Studying behavior of systems through networks is important because it allows to understand them and make decisions based on this knowledge. Community detection is one of the tools used in this sense, for detection of groups in graphs. This can be done not only considering connections between nodes, but also including their attributes. Also, objects can be part of different groups with varying degrees, so overlapping fuzzy assignment is relevant in this context. Furthermore, most networks change overtime, so including this aspect also enhance the benefits of using community detection. Hence, in this doctoral thesis we propose to design models for overlapping community detection for static and dynamic networks with node attributes. Firstly, an approach based on a nonnegative matrix factorization generative model that automatically detects the number of communities in the network, is designed. Secondly, tensor factorization is used in order to overcome some of the challenges faced in the first model.
{"title":"Overlapping Community Detection in Static and Dynamic Networks","authors":"Renny Márquez","doi":"10.1145/3336191.3372185","DOIUrl":"https://doi.org/10.1145/3336191.3372185","url":null,"abstract":"Studying behavior of systems through networks is important because it allows to understand them and make decisions based on this knowledge. Community detection is one of the tools used in this sense, for detection of groups in graphs. This can be done not only considering connections between nodes, but also including their attributes. Also, objects can be part of different groups with varying degrees, so overlapping fuzzy assignment is relevant in this context. Furthermore, most networks change overtime, so including this aspect also enhance the benefits of using community detection. Hence, in this doctoral thesis we propose to design models for overlapping community detection for static and dynamic networks with node attributes. Firstly, an approach based on a nonnegative matrix factorization generative model that automatically detects the number of communities in the network, is designed. Secondly, tensor factorization is used in order to overcome some of the challenges faced in the first model.","PeriodicalId":319008,"journal":{"name":"Proceedings of the 13th International Conference on Web Search and Data Mining","volume":"17 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-01-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122257595","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}
H. Steck, Maria Dimakopoulou, Nickolai Riabov, T. Jebara
The Sparse Linear Method (SLIM) is a well-established approach for top-N recommendations. This article proposes several improvements that are enabled by the Alternating Directions Method of Multipliers (ADMM), a well-known optimization method with many application areas. First, we show that optimizing the original SLIM-objective by ADMM results in an approach where the training time is independent of the number of users in the training data, and hence trivially scales to large numbers of users. Second, the flexibility of ADMM allows us to switch on and off the various constraints and regularization terms in the original SLIM-objective, in order to empirically assess their contributions to ranking accuracy on given data. Third, we also propose two extensions to the original SLIM training-objective in order to improve recommendation accuracy further without increasing the computational cost. In our experiments on three well-known data-sets, we first compare to the original SLIM-implementation and find that not only ADMM reduces training time considerably, but also achieves an improvement in recommendation accuracy due to better optimization. We then compare to various state-of-the-art approaches and observe up to 25% improvement in recommendation accuracy in our experiments. Finally, we evaluate the importance of sparsity and the non-negativity constraint in the original SLIM-objective with sub-sampling experiments that simulate scenarios of cold-starting and large catalog sizes compared to relatively small user base, which often occur in practice.
{"title":"ADMM SLIM: Sparse Recommendations for Many Users","authors":"H. Steck, Maria Dimakopoulou, Nickolai Riabov, T. Jebara","doi":"10.1145/3336191.3371774","DOIUrl":"https://doi.org/10.1145/3336191.3371774","url":null,"abstract":"The Sparse Linear Method (SLIM) is a well-established approach for top-N recommendations. This article proposes several improvements that are enabled by the Alternating Directions Method of Multipliers (ADMM), a well-known optimization method with many application areas. First, we show that optimizing the original SLIM-objective by ADMM results in an approach where the training time is independent of the number of users in the training data, and hence trivially scales to large numbers of users. Second, the flexibility of ADMM allows us to switch on and off the various constraints and regularization terms in the original SLIM-objective, in order to empirically assess their contributions to ranking accuracy on given data. Third, we also propose two extensions to the original SLIM training-objective in order to improve recommendation accuracy further without increasing the computational cost. In our experiments on three well-known data-sets, we first compare to the original SLIM-implementation and find that not only ADMM reduces training time considerably, but also achieves an improvement in recommendation accuracy due to better optimization. We then compare to various state-of-the-art approaches and observe up to 25% improvement in recommendation accuracy in our experiments. Finally, we evaluate the importance of sparsity and the non-negativity constraint in the original SLIM-objective with sub-sampling experiments that simulate scenarios of cold-starting and large catalog sizes compared to relatively small user base, which often occur in practice.","PeriodicalId":319008,"journal":{"name":"Proceedings of the 13th International Conference on Web Search and Data Mining","volume":"31 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-01-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122385755","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 essential to fully understand user intents for the optimization of downstream tasks such as document ranking and query suggestion in web search. As users tend to submit ambiguous queries, numer- ous studies utilize contextual information such as query sequence and user clicks for the auxiliary of user intent modeling. Most of these work adopted Recurrent Neural Network (RNN) based frame- works to encode sequential information within a session, which is hard to realize parallel computation. To this end, we plan to adopt attention-based units to generate context-aware representations for elements in sessions. As intra-session contexts are deficient for handling the data sparsity and cold-start problems in session search, we would also attempt to integrate cross-session dependen- cies by constructing session graphs on the whole corpus to enrich the representation of queries and documents.
{"title":"Beyond Sessions: Exploiting Hybrid Contextual Information for Web Search","authors":"Jia Chen","doi":"10.1145/3336191.3372179","DOIUrl":"https://doi.org/10.1145/3336191.3372179","url":null,"abstract":"It is essential to fully understand user intents for the optimization of downstream tasks such as document ranking and query suggestion in web search. As users tend to submit ambiguous queries, numer- ous studies utilize contextual information such as query sequence and user clicks for the auxiliary of user intent modeling. Most of these work adopted Recurrent Neural Network (RNN) based frame- works to encode sequential information within a session, which is hard to realize parallel computation. To this end, we plan to adopt attention-based units to generate context-aware representations for elements in sessions. As intra-session contexts are deficient for handling the data sparsity and cold-start problems in session search, we would also attempt to integrate cross-session dependen- cies by constructing session graphs on the whole corpus to enrich the representation of queries and documents.","PeriodicalId":319008,"journal":{"name":"Proceedings of the 13th International Conference on Web Search and Data Mining","volume":"44 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-01-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126150727","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 goal of personalized search is to tailor the document ranking list to meet user's individual needs. Previous studies showed users usually look for the information that has been searched before. This is called re-finding behavior which is widely explored in existing personalized search approaches. However, most existing methods for identifying re-finding behavior focus on simple lexical similarities between queries. In this paper, we propose to construct memory networks (MN) to support the identification of more complex re-finding behavior. Specifically, incorporating semantic information, we devise two external memories to make an expansion of re-finding based on the query and the document respectively. We further design an intent memory to recognize session-based re-finding behavior. Endowed with these memory networks, we can build a fine-grained user model dynamically based on the current query and documents, and use the model to re-rank the results. Experimental results show the significant improvement of our model compared with traditional methods.
{"title":"Enhancing Re-finding Behavior with External Memories for Personalized Search","authors":"Yujia Zhou, Zhicheng Dou, Ji-rong Wen","doi":"10.1145/3336191.3371794","DOIUrl":"https://doi.org/10.1145/3336191.3371794","url":null,"abstract":"The goal of personalized search is to tailor the document ranking list to meet user's individual needs. Previous studies showed users usually look for the information that has been searched before. This is called re-finding behavior which is widely explored in existing personalized search approaches. However, most existing methods for identifying re-finding behavior focus on simple lexical similarities between queries. In this paper, we propose to construct memory networks (MN) to support the identification of more complex re-finding behavior. Specifically, incorporating semantic information, we devise two external memories to make an expansion of re-finding based on the query and the document respectively. We further design an intent memory to recognize session-based re-finding behavior. Endowed with these memory networks, we can build a fine-grained user model dynamically based on the current query and documents, and use the model to re-rank the results. Experimental results show the significant improvement of our model compared with traditional methods.","PeriodicalId":319008,"journal":{"name":"Proceedings of the 13th International Conference on Web Search and Data Mining","volume":"15 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-01-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114497326","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}
Xin Mao, Wenting Wang, Huimin Xu, Man Lan, Yuanbin Wu
Entity alignment to find equivalent entities in cross-lingual Knowledge Graphs (KGs) plays a vital role in automatically integrating multiple KGs. Existing translation-based entity alignment methods jointly model the cross-lingual knowledge and monolingual knowledge into one unified optimization problem. On the other hand, the Graph Neural Network (GNN) based methods either ignore the node differentiations, or represent relation through entity or triple instances. They all fail to model the meta semantics embedded in relation nor complex relations such as n-to-n and multi-graphs. To tackle these challenges, we propose a novel Meta Relation Aware Entity Alignment (MRAEA) to directly model cross-lingual entity embeddings by attending over the node's incoming and outgoing neighbors and its connected relations' meta semantics. In addition, we also propose a simple and effective bi-directional iterative strategy to add new aligned seeds during training. Our experiments on all three benchmark entity alignment datasets show that our approach consistently outperforms the state-of-the-art methods, exceeding by 15%-58% on Hit@1. Through an extensive ablation study, we validate that the proposed meta relation aware representations, relation aware self-attention and bi-directional iterative strategy of new seed selection all make contributions to significant performance improvement. The code is available at https://github.com/MaoXinn/MRAEA.
{"title":"MRAEA","authors":"Xin Mao, Wenting Wang, Huimin Xu, Man Lan, Yuanbin Wu","doi":"10.1145/3336191.3371804","DOIUrl":"https://doi.org/10.1145/3336191.3371804","url":null,"abstract":"Entity alignment to find equivalent entities in cross-lingual Knowledge Graphs (KGs) plays a vital role in automatically integrating multiple KGs. Existing translation-based entity alignment methods jointly model the cross-lingual knowledge and monolingual knowledge into one unified optimization problem. On the other hand, the Graph Neural Network (GNN) based methods either ignore the node differentiations, or represent relation through entity or triple instances. They all fail to model the meta semantics embedded in relation nor complex relations such as n-to-n and multi-graphs. To tackle these challenges, we propose a novel Meta Relation Aware Entity Alignment (MRAEA) to directly model cross-lingual entity embeddings by attending over the node's incoming and outgoing neighbors and its connected relations' meta semantics. In addition, we also propose a simple and effective bi-directional iterative strategy to add new aligned seeds during training. Our experiments on all three benchmark entity alignment datasets show that our approach consistently outperforms the state-of-the-art methods, exceeding by 15%-58% on Hit@1. Through an extensive ablation study, we validate that the proposed meta relation aware representations, relation aware self-attention and bi-directional iterative strategy of new seed selection all make contributions to significant performance improvement. The code is available at https://github.com/MaoXinn/MRAEA.","PeriodicalId":319008,"journal":{"name":"Proceedings of the 13th International Conference on Web Search and Data Mining","volume":"23 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-01-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122275021","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}
Recommender systems are embracing conversational technologies to obtain user preferences dynamically, and to overcome inherent limitations of their static models. A successful Conversational Recommender System (CRS) requires proper handling of interactions between conversation and recommendation. We argue that three fundamental problems need to be solved: 1) what questions to ask regarding item attributes, 2) when to recommend items, and 3) how to adapt to the users' online feedback. To the best of our knowledge, there lacks a unified framework that addresses these problems. In this work, we fill this missing interaction framework gap by proposing a new CRS framework named Estimation"Action" Reflection, or EAR, which consists of three stages to better converse with users. (1) Estimation, which builds predictive models to estimate user preference on both items and item attributes; (2) Action, which learns a dialogue policy to determine whether to ask attributes or recommend items, based on Estimation stage and conversation history; and (3) Reflection, which updates the recommender model when a user rejects the recommendations made by the Action stage. We present two conversation scenarios on binary and enumerated questions, and conduct extensive experiments on two datasets from Yelp and LastFM, for each scenario, respectively. Our experiments demonstrate significant improvements over the state-of-the-art method CRM [32], corresponding to fewer conversation turns and a higher level of recommendation hits.
{"title":"Estimation-Action-Reflection: Towards Deep Interaction Between Conversational and Recommender Systems","authors":"Wenqiang Lei, Xiangnan He, Yisong Miao, Qingyun Wu, Richang Hong, Min-Yen Kan, Tat-Seng Chua","doi":"10.1145/3336191.3371769","DOIUrl":"https://doi.org/10.1145/3336191.3371769","url":null,"abstract":"Recommender systems are embracing conversational technologies to obtain user preferences dynamically, and to overcome inherent limitations of their static models. A successful Conversational Recommender System (CRS) requires proper handling of interactions between conversation and recommendation. We argue that three fundamental problems need to be solved: 1) what questions to ask regarding item attributes, 2) when to recommend items, and 3) how to adapt to the users' online feedback. To the best of our knowledge, there lacks a unified framework that addresses these problems. In this work, we fill this missing interaction framework gap by proposing a new CRS framework named Estimation\"Action\" Reflection, or EAR, which consists of three stages to better converse with users. (1) Estimation, which builds predictive models to estimate user preference on both items and item attributes; (2) Action, which learns a dialogue policy to determine whether to ask attributes or recommend items, based on Estimation stage and conversation history; and (3) Reflection, which updates the recommender model when a user rejects the recommendations made by the Action stage. We present two conversation scenarios on binary and enumerated questions, and conduct extensive experiments on two datasets from Yelp and LastFM, for each scenario, respectively. Our experiments demonstrate significant improvements over the state-of-the-art method CRM [32], corresponding to fewer conversation turns and a higher level of recommendation hits.","PeriodicalId":319008,"journal":{"name":"Proceedings of the 13th International Conference on Web Search and Data Mining","volume":"17 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-01-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123673216","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}
Hierarchical user profiling that aims to model users' real-time interests in different granularity is an essential issue for personalized recommendations in E-commerce. On one hand, items (i.e. products) are usually organized hierarchically in categories, and correspondingly users' interests are naturally hierarchical on different granularity of items and categories. On the other hand, multiple granularity oriented recommendations become very popular in E-commerce sites, which require hierarchical user profiling in different granularity as well. In this paper, we propose HUP, a Hierarchical User Profiling framework to solve the hierarchical user profiling problem in E-commerce recommender systems. In HUP, we provide a Pyramid Recurrent Neural Networks, equipped with Behavior-LSTM to formulate users' hierarchical real-time interests at multiple scales. Furthermore, instead of simply utilizing users' item-level behaviors (e.g., ratings or clicks) in conventional methods, HUP harvests the sequential information of users' temporal finely-granular interactions (micro-behaviors, e.g., clicks on components of items like pictures or comments, browses with navigation of the search engines or recommendations) for modeling. Extensive experiments on two real-world E-commerce datasets demonstrate the significant performance gains of the HUP against state-of-the-art methods for the hierarchical user profiling and recommendation problems. We release the codes and datasets at https://github.com/guyulongcs/WSDM2020_HUP.
{"title":"Hierarchical User Profiling for E-commerce Recommender Systems","authors":"Yulong Gu, Zhuoye Ding, Shuaiqiang Wang, Dawei Yin","doi":"10.1145/3336191.3371827","DOIUrl":"https://doi.org/10.1145/3336191.3371827","url":null,"abstract":"Hierarchical user profiling that aims to model users' real-time interests in different granularity is an essential issue for personalized recommendations in E-commerce. On one hand, items (i.e. products) are usually organized hierarchically in categories, and correspondingly users' interests are naturally hierarchical on different granularity of items and categories. On the other hand, multiple granularity oriented recommendations become very popular in E-commerce sites, which require hierarchical user profiling in different granularity as well. In this paper, we propose HUP, a Hierarchical User Profiling framework to solve the hierarchical user profiling problem in E-commerce recommender systems. In HUP, we provide a Pyramid Recurrent Neural Networks, equipped with Behavior-LSTM to formulate users' hierarchical real-time interests at multiple scales. Furthermore, instead of simply utilizing users' item-level behaviors (e.g., ratings or clicks) in conventional methods, HUP harvests the sequential information of users' temporal finely-granular interactions (micro-behaviors, e.g., clicks on components of items like pictures or comments, browses with navigation of the search engines or recommendations) for modeling. Extensive experiments on two real-world E-commerce datasets demonstrate the significant performance gains of the HUP against state-of-the-art methods for the hierarchical user profiling and recommendation problems. We release the codes and datasets at https://github.com/guyulongcs/WSDM2020_HUP.","PeriodicalId":319008,"journal":{"name":"Proceedings of the 13th International Conference on Web Search and Data Mining","volume":"8 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-01-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126574228","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}
Armineh Nourbakhsh, M. Ghassemi, Steven Pomerville
In this paper, we present SPread, an automated financial metric extraction and spreading tool from earnings reports. The tool is created in a document-agnostic fashion, and uses an interpolation of tagging methods to capture arbitrarily complicated expressions. SPread can handle single-line items as well as metrics broken down into sub-items. A validation layer further improves the performance of upstream modules and enables the tool to reach an F1 performance of more than 87% for metrics expressed in tabular format, and 76% for metrics in free-form text. The results are displayed to end-users in an interactive web interface, which allows them to locate, compare, validate, adjust, and export the values.
{"title":"SPread: Automated Financial Metric Extraction and Spreading Tool from Earnings Reports","authors":"Armineh Nourbakhsh, M. Ghassemi, Steven Pomerville","doi":"10.1145/3336191.3371869","DOIUrl":"https://doi.org/10.1145/3336191.3371869","url":null,"abstract":"In this paper, we present SPread, an automated financial metric extraction and spreading tool from earnings reports. The tool is created in a document-agnostic fashion, and uses an interpolation of tagging methods to capture arbitrarily complicated expressions. SPread can handle single-line items as well as metrics broken down into sub-items. A validation layer further improves the performance of upstream modules and enables the tool to reach an F1 performance of more than 87% for metrics expressed in tabular format, and 76% for metrics in free-form text. The results are displayed to end-users in an interactive web interface, which allows them to locate, compare, validate, adjust, and export the values.","PeriodicalId":319008,"journal":{"name":"Proceedings of the 13th International Conference on Web Search and Data Mining","volume":"40 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-01-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116779073","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}
Oluwaseyi Feyisetan, S. Ghanavati, Patricia Thaine
Privacy-preserving data analysis has become essential in Machine Learning (ML), where access to vast amounts of data can provide large gains the in accuracies of tuned models. A large proportion of user-contributed data comes from natural language e.g., text transcriptions from voice assistants. It is therefore important for curated natural language datasets to preserve the privacy of the users whose data is collected and for the models trained on sensitive data to only retain non-identifying (i.e., generalizable) information. The workshop aims to bring together researchers and practitioners from academia and industry to discuss the challenges and approaches to designing, building, verifying, and testing privacy-preserving systems in the context of Natural Language Processing (NLP).
{"title":"Workshop on Privacy in NLP (PrivateNLP 2020)","authors":"Oluwaseyi Feyisetan, S. Ghanavati, Patricia Thaine","doi":"10.1145/3336191.3371881","DOIUrl":"https://doi.org/10.1145/3336191.3371881","url":null,"abstract":"Privacy-preserving data analysis has become essential in Machine Learning (ML), where access to vast amounts of data can provide large gains the in accuracies of tuned models. A large proportion of user-contributed data comes from natural language e.g., text transcriptions from voice assistants. It is therefore important for curated natural language datasets to preserve the privacy of the users whose data is collected and for the models trained on sensitive data to only retain non-identifying (i.e., generalizable) information. The workshop aims to bring together researchers and practitioners from academia and industry to discuss the challenges and approaches to designing, building, verifying, and testing privacy-preserving systems in the context of Natural Language Processing (NLP).","PeriodicalId":319008,"journal":{"name":"Proceedings of the 13th International Conference on Web Search and Data Mining","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-01-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114697931","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}
Congcong Miao, Ziyan Luo, Fengzhu Zeng, Jilong Wang
Predicting human mobility is an important trajectory mining task for various applications, ranging from smart city planning to personalized recommendation system. While most of previous works adopt GPS tracking data to model human mobility, the recent fast-growing geo-tagged social media (GTSM) data brings new opportunities to this task. However, predicting human mobility on GTSM data is not trivial because of three challenges: 1) extreme data sparsity; 2) high order sequential patterns of human mobility and 3) evolving preference of users for tagging. In this paper, we propose ACN, an attentive convolutional network model for predicting human mobility from sparse and complex GTSM data. In ACN, we firstly design a multi-dimension embedding layer which jointly embeds key features (i.e., spatial, temporal and user features) that govern human mobility. Then, we regard the embedded trajectory as an "image" and learn short-term sequential patterns as local features of the image using convolution filters. Instead of directly using convention filters, we design hybrid dilated and separable convolution filters to effectively capture high order sequential patterns from lengthy trajectory. In addition, we propose an attention mechanism which learns the user long-term preference to augment convolutional network for mobility prediction. We conduct extensive experiments on three publicly available GTSM datasets to evaluate the effectiveness of our model. The results demonstrate that ACN consistently outperforms existing state-of-art mobility prediction approaches on a variety of common evaluation metrics.
{"title":"Predicting Human Mobility via Attentive Convolutional Network","authors":"Congcong Miao, Ziyan Luo, Fengzhu Zeng, Jilong Wang","doi":"10.1145/3336191.3371846","DOIUrl":"https://doi.org/10.1145/3336191.3371846","url":null,"abstract":"Predicting human mobility is an important trajectory mining task for various applications, ranging from smart city planning to personalized recommendation system. While most of previous works adopt GPS tracking data to model human mobility, the recent fast-growing geo-tagged social media (GTSM) data brings new opportunities to this task. However, predicting human mobility on GTSM data is not trivial because of three challenges: 1) extreme data sparsity; 2) high order sequential patterns of human mobility and 3) evolving preference of users for tagging. In this paper, we propose ACN, an attentive convolutional network model for predicting human mobility from sparse and complex GTSM data. In ACN, we firstly design a multi-dimension embedding layer which jointly embeds key features (i.e., spatial, temporal and user features) that govern human mobility. Then, we regard the embedded trajectory as an \"image\" and learn short-term sequential patterns as local features of the image using convolution filters. Instead of directly using convention filters, we design hybrid dilated and separable convolution filters to effectively capture high order sequential patterns from lengthy trajectory. In addition, we propose an attention mechanism which learns the user long-term preference to augment convolutional network for mobility prediction. We conduct extensive experiments on three publicly available GTSM datasets to evaluate the effectiveness of our model. The results demonstrate that ACN consistently outperforms existing state-of-art mobility prediction approaches on a variety of common evaluation metrics.","PeriodicalId":319008,"journal":{"name":"Proceedings of the 13th International Conference on Web Search and Data Mining","volume":"65 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-01-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125273952","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：