conference series. The conference was hosted in Bratislava, Slovakia, from 9 to 12 July 2017. The conference consisted of five tracks that represent the variety of disciplines and application areas in which user modeling, adaptation and personalization play a role. User interface aspects, including adaptive presentation and navigation, were covered by the tracks Intelligent User Interfaces and Adaptive Hypermedia. As one of the most visible and largest application area of personalization is the Social Web, we received in the corresponding track submissions that both analyzed user behavior to function as input for personalization, as well as the effect of personalization on user behavior. Being the most prominent and most applied adaptive technique, Recommender Systems were given a dedicated track as well. Finally, we dedicated a track to the field of Technology-Enhanced Adaptive Learning, as this is an application area with important and tangible impacts on society. The papers in this special issue belong to the latter two areas. Three papers are situated in the field of Technology-Enhanced Learning. The first paper, “Analysis and Design of Mastery Learning Criteria” (Pelánek and Řihák), shows that, under the assumption of isolated skills, the decision over skill mastery, and whether a system should let the student move on to the next skill to learn, can rest on a simple exponential moving average rather than on the more sophisticated Bayesian and logistic approaches to learner modeling. They also show that the choice of an appropriate mastery threshold and of the source of information is more influential than the choice of the learner modeling technique. The second paper focuses on open learner models, which is an approach for making a student’s learner model explicit to the student, in order to enhance reflection, self-awareness and self-regulation of the learning process. In “Navigation Support in Complex Learner Models: Assessing Visual Design Alternatives” (Guerra, Schunn, Bull, BarríaPineda and Brusilovsky), six alternative prototypes were
{"title":"Introduction","authors":"E. Herder, M. Bieliková, F. Cena, M. Desmarais","doi":"10.1080/13614568.2018.1527114","DOIUrl":"https://doi.org/10.1080/13614568.2018.1527114","url":null,"abstract":"ACM UMAP is an annual conference on user modeling, adaptation and personalization. User modeling concerns the process of understanding the user’s needs, preferences, interests, knowledge and other aspects. This is achieved by reasoning about and extracting knowledge from user data, which includes both data that is explicitly provided by the user—such as profile data—and implicitly gathered usage data—such as browsing data. Adaptation and personalization techniques exploit the user models in order to better tailor a software system, such as a website, to the user needs. Recommender systems are the best known type of personalized systems, but the field is much wider and includes among others personalized search, adaptive user interfaces, personalized advice, and personalized technology-enhanced learning. This special issue contains extended versions of selected papers from UMAP 2017, the 25th edition of the conference series. The conference was hosted in Bratislava, Slovakia, from 9 to 12 July 2017. The conference consisted of five tracks that represent the variety of disciplines and application areas in which user modeling, adaptation and personalization play a role. User interface aspects, including adaptive presentation and navigation, were covered by the tracks Intelligent User Interfaces and Adaptive Hypermedia. As one of the most visible and largest application area of personalization is the Social Web, we received in the corresponding track submissions that both analyzed user behavior to function as input for personalization, as well as the effect of personalization on user behavior. Being the most prominent and most applied adaptive technique, Recommender Systems were given a dedicated track as well. Finally, we dedicated a track to the field of Technology-Enhanced Adaptive Learning, as this is an application area with important and tangible impacts on society. The papers in this special issue belong to the latter two areas. Three papers are situated in the field of Technology-Enhanced Learning. The first paper, “Analysis and Design of Mastery Learning Criteria” (Pelánek and Řihák), shows that, under the assumption of isolated skills, the decision over skill mastery, and whether a system should let the student move on to the next skill to learn, can rest on a simple exponential moving average rather than on the more sophisticated Bayesian and logistic approaches to learner modeling. They also show that the choice of an appropriate mastery threshold and of the source of information is more influential than the choice of the learner modeling technique. The second paper focuses on open learner models, which is an approach for making a student’s learner model explicit to the student, in order to enhance reflection, self-awareness and self-regulation of the learning process. In “Navigation Support in Complex Learner Models: Assessing Visual Design Alternatives” (Guerra, Schunn, Bull, BarríaPineda and Brusilovsky), six alternative prototypes were","PeriodicalId":54386,"journal":{"name":"New Review of Hypermedia and Multimedia","volume":null,"pages":null},"PeriodicalIF":1.2,"publicationDate":"2018-07-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://sci-hub-pdf.com/10.1080/13614568.2018.1527114","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"45774943","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2018-06-25DOI: 10.1080/13614568.2018.1482375
Julio Guerra, C. Schunn, S. Bull, Jordan Barria-Pineda, Peter Brusilovsky
ABSTRACT Open Learner Models are used in modern e-learning to show system users the content of their learner models. This approach is known to prompt reflection, facilitate planning and navigation. Open Learner Models may show different levels of detail of the underlying learner model, and may structure the information differently. However, a trade-off exists between useful information and the complexity of the information. This paper investigates whether offering richer information is assessed positively by learners and results in more effective support for learning tasks. An interview pre-study revealed which information within the complex learner model is of interest. A controlled user study examined six alternative visualisation prototypes of varying complexity and resulted in the implementation of one of the designs. A second controlled study involved students interacting with variations of the visualisation while searching for suitable learning material, and revealed the value of the design alternative and its variations. The work contributes to developing complex open learner models by stressing the need to balance complexity and support. It also suggests that the expressiveness of open learner models can be improved with visual elements that strategically summarise the complex information being displayed in detail.
{"title":"Navigation support in complex open learner models: assessing visual design alternatives","authors":"Julio Guerra, C. Schunn, S. Bull, Jordan Barria-Pineda, Peter Brusilovsky","doi":"10.1080/13614568.2018.1482375","DOIUrl":"https://doi.org/10.1080/13614568.2018.1482375","url":null,"abstract":"ABSTRACT Open Learner Models are used in modern e-learning to show system users the content of their learner models. This approach is known to prompt reflection, facilitate planning and navigation. Open Learner Models may show different levels of detail of the underlying learner model, and may structure the information differently. However, a trade-off exists between useful information and the complexity of the information. This paper investigates whether offering richer information is assessed positively by learners and results in more effective support for learning tasks. An interview pre-study revealed which information within the complex learner model is of interest. A controlled user study examined six alternative visualisation prototypes of varying complexity and resulted in the implementation of one of the designs. A second controlled study involved students interacting with variations of the visualisation while searching for suitable learning material, and revealed the value of the design alternative and its variations. The work contributes to developing complex open learner models by stressing the need to balance complexity and support. It also suggests that the expressiveness of open learner models can be improved with visual elements that strategically summarise the complex information being displayed in detail.","PeriodicalId":54386,"journal":{"name":"New Review of Hypermedia and Multimedia","volume":null,"pages":null},"PeriodicalIF":1.2,"publicationDate":"2018-06-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://sci-hub-pdf.com/10.1080/13614568.2018.1482375","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"46379225","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2018-06-21DOI: 10.1080/13614568.2018.1477999
J. Okpo, J. Masthoff, Matt Dennis, N. Beacham
ABSTRACT Adapting to learner characteristics is essential when selecting exercises for learners in an intelligent tutoring system. This paper investigates how humans adapt next exercise selection (in particular difficulty level) to learner personality, invested mental effort, and performance to inspire an adaptive exercise selection algorithm. First, the paper describes the investigations to produce validated materials for the main studies, namely the creation and validation of self-esteem personality stories, mental effort statements, and mathematical exercises with varying levels of difficulty. Next, through empirical studies, we investigate the impact on exercise selection of learner's self-esteem (low versus high self-esteem) and effort (minimal, little, moderate, much, and all possible effort). Three studies investigate this for learners who had different performances on a previous exercise: just passing, just failing, and performed well. Participants considered a fictional learner with a certain performance, self-esteem and effort, and selected the difficulty level of the next mathematical exercise. We found that self-esteem, mental effort, and performance all impacted the difficulty level of the exercises selected for learners. Finally, using the results from the studies, we propose an algorithm that selects exercises with varying difficulty levels adapted to learner characteristics.
{"title":"Adapting exercise selection to performance, effort and self-esteem","authors":"J. Okpo, J. Masthoff, Matt Dennis, N. Beacham","doi":"10.1080/13614568.2018.1477999","DOIUrl":"https://doi.org/10.1080/13614568.2018.1477999","url":null,"abstract":"ABSTRACT Adapting to learner characteristics is essential when selecting exercises for learners in an intelligent tutoring system. This paper investigates how humans adapt next exercise selection (in particular difficulty level) to learner personality, invested mental effort, and performance to inspire an adaptive exercise selection algorithm. First, the paper describes the investigations to produce validated materials for the main studies, namely the creation and validation of self-esteem personality stories, mental effort statements, and mathematical exercises with varying levels of difficulty. Next, through empirical studies, we investigate the impact on exercise selection of learner's self-esteem (low versus high self-esteem) and effort (minimal, little, moderate, much, and all possible effort). Three studies investigate this for learners who had different performances on a previous exercise: just passing, just failing, and performed well. Participants considered a fictional learner with a certain performance, self-esteem and effort, and selected the difficulty level of the next mathematical exercise. We found that self-esteem, mental effort, and performance all impacted the difficulty level of the exercises selected for learners. Finally, using the results from the studies, we propose an algorithm that selects exercises with varying difficulty levels adapted to learner characteristics.","PeriodicalId":54386,"journal":{"name":"New Review of Hypermedia and Multimedia","volume":null,"pages":null},"PeriodicalIF":1.2,"publicationDate":"2018-06-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://sci-hub-pdf.com/10.1080/13614568.2018.1477999","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"47437754","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2018-05-28DOI: 10.1080/13614568.2018.1476596
Radek Pelánek, Jirí Rihák
ABSTRACT A common personalisation approach in educational systems is mastery learning. A key step in this approach is a criterion that determines whether a learner has already achieved mastery. We thoroughly analyse several mastery criteria for the basic case of a single well-specified knowledge component. For the analysis we use experiments with both simulated and real data. The results show that the choice of data sources used for mastery decision and the setting of thresholds are more important than the choice of a learner modelling technique. We argue that a simple exponential moving average method is a suitable technique for mastery criterion and discuss techniques for the choice of a mastery threshold. We also propose an extension of the exponential moving average method that takes into account practical aspects like time intensity of items and we report on a practical application of this mastery criterion in a widely used educational system.
{"title":"Analysis and design of mastery learning criteria","authors":"Radek Pelánek, Jirí Rihák","doi":"10.1080/13614568.2018.1476596","DOIUrl":"https://doi.org/10.1080/13614568.2018.1476596","url":null,"abstract":"ABSTRACT A common personalisation approach in educational systems is mastery learning. A key step in this approach is a criterion that determines whether a learner has already achieved mastery. We thoroughly analyse several mastery criteria for the basic case of a single well-specified knowledge component. For the analysis we use experiments with both simulated and real data. The results show that the choice of data sources used for mastery decision and the setting of thresholds are more important than the choice of a learner modelling technique. We argue that a simple exponential moving average method is a suitable technique for mastery criterion and discuss techniques for the choice of a mastery threshold. We also propose an extension of the exponential moving average method that takes into account practical aspects like time intensity of items and we report on a practical application of this mastery criterion in a widely used educational system.","PeriodicalId":54386,"journal":{"name":"New Review of Hypermedia and Multimedia","volume":null,"pages":null},"PeriodicalIF":1.2,"publicationDate":"2018-05-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://sci-hub-pdf.com/10.1080/13614568.2018.1476596","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"45132903","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2018-04-03DOI: 10.1080/13614568.2018.1489001
Mainack Mondal, Leandro Araújo Silva, D. Correa, Fabrício Benevenuto
ABSTRACT Social media platforms provide an inexpensive communication medium that allows anyone to publish content and anyone interested in the content can obtain it. However, this same potential of social media provide space for discourses that are harmful to certain groups of people. Examples of these discourses include bullying, offensive content, and hate speech. Out of these discourses hate speech is rapidly recognized as a serious problem by authorities of many countries. In this paper, we provide the first of a kind systematic large-scale measurement and analysis study of explicit expressions of hate speech in online social media. We aim to understand the abundance of hate speech in online social media, the most common hate expressions, the effect of anonymity on hate speech, the sensitivity of hate speech and the most hated groups across regions. In order to achieve our objectives, we gather traces from two social media systems: Whisper and Twitter. We then develop and validate a methodology to identify hate speech on both of these systems. Our results identify hate speech forms and unveil a set of important patterns, providing not only a broader understanding of online hate speech, but also offering directions for detection and prevention approaches.
{"title":"Characterizing usage of explicit hate expressions in social media","authors":"Mainack Mondal, Leandro Araújo Silva, D. Correa, Fabrício Benevenuto","doi":"10.1080/13614568.2018.1489001","DOIUrl":"https://doi.org/10.1080/13614568.2018.1489001","url":null,"abstract":"ABSTRACT Social media platforms provide an inexpensive communication medium that allows anyone to publish content and anyone interested in the content can obtain it. However, this same potential of social media provide space for discourses that are harmful to certain groups of people. Examples of these discourses include bullying, offensive content, and hate speech. Out of these discourses hate speech is rapidly recognized as a serious problem by authorities of many countries. In this paper, we provide the first of a kind systematic large-scale measurement and analysis study of explicit expressions of hate speech in online social media. We aim to understand the abundance of hate speech in online social media, the most common hate expressions, the effect of anonymity on hate speech, the sensitivity of hate speech and the most hated groups across regions. In order to achieve our objectives, we gather traces from two social media systems: Whisper and Twitter. We then develop and validate a methodology to identify hate speech on both of these systems. Our results identify hate speech forms and unveil a set of important patterns, providing not only a broader understanding of online hate speech, but also offering directions for detection and prevention approaches.","PeriodicalId":54386,"journal":{"name":"New Review of Hypermedia and Multimedia","volume":null,"pages":null},"PeriodicalIF":1.2,"publicationDate":"2018-04-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://sci-hub-pdf.com/10.1080/13614568.2018.1489001","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"47426260","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2018-04-03DOI: 10.1080/13614568.2018.1504520
F. Bonchi, Peter Dolog, D. Helic, P. Vojtás
This Special Issue presents three invited papers, selected from among the best contributions that were presented at the 2017 ACM International Conference on Hypertext and Social Media (HT 2017) held in Prague, Czech Republic on 4–7th July 2017. Since 1987, HT has successfully brought together leading researchers and developers from the Hypertext community. It is concerned with all aspects of modern hypertext research, including social media, adaptation, personalisation, recommendations, user modelling, linked data and semantic web, dynamic and computed hypertext, and its application in digital humanities, as well as with interplay between those aspects such as linking stories with data or linking people with resources. The call for papers of HT 2017 was organised into four technical tracks: Social Networks and Digital Humanities (Linking people), Semantic Web and Linked Data (Linking data), Adaptive Hypertext and Recommendations (Linking resources), News and Storytelling (Linking stories). The Program Committee of HT 2017 accepted 19 papers (acceptance rate 27%) for regular presentation, and an additional 12 short-presentation papers. In addition, the conference featured four demonstrations and two keynotes: Kristina Lerman and Peter Mika. The three papers selected for this Special Issue cover a diverse set of topics, well representing the spectrum of topics that were discussed at HT 2017. The first paper, entitled “Implicit Negative Link Detection on Online Political Networks via Matrix Tri-Factorizations” (Ozer, Yildirim and Davulcu), deals with the prediction of negative connections between users of online political networks. Currently, the majority of social media sites do not support explicit negative links between participating users. However, the very nature of the political discourse often involves users in discussing controversial political issues, which results in a series of agreements and disagreements. The authors present a technically sound approach to extracting negative links from a variety of online political platforms by using a matrix factorisation approach. Matrix factorisation is extended in multiple ways to reflect the information that can be found in the sentiment of the written comments as well as the social balance theory known from the social sciences. The paper concludes with a range of experiments on the real datasets using the Twitter accounts of the politicians of the major UK political parties. The experiments show an improved accuracy of the community detection methods applied on the networks with the extracted negative interaction links as compared to the application of these methods on the networks having only positive links. The second paper, entitled “Hybrid Recommendations by Content-Aligned Bayesian Personalized Ranking” (Peska) focuses on recommender systems that seek to predict the "rating" or "preference" a user would give to an item and hence enabling to display items in order the user might find interest
{"title":"Invited papers from the ACM conference on hypertext and social media","authors":"F. Bonchi, Peter Dolog, D. Helic, P. Vojtás","doi":"10.1080/13614568.2018.1504520","DOIUrl":"https://doi.org/10.1080/13614568.2018.1504520","url":null,"abstract":"This Special Issue presents three invited papers, selected from among the best contributions that were presented at the 2017 ACM International Conference on Hypertext and Social Media (HT 2017) held in Prague, Czech Republic on 4–7th July 2017. Since 1987, HT has successfully brought together leading researchers and developers from the Hypertext community. It is concerned with all aspects of modern hypertext research, including social media, adaptation, personalisation, recommendations, user modelling, linked data and semantic web, dynamic and computed hypertext, and its application in digital humanities, as well as with interplay between those aspects such as linking stories with data or linking people with resources. The call for papers of HT 2017 was organised into four technical tracks: Social Networks and Digital Humanities (Linking people), Semantic Web and Linked Data (Linking data), Adaptive Hypertext and Recommendations (Linking resources), News and Storytelling (Linking stories). The Program Committee of HT 2017 accepted 19 papers (acceptance rate 27%) for regular presentation, and an additional 12 short-presentation papers. In addition, the conference featured four demonstrations and two keynotes: Kristina Lerman and Peter Mika. The three papers selected for this Special Issue cover a diverse set of topics, well representing the spectrum of topics that were discussed at HT 2017. The first paper, entitled “Implicit Negative Link Detection on Online Political Networks via Matrix Tri-Factorizations” (Ozer, Yildirim and Davulcu), deals with the prediction of negative connections between users of online political networks. Currently, the majority of social media sites do not support explicit negative links between participating users. However, the very nature of the political discourse often involves users in discussing controversial political issues, which results in a series of agreements and disagreements. The authors present a technically sound approach to extracting negative links from a variety of online political platforms by using a matrix factorisation approach. Matrix factorisation is extended in multiple ways to reflect the information that can be found in the sentiment of the written comments as well as the social balance theory known from the social sciences. The paper concludes with a range of experiments on the real datasets using the Twitter accounts of the politicians of the major UK political parties. The experiments show an improved accuracy of the community detection methods applied on the networks with the extracted negative interaction links as compared to the application of these methods on the networks having only positive links. The second paper, entitled “Hybrid Recommendations by Content-Aligned Bayesian Personalized Ranking” (Peska) focuses on recommender systems that seek to predict the \"rating\" or \"preference\" a user would give to an item and hence enabling to display items in order the user might find interest","PeriodicalId":54386,"journal":{"name":"New Review of Hypermedia and Multimedia","volume":null,"pages":null},"PeriodicalIF":1.2,"publicationDate":"2018-04-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://sci-hub-pdf.com/10.1080/13614568.2018.1504520","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"48194211","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2018-04-03DOI: 10.1080/13614568.2018.1489002
Ladislav Peška
ABSTRACT In many application domains of recommender systems, content-based (CB) information are available for users, objects or both. CB information plays an important role in the process of recommendation, especially in cold-start scenarios, where the volume of feedback data is low. However, CB information may come from several, possibly external, sources varying in reliability, coverage or relevance to the recommending task. Therefore, each content source or attribute possess a different level of informativeness, which should be taken into consideration during the process of recommendation. In this paper, we propose a Content-Aligned Bayesian Personalized Ranking Matrix Factorization method (CABPR), extending Bayesian Personalized Ranking Matrix Factorization (BPR) by incorporating multiple sources of content information into the BPR’s optimization procedure. The working principle of CABPR is to calculate user-to-user and object-to-object similarity matrices based on the content information and penalize differences in latent factors of closely related users’ or objects’. CABPR further estimates relevance of similarity matrices as a part of the optimization procedure. CABPR method is a significant extension of a previously published BPR_MCA method, featuring additional variants of optimization criterion and improved optimization procedure. Four variants of CABPR were evaluated on two publicly available datasets: MovieLens 1M dataset, extended by data from IMDB, DBTropes and ZIP code statistics and LOD-RecSys dataset extended by the information available from DBPedia. Experiments shown that CABPR significantly improves over standard BPR as well as BPR_MCA method w.r.t. several cold-start scenarios.
{"title":"Hybrid recommendations by content-aligned Bayesian personalized ranking","authors":"Ladislav Peška","doi":"10.1080/13614568.2018.1489002","DOIUrl":"https://doi.org/10.1080/13614568.2018.1489002","url":null,"abstract":"ABSTRACT In many application domains of recommender systems, content-based (CB) information are available for users, objects or both. CB information plays an important role in the process of recommendation, especially in cold-start scenarios, where the volume of feedback data is low. However, CB information may come from several, possibly external, sources varying in reliability, coverage or relevance to the recommending task. Therefore, each content source or attribute possess a different level of informativeness, which should be taken into consideration during the process of recommendation. In this paper, we propose a Content-Aligned Bayesian Personalized Ranking Matrix Factorization method (CABPR), extending Bayesian Personalized Ranking Matrix Factorization (BPR) by incorporating multiple sources of content information into the BPR’s optimization procedure. The working principle of CABPR is to calculate user-to-user and object-to-object similarity matrices based on the content information and penalize differences in latent factors of closely related users’ or objects’. CABPR further estimates relevance of similarity matrices as a part of the optimization procedure. CABPR method is a significant extension of a previously published BPR_MCA method, featuring additional variants of optimization criterion and improved optimization procedure. Four variants of CABPR were evaluated on two publicly available datasets: MovieLens 1M dataset, extended by data from IMDB, DBTropes and ZIP code statistics and LOD-RecSys dataset extended by the information available from DBPedia. Experiments shown that CABPR significantly improves over standard BPR as well as BPR_MCA method w.r.t. several cold-start scenarios.","PeriodicalId":54386,"journal":{"name":"New Review of Hypermedia and Multimedia","volume":null,"pages":null},"PeriodicalIF":1.2,"publicationDate":"2018-04-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://sci-hub-pdf.com/10.1080/13614568.2018.1489002","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"47916531","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2018-04-03DOI: 10.1080/13614568.2018.1482964
M. Ozer, M. Yildirim, H. Davulcu
ABSTRACT Political conversations have become a ubiquitous part of social media. When users interact and engage in discussions, there are usually two mediums available to them; textual conversations and platform-specific interactions such as like, share (Facebook) or retweet (Twitter). Major social media platforms do not facilitate users with negative interaction options. However, many political network analysis tasks rely on not only positive but also negative linkages. Thus, detecting implicit negative links is an important and a challenging task. In this work, we propose an unsupervised framework utilising positive interactions, sentiment cues, and socially balanced triads for detecting implicit negative links. We also present an online variant of it for streaming data tasks. We show the effectiveness of both frameworks with experiments on two annotated datasets of politician Twitter accounts. Our experiments show that the proposed frameworks outperform other well-known and proposed baselines. To illustrate the detected implicit negative links' contribution, we compare the community detection accuracies using unsigned and signed networks. Experimental results using detected negative links show superiority on the three datasets where the camps are known a priori. We also present qualitative evaluations of polarisation patterns between communities which are only possible in the presence of negative links.
{"title":"Implicit negative link detection on online political networks via matrix tri-factorizations","authors":"M. Ozer, M. Yildirim, H. Davulcu","doi":"10.1080/13614568.2018.1482964","DOIUrl":"https://doi.org/10.1080/13614568.2018.1482964","url":null,"abstract":"ABSTRACT Political conversations have become a ubiquitous part of social media. When users interact and engage in discussions, there are usually two mediums available to them; textual conversations and platform-specific interactions such as like, share (Facebook) or retweet (Twitter). Major social media platforms do not facilitate users with negative interaction options. However, many political network analysis tasks rely on not only positive but also negative linkages. Thus, detecting implicit negative links is an important and a challenging task. In this work, we propose an unsupervised framework utilising positive interactions, sentiment cues, and socially balanced triads for detecting implicit negative links. We also present an online variant of it for streaming data tasks. We show the effectiveness of both frameworks with experiments on two annotated datasets of politician Twitter accounts. Our experiments show that the proposed frameworks outperform other well-known and proposed baselines. To illustrate the detected implicit negative links' contribution, we compare the community detection accuracies using unsigned and signed networks. Experimental results using detected negative links show superiority on the three datasets where the camps are known a priori. We also present qualitative evaluations of polarisation patterns between communities which are only possible in the presence of negative links.","PeriodicalId":54386,"journal":{"name":"New Review of Hypermedia and Multimedia","volume":null,"pages":null},"PeriodicalIF":1.2,"publicationDate":"2018-04-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://sci-hub-pdf.com/10.1080/13614568.2018.1482964","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"43920227","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2018-01-02DOI: 10.1080/13614568.2018.1460403
Yuan Meng, Hongwei Wang, Lijuan Zheng
ABSTRACT Facing with thousands of online product reviews, consumers usually pay close attention to those valuable ones which provide more specific and credible evaluations on products. Whether a close association exists between product review quality and sales is thus examined in this paper. By employing text mining techniques on multiple review features, a review is measured as one of the following two levels: high-quality or low-quality. In doing so, aggregate quality level of product’s whole reviews is also identified. Then, a two-level econometrical analysis is conducted on the real datasets from Amazon.cn. The results reveal that aggregate quality level of positive reviews and negative reviews interactively influence sales. In the situation the aggregate quality level of positive reviews is high meanwhile that of negative reviews’ is low, product sale is the highest, while in the opposite situation product sale is the lowest. The results also reveal that consumers understand product’s value from weighting positive and negative reviews of high-quality level, which then positively relates to product sales and exerts a dynamic effect on sales by the moderating role of product selling stage and popularity. The paper innovatively integrates the quantitative and qualitative characteristics of reviews to estimate their economic effect.
{"title":"Impact of online word-of-mouth on sales: the moderating role of product review quality","authors":"Yuan Meng, Hongwei Wang, Lijuan Zheng","doi":"10.1080/13614568.2018.1460403","DOIUrl":"https://doi.org/10.1080/13614568.2018.1460403","url":null,"abstract":"ABSTRACT Facing with thousands of online product reviews, consumers usually pay close attention to those valuable ones which provide more specific and credible evaluations on products. Whether a close association exists between product review quality and sales is thus examined in this paper. By employing text mining techniques on multiple review features, a review is measured as one of the following two levels: high-quality or low-quality. In doing so, aggregate quality level of product’s whole reviews is also identified. Then, a two-level econometrical analysis is conducted on the real datasets from Amazon.cn. The results reveal that aggregate quality level of positive reviews and negative reviews interactively influence sales. In the situation the aggregate quality level of positive reviews is high meanwhile that of negative reviews’ is low, product sale is the highest, while in the opposite situation product sale is the lowest. The results also reveal that consumers understand product’s value from weighting positive and negative reviews of high-quality level, which then positively relates to product sales and exerts a dynamic effect on sales by the moderating role of product selling stage and popularity. The paper innovatively integrates the quantitative and qualitative characteristics of reviews to estimate their economic effect.","PeriodicalId":54386,"journal":{"name":"New Review of Hypermedia and Multimedia","volume":null,"pages":null},"PeriodicalIF":1.2,"publicationDate":"2018-01-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://sci-hub-pdf.com/10.1080/13614568.2018.1460403","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"43911767","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2018-01-02DOI: 10.1080/13614568.2018.1448007
M. S. Missen, Mickaël Coustaty, N. Salamat, V. B. Surya Prasath
ABSTRACT The amount of opinionated data on the web has exponentially increased especially after the emergence of online social networks. To deal with these huge deluge of data, we need to have robust mechanisms that can help identify all aspects of opinion segment and support the automatic processing of opinion data. Recently, there have been a few developments made in this direction, and different sentiment annotation schemes have been proposed such as the SentiML, OpinionMiningML, and EmotionML. In this work, we propose SentiML++, an extension of SentiML with a focus on annotating opinions, and further answering aspects of the general question “who has what opinion about whom in which context?”. A detailed comparison with SentiML and other existing annotation schemes is also presented. The data collection annotated with SentiML has been annotated with SentiML++ and is available for download for further research purposes. Experiments with data collections annotated with SentiML and SentiML++ proves that SentiML++ is a significant and valuable addition to SentiML.
{"title":"SentiML ++: an extension of the SentiML sentiment annotation scheme","authors":"M. S. Missen, Mickaël Coustaty, N. Salamat, V. B. Surya Prasath","doi":"10.1080/13614568.2018.1448007","DOIUrl":"https://doi.org/10.1080/13614568.2018.1448007","url":null,"abstract":"ABSTRACT The amount of opinionated data on the web has exponentially increased especially after the emergence of online social networks. To deal with these huge deluge of data, we need to have robust mechanisms that can help identify all aspects of opinion segment and support the automatic processing of opinion data. Recently, there have been a few developments made in this direction, and different sentiment annotation schemes have been proposed such as the SentiML, OpinionMiningML, and EmotionML. In this work, we propose SentiML++, an extension of SentiML with a focus on annotating opinions, and further answering aspects of the general question “who has what opinion about whom in which context?”. A detailed comparison with SentiML and other existing annotation schemes is also presented. The data collection annotated with SentiML has been annotated with SentiML++ and is available for download for further research purposes. Experiments with data collections annotated with SentiML and SentiML++ proves that SentiML++ is a significant and valuable addition to SentiML.","PeriodicalId":54386,"journal":{"name":"New Review of Hypermedia and Multimedia","volume":null,"pages":null},"PeriodicalIF":1.2,"publicationDate":"2018-01-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://sci-hub-pdf.com/10.1080/13614568.2018.1448007","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"41873252","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}