{"title":"WIMS 2020: The 10th International Conference on Web Intelligence, Mining and Semantics, Biarritz, France, June 30 - July 3, 2020","authors":"","doi":"10.1145/3405962","DOIUrl":"https://doi.org/10.1145/3405962","url":null,"abstract":"","PeriodicalId":20685,"journal":{"name":"Proceedings of the 7th International Conference on Web Intelligence, Mining and Semantics","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2020-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"77462984","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The evolution of the social web opens a new channel that allows bidirectional electronic interactions directly with customers in real-time. By accessing social media content via application programming interfaces (API), businesses may enrich their information on customers, which are usually represented in customer profiles. However, these profiles are often incomplete since additional meaningful data on the customer's context are missing. Based on this idea, this research in progress paper describes first ideas on how data from social media, which are available through API, may be matched with customer profiles via a customer context model.
{"title":"Capturing customer context from social media: mapping social media API and CRM profile data","authors":"Matthias Wittwer, Olaf Reinhold, R. Alt","doi":"10.1145/3106426.3117762","DOIUrl":"https://doi.org/10.1145/3106426.3117762","url":null,"abstract":"The evolution of the social web opens a new channel that allows bidirectional electronic interactions directly with customers in real-time. By accessing social media content via application programming interfaces (API), businesses may enrich their information on customers, which are usually represented in customer profiles. However, these profiles are often incomplete since additional meaningful data on the customer's context are missing. Based on this idea, this research in progress paper describes first ideas on how data from social media, which are available through API, may be matched with customer profiles via a customer context model.","PeriodicalId":20685,"journal":{"name":"Proceedings of the 7th International Conference on Web Intelligence, Mining and Semantics","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2017-08-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"79476348","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The problem with distributed representations generated by neural networks is that the meaning of the features is difficult to understand. We propose a new method that gives a specific meaning to each node of a hidden layer by introducing a manually created word semantic vector dictionary into the initial weights and by using paragraph vector models. Our experimental results demonstrated that weights obtained based on learning and weights based on the dictionary are more strongly correlated in a closed test and more weakly correlated in an open test, compared with the results of a control test. Additionally, we found that the learned vector are better than the performance of the existing paragraph vector in the evaluation of the sentiment analysis task. Finally, we determined the readability of document embedding in a user test. The definition of readability in this paper is that people can understand the meaning of large weighted features of distributed representations. A total of 52.4% of the top five weighted hidden nodes were related to tweets where one of the paragraph vector models learned the document embedding. Because each hidden node maintains a specific meaning, the proposed method succeeds in improving readability.
{"title":"Semantically readable distributed representation learning for social media mining","authors":"Ikuo Keshi, Yumiko Suzuki, Koichiro Yoshino, Satoshi Nakamura","doi":"10.1145/3106426.3106521","DOIUrl":"https://doi.org/10.1145/3106426.3106521","url":null,"abstract":"The problem with distributed representations generated by neural networks is that the meaning of the features is difficult to understand. We propose a new method that gives a specific meaning to each node of a hidden layer by introducing a manually created word semantic vector dictionary into the initial weights and by using paragraph vector models. Our experimental results demonstrated that weights obtained based on learning and weights based on the dictionary are more strongly correlated in a closed test and more weakly correlated in an open test, compared with the results of a control test. Additionally, we found that the learned vector are better than the performance of the existing paragraph vector in the evaluation of the sentiment analysis task. Finally, we determined the readability of document embedding in a user test. The definition of readability in this paper is that people can understand the meaning of large weighted features of distributed representations. A total of 52.4% of the top five weighted hidden nodes were related to tweets where one of the paragraph vector models learned the document embedding. Because each hidden node maintains a specific meaning, the proposed method succeeds in improving readability.","PeriodicalId":20685,"journal":{"name":"Proceedings of the 7th International Conference on Web Intelligence, Mining and Semantics","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2017-08-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"85664907","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
In this investigation, we discuss high order data structure (called tensor) for efficient information retrieval and show especially how well reduction techniques of dimensionality goes while preserving Euclid distance between information. High order data structure requires much amount of space. One of the effective approaches comes from dimensionality reduction such as Latent Semantic Indexing (LSI) and Random Projection (RP) which allows us to reduce complexity of time and space dramatically. The reduction techniques can be applied to high order data structure. Here we examine High Order Random Projection (HORP) which provides us with efficient information retrieval keeping feasible dimensionality reduction.
{"title":"Modeling random projection for tensor objects","authors":"Ryohei Yokobayashi, T. Miura","doi":"10.1145/3106426.3106504","DOIUrl":"https://doi.org/10.1145/3106426.3106504","url":null,"abstract":"In this investigation, we discuss high order data structure (called tensor) for efficient information retrieval and show especially how well reduction techniques of dimensionality goes while preserving Euclid distance between information. High order data structure requires much amount of space. One of the effective approaches comes from dimensionality reduction such as Latent Semantic Indexing (LSI) and Random Projection (RP) which allows us to reduce complexity of time and space dramatically. The reduction techniques can be applied to high order data structure. Here we examine High Order Random Projection (HORP) which provides us with efficient information retrieval keeping feasible dimensionality reduction.","PeriodicalId":20685,"journal":{"name":"Proceedings of the 7th International Conference on Web Intelligence, Mining and Semantics","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2017-08-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"80960740","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
It is important to analyze the reputation or demands for a regional event, such as a school festival. In our work, we use sentiment polarity classification in order to coordinate regional event reputation. We proposed sentiment polarity classification based on bag-of-words models in the previous works. To get over the traditional models, we proposed several classifier models based on deep learning models. As the application, we also described the overview of a system supports to analyze regional event reputation and an example of regional event analysis using our system. In this paper, we described how to improve the performance of the sentiment polarity classification using deep learning models. We compared the performance of four models in terms of the classification accuracy and the training speed. We found the Convolutional Neural Networks based model, three words convolutions, was the best model among the four models.
{"title":"A sentiment polarity classifier for regional event reputation analysis","authors":"Tatsuya Ohbe, Tadachika Ozono, T. Shintani","doi":"10.1145/3106426.3109416","DOIUrl":"https://doi.org/10.1145/3106426.3109416","url":null,"abstract":"It is important to analyze the reputation or demands for a regional event, such as a school festival. In our work, we use sentiment polarity classification in order to coordinate regional event reputation. We proposed sentiment polarity classification based on bag-of-words models in the previous works. To get over the traditional models, we proposed several classifier models based on deep learning models. As the application, we also described the overview of a system supports to analyze regional event reputation and an example of regional event analysis using our system. In this paper, we described how to improve the performance of the sentiment polarity classification using deep learning models. We compared the performance of four models in terms of the classification accuracy and the training speed. We found the Convolutional Neural Networks based model, three words convolutions, was the best model among the four models.","PeriodicalId":20685,"journal":{"name":"Proceedings of the 7th International Conference on Web Intelligence, Mining and Semantics","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2017-08-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"83932117","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Maintaining online communities is vital in order to increase and retain their economic and social value. That is why community managers look to gauge the success of their communities by measuring a variety of user behaviour, such as member activity, turnover and interaction. However, such communities vary widely in their purpose, implementation and user demographics, and although many success indicators have been proposed in the literature, we will show that there is no one-fits-all approach to community success: Different success criteria depend on different user behaviour. To demonstrate this, we put together a set of user behaviour features, including many that have been used in the literature as indicators of success, and then we define and predict community success in three different types of online communities: Questions & Answers (Q&A), Healthcare and Emotional Support (Life & Health), and Encyclopaedic Knowledge Creation. The results show that it is feasible to relate community success to specific user behaviour with an accuracy of 0.67--0.93 F1 score and 0.77--1.0 AUC.
{"title":"The path to success: a study of user behaviour and success criteria in online communities","authors":"Erik Aumayr, Conor Hayes","doi":"10.1145/3106426.3106469","DOIUrl":"https://doi.org/10.1145/3106426.3106469","url":null,"abstract":"Maintaining online communities is vital in order to increase and retain their economic and social value. That is why community managers look to gauge the success of their communities by measuring a variety of user behaviour, such as member activity, turnover and interaction. However, such communities vary widely in their purpose, implementation and user demographics, and although many success indicators have been proposed in the literature, we will show that there is no one-fits-all approach to community success: Different success criteria depend on different user behaviour. To demonstrate this, we put together a set of user behaviour features, including many that have been used in the literature as indicators of success, and then we define and predict community success in three different types of online communities: Questions & Answers (Q&A), Healthcare and Emotional Support (Life & Health), and Encyclopaedic Knowledge Creation. The results show that it is feasible to relate community success to specific user behaviour with an accuracy of 0.67--0.93 F1 score and 0.77--1.0 AUC.","PeriodicalId":20685,"journal":{"name":"Proceedings of the 7th International Conference on Web Intelligence, Mining and Semantics","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2017-08-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"82953884","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Ke Xu, Y. Cai, Huaqing Min, Xushen Zheng, Haoran Xie, Tak-Lam Wong
The rapid growth of population has posed a challenge to people for discovering new followees in uni-directional social networks. Intuitively, a user's adoption of others as followees may motivated by her interest as well as social connection. Therefore, it is worth-while to consider both factors at the same time for better recommendations. Previous recommender works on implicit follow or not feedbacks become unqualified, mainly because of the coarse users' preferences inferring, which cannot distinguish whether one follows the other is based on her social connection or individual interest. In this paper, we present a new user recommendation method which is capable of recommending candidate followees who have similar interest and closer social connection relevant to a target user. As its core, a novel topic model namely UIS-LDA is designed to jointly model a user's preferences with respect to the set of latent interest topics and social topics. The experiments using Twitter dataset proves that our proposed method effective in improving the Precision, Conversion Rate F1 score and NDCG.
{"title":"UIS-LDA: a user recommendation based on social connections and interests of users in uni-directional social networks","authors":"Ke Xu, Y. Cai, Huaqing Min, Xushen Zheng, Haoran Xie, Tak-Lam Wong","doi":"10.1145/3106426.3106494","DOIUrl":"https://doi.org/10.1145/3106426.3106494","url":null,"abstract":"The rapid growth of population has posed a challenge to people for discovering new followees in uni-directional social networks. Intuitively, a user's adoption of others as followees may motivated by her interest as well as social connection. Therefore, it is worth-while to consider both factors at the same time for better recommendations. Previous recommender works on implicit follow or not feedbacks become unqualified, mainly because of the coarse users' preferences inferring, which cannot distinguish whether one follows the other is based on her social connection or individual interest. In this paper, we present a new user recommendation method which is capable of recommending candidate followees who have similar interest and closer social connection relevant to a target user. As its core, a novel topic model namely UIS-LDA is designed to jointly model a user's preferences with respect to the set of latent interest topics and social topics. The experiments using Twitter dataset proves that our proposed method effective in improving the Precision, Conversion Rate F1 score and NDCG.","PeriodicalId":20685,"journal":{"name":"Proceedings of the 7th International Conference on Web Intelligence, Mining and Semantics","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2017-08-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"91474433","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Michael Aleithe, P. Skowron, Bogdan Franczyk, B. Sommer
In the digital age, where research is data-driven, understanding all involved fields of research becomes more and more important. Understanding various data sources within interdisciplinary research and beyond domain boundaries is a significant core competency. All participants should have a same-level understanding of significant information, which can be created from various data sources. Based on this fact, the paper at hand demonstrates a modeling approach for the generation of a unified data model in terms of smart urban objects. These smart objects are represented by interconnected data structures which is a prime example in context of Internet of Things. Further, an implementation of the graph database Neo4J and a correlated visualization of intuitive structuring of data sources beyond domain boundaries will be demonstrated.
{"title":"Data modeling of smart urban object networks","authors":"Michael Aleithe, P. Skowron, Bogdan Franczyk, B. Sommer","doi":"10.1145/3106426.3117759","DOIUrl":"https://doi.org/10.1145/3106426.3117759","url":null,"abstract":"In the digital age, where research is data-driven, understanding all involved fields of research becomes more and more important. Understanding various data sources within interdisciplinary research and beyond domain boundaries is a significant core competency. All participants should have a same-level understanding of significant information, which can be created from various data sources. Based on this fact, the paper at hand demonstrates a modeling approach for the generation of a unified data model in terms of smart urban objects. These smart objects are represented by interconnected data structures which is a prime example in context of Internet of Things. Further, an implementation of the graph database Neo4J and a correlated visualization of intuitive structuring of data sources beyond domain boundaries will be demonstrated.","PeriodicalId":20685,"journal":{"name":"Proceedings of the 7th International Conference on Web Intelligence, Mining and Semantics","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2017-08-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"79454549","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The paradigm of lifelong learning supported by technology is redefining the way we learn as well as the way we search and consume the ever growing corpus of information available in the Web to acquire knowledge on a particular subject. This research addresses the problem of finding and organizing learning content to support self-directed learners in achieving a learning goal through the search, selection and sequencing of Web content that might or might not have been conceived as learning resources. We plan to build an automatic process driven by the knowledge available in datasets belonging to the Linked Open Initiative and open non-structured information such as courses syllabi and books table of contents. Our proposed service have two main components: (i) a graph of interrelated learning concepts from which is possible infer what concepts must be addressed first before others in the learning process (prerequisite relationships), and (ii) a component for the creation of learning resources sequences based on a learning goal and a learner profile.
{"title":"Towards automatic learning content sequence via linked open data","authors":"R. Manrique","doi":"10.1145/3106426.3110322","DOIUrl":"https://doi.org/10.1145/3106426.3110322","url":null,"abstract":"The paradigm of lifelong learning supported by technology is redefining the way we learn as well as the way we search and consume the ever growing corpus of information available in the Web to acquire knowledge on a particular subject. This research addresses the problem of finding and organizing learning content to support self-directed learners in achieving a learning goal through the search, selection and sequencing of Web content that might or might not have been conceived as learning resources. We plan to build an automatic process driven by the knowledge available in datasets belonging to the Linked Open Initiative and open non-structured information such as courses syllabi and books table of contents. Our proposed service have two main components: (i) a graph of interrelated learning concepts from which is possible infer what concepts must be addressed first before others in the learning process (prerequisite relationships), and (ii) a component for the creation of learning resources sequences based on a learning goal and a learner profile.","PeriodicalId":20685,"journal":{"name":"Proceedings of the 7th International Conference on Web Intelligence, Mining and Semantics","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2017-08-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"89357084","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Recommendation system is able to recommend items that are likely to be preferred by the user. Hybrid recommender systems combine the advantages of the collaborative filtering and content-based filtering for improved recommendation. Hybrid recommendation methods use as many significant factors as possible to generate recommendation, which is practically very functional in real scenarios. However, such method has not been applied to book recommendation yet. Thus, in this paper, we propose a set of novel features which can be categorized into three types: latent features, derived features and content features. These features can be combined to form a new hybrid feature vector containing rating information and content information. Then, we adopted learning-to-rank to use the proposed feature vector as the input for book recommendation. Collaborative Ranking (CR) and Probabilistic Matrix Factorization (PMF) are compared with our proposed method. The experimental results show that the proposed method outperforms CR and PMF. It shows that, on NDCG@1, PMF achieves 0.713818, CR achieves 0.690072 vs. our method achieves 0.742689 which is 4.04% over PMF and 7.62% over CR.
{"title":"A novel learning-to-rank based hybrid method for book recommendation","authors":"Y. Liu, Jiajun Yang","doi":"10.1145/3106426.3106547","DOIUrl":"https://doi.org/10.1145/3106426.3106547","url":null,"abstract":"Recommendation system is able to recommend items that are likely to be preferred by the user. Hybrid recommender systems combine the advantages of the collaborative filtering and content-based filtering for improved recommendation. Hybrid recommendation methods use as many significant factors as possible to generate recommendation, which is practically very functional in real scenarios. However, such method has not been applied to book recommendation yet. Thus, in this paper, we propose a set of novel features which can be categorized into three types: latent features, derived features and content features. These features can be combined to form a new hybrid feature vector containing rating information and content information. Then, we adopted learning-to-rank to use the proposed feature vector as the input for book recommendation. Collaborative Ranking (CR) and Probabilistic Matrix Factorization (PMF) are compared with our proposed method. The experimental results show that the proposed method outperforms CR and PMF. It shows that, on NDCG@1, PMF achieves 0.713818, CR achieves 0.690072 vs. our method achieves 0.742689 which is 4.04% over PMF and 7.62% over CR.","PeriodicalId":20685,"journal":{"name":"Proceedings of the 7th International Conference on Web Intelligence, Mining and Semantics","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2017-08-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"86693356","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：