When people talk, they tend to adopt the behaviors, gestures, and language of their conversational partners. This "accommodation" to one's partners is largely automatic, but the degree to which it occurs is influenced by social factors, such as gender, relative power, and attraction. In settings where such social information is not known, this accommodation can be a useful cue for the missing information. This is especially important in web-based communication, where social dynamics are often fluid and rarely stated explicitly. But connecting accommodation and social dynamics on the web requires accurate quantification of the different amounts of accommodation being made. We focus specifically on accommodation in the form of "linguistic alignment": the amount that one person's word use is influenced by another's. Previous studies have used many measures for linguistic alignment, with no clear standard. In this paper, we lay out a set of desiderata for a linguistic alignment measure, including robustness to sparse and short messages, explicit conditionality, and consistency across linguistic features with different baseline frequencies. We propose a straightforward and flexible model-based framework for calculating linguistic alignment, with a focus on the sparse data and limited social information observed in social media. We show that this alignment measure fulfills our desiderata on simulated data. We then analyze a large corpus of Twitter data, both replicating previous results and extending them: Our measure's improved resolution reveals a previously undetectable effect of interpersonal power in Twitter interactions.
{"title":"A Robust Framework for Estimating Linguistic Alignment in Twitter Conversations","authors":"Gabriel Doyle, D. Yurovsky, Michael C. Frank","doi":"10.1145/2872427.2883091","DOIUrl":"https://doi.org/10.1145/2872427.2883091","url":null,"abstract":"When people talk, they tend to adopt the behaviors, gestures, and language of their conversational partners. This \"accommodation\" to one's partners is largely automatic, but the degree to which it occurs is influenced by social factors, such as gender, relative power, and attraction. In settings where such social information is not known, this accommodation can be a useful cue for the missing information. This is especially important in web-based communication, where social dynamics are often fluid and rarely stated explicitly. But connecting accommodation and social dynamics on the web requires accurate quantification of the different amounts of accommodation being made. We focus specifically on accommodation in the form of \"linguistic alignment\": the amount that one person's word use is influenced by another's. Previous studies have used many measures for linguistic alignment, with no clear standard. In this paper, we lay out a set of desiderata for a linguistic alignment measure, including robustness to sparse and short messages, explicit conditionality, and consistency across linguistic features with different baseline frequencies. We propose a straightforward and flexible model-based framework for calculating linguistic alignment, with a focus on the sparse data and limited social information observed in social media. We show that this alignment measure fulfills our desiderata on simulated data. We then analyze a large corpus of Twitter data, both replicating previous results and extending them: Our measure's improved resolution reveals a previously undetectable effect of interpersonal power in Twitter interactions.","PeriodicalId":20455,"journal":{"name":"Proceedings of the 25th International Conference on World Wide Web","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2016-04-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"77379355","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}
Desislava Hristova, Matthew J. Williams, Mirco Musolesi, P. Panzarasa, C. Mascolo
Large metropolitan cities bring together diverse individuals, creating opportunities for cultural and intellectual exchanges, which can ultimately lead to social and economic enrichment. In this work, we present a novel network perspective on the interconnected nature of people and places, allowing us to capture the social diversity of urban locations through the social network and mobility patterns of their visitors. We use a dataset of approximately 37K users and 42K venues in London to build a network of Foursquare places and the parallel Twitter social network of visitors through check-ins. We define four metrics of the social diversity of places which relate to their social brokerage role, their entropy, the homogeneity of their visitors and the amount of serendipitous encounters they are able to induce. This allows us to distinguish between places that bring together strangers versus those which tend to bring together friends, as well as places that attract diverse individuals as opposed to those which attract regulars. We correlate these properties with wellbeing indicators for London neighbourhoods and discover signals of gentrification in deprived areas with high entropy and brokerage, where an influx of more affluent and diverse visitors points to an overall improvement of their rank according to the UK Index of Multiple Deprivation for the area over the five-year census period. Our analysis sheds light on the relationship between the prosperity of people and places, distinguishing between different categories and urban geographies of consequence to the development of urban policy and the next generation of socially-aware location-based applications.
{"title":"Measuring Urban Social Diversity Using Interconnected Geo-Social Networks","authors":"Desislava Hristova, Matthew J. Williams, Mirco Musolesi, P. Panzarasa, C. Mascolo","doi":"10.1145/2872427.2883065","DOIUrl":"https://doi.org/10.1145/2872427.2883065","url":null,"abstract":"Large metropolitan cities bring together diverse individuals, creating opportunities for cultural and intellectual exchanges, which can ultimately lead to social and economic enrichment. In this work, we present a novel network perspective on the interconnected nature of people and places, allowing us to capture the social diversity of urban locations through the social network and mobility patterns of their visitors. We use a dataset of approximately 37K users and 42K venues in London to build a network of Foursquare places and the parallel Twitter social network of visitors through check-ins. We define four metrics of the social diversity of places which relate to their social brokerage role, their entropy, the homogeneity of their visitors and the amount of serendipitous encounters they are able to induce. This allows us to distinguish between places that bring together strangers versus those which tend to bring together friends, as well as places that attract diverse individuals as opposed to those which attract regulars. We correlate these properties with wellbeing indicators for London neighbourhoods and discover signals of gentrification in deprived areas with high entropy and brokerage, where an influx of more affluent and diverse visitors points to an overall improvement of their rank according to the UK Index of Multiple Deprivation for the area over the five-year census period. Our analysis sheds light on the relationship between the prosperity of people and places, distinguishing between different categories and urban geographies of consequence to the development of urban policy and the next generation of socially-aware location-based applications.","PeriodicalId":20455,"journal":{"name":"Proceedings of the 25th International Conference on World Wide Web","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2016-04-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"80891445","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}
Mahanth K. Gowda, Ashutosh Dhekne, Romit Roy Choudhury
This paper explores the possibility of injecting mobility into wireless network infrastructure. We envision WiFi access points on wheels that move to optimize user performance. Movements need not be all around the floor, neither do they have to operate on batteries. As a first step, WiFi APs at home could remain tethered to power and Ethernet outlets while moving in small areas (perhaps under the couch). If such systems prove successful, perhaps future buildings and cities could offer explicit support for network infrastructure mobility. This paper begins with a higher level discussion of robotic wireless networks -- the opportunities and the hurdles -- and then pivots by developing a smaller slice of the vision through a system called iMob. With iMob, a WiFi AP is mounted on a Roomba robot and made to periodically move within a 2x2 sqft region. The core research questions pertain to finding the best location to move to, such that the SNRs from its clients are strong, and the interferences from other APs are weak. Our measurements show that the richness of wireless multipath offers significant opportunities -- even within a 2x2 sqft region, locations exist that are 1.7x better than the average location in terms of throughput. When multiple APs in a neighborhood coordinate, the gains can be even higher. In sum, although infrastructure mobility has been discussed in the context of Google Balloons, ad hoc networks, and delay tolerant networks, we believe that the possibility of moving our personal devices in homes and offices is relatively unexplored, and could open doors to new kinds of innovation.
{"title":"The Case for Robotic Wireless Networks","authors":"Mahanth K. Gowda, Ashutosh Dhekne, Romit Roy Choudhury","doi":"10.1145/2872427.2882986","DOIUrl":"https://doi.org/10.1145/2872427.2882986","url":null,"abstract":"This paper explores the possibility of injecting mobility into wireless network infrastructure. We envision WiFi access points on wheels that move to optimize user performance. Movements need not be all around the floor, neither do they have to operate on batteries. As a first step, WiFi APs at home could remain tethered to power and Ethernet outlets while moving in small areas (perhaps under the couch). If such systems prove successful, perhaps future buildings and cities could offer explicit support for network infrastructure mobility. This paper begins with a higher level discussion of robotic wireless networks -- the opportunities and the hurdles -- and then pivots by developing a smaller slice of the vision through a system called iMob. With iMob, a WiFi AP is mounted on a Roomba robot and made to periodically move within a 2x2 sqft region. The core research questions pertain to finding the best location to move to, such that the SNRs from its clients are strong, and the interferences from other APs are weak. Our measurements show that the richness of wireless multipath offers significant opportunities -- even within a 2x2 sqft region, locations exist that are 1.7x better than the average location in terms of throughput. When multiple APs in a neighborhood coordinate, the gains can be even higher. In sum, although infrastructure mobility has been discussed in the context of Google Balloons, ad hoc networks, and delay tolerant networks, we believe that the possibility of moving our personal devices in homes and offices is relatively unexplored, and could open doors to new kinds of innovation.","PeriodicalId":20455,"journal":{"name":"Proceedings of the 25th International Conference on World Wide Web","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2016-04-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"79251615","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}
Isabel M. Kloumann, Chenhao Tan, J. Kleinberg, Lillian Lee
Despite the existence of highly successful Internet collaborations on complex projects, including open-source software, little is known about how Internet collaborations work for solving "extremely" difficult problems, such as open-ended research questions. We quantitatively investigate a series of efforts known as the Polymath projects, which tackle mathematical research problems through open online discussion. A key analytical insight is that we can contrast the polymath projects with mini-polymaths -- spinoffs that were conducted in the same manner as the polymaths but aimed at addressing math Olympiad questions, which, while quite difficult, are known to be feasible. Our comparative analysis shifts between three elements of the projects: the roles and relationships of the authors, the temporal dynamics of how the projects evolved, and the linguistic properties of the discussions themselves. We find interesting differences between the two domains through each of these analyses, and present these analyses as a template to facilitate comparison between Polymath and other domains for collaboration and communication. We also develop models that have strong performance in distinguishing research-level comments based on any of our groups of features. Finally, we examine whether comments representing research breakthroughs can be recognized more effectively based on their intrinsic features, or by the (re-)actions of others, and find good predictive power in linguistic features.
{"title":"Internet Collaboration on Extremely Difficult Problems: Research versus Olympiad Questions on the Polymath Site","authors":"Isabel M. Kloumann, Chenhao Tan, J. Kleinberg, Lillian Lee","doi":"10.1145/2872427.2883023","DOIUrl":"https://doi.org/10.1145/2872427.2883023","url":null,"abstract":"Despite the existence of highly successful Internet collaborations on complex projects, including open-source software, little is known about how Internet collaborations work for solving \"extremely\" difficult problems, such as open-ended research questions. We quantitatively investigate a series of efforts known as the Polymath projects, which tackle mathematical research problems through open online discussion. A key analytical insight is that we can contrast the polymath projects with mini-polymaths -- spinoffs that were conducted in the same manner as the polymaths but aimed at addressing math Olympiad questions, which, while quite difficult, are known to be feasible. Our comparative analysis shifts between three elements of the projects: the roles and relationships of the authors, the temporal dynamics of how the projects evolved, and the linguistic properties of the discussions themselves. We find interesting differences between the two domains through each of these analyses, and present these analyses as a template to facilitate comparison between Polymath and other domains for collaboration and communication. We also develop models that have strong performance in distinguishing research-level comments based on any of our groups of features. Finally, we examine whether comments representing research breakthroughs can be recognized more effectively based on their intrinsic features, or by the (re-)actions of others, and find good predictive power in linguistic features.","PeriodicalId":20455,"journal":{"name":"Proceedings of the 25th International Conference on World Wide Web","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2016-04-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"79359643","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}
Ontology & schema matching predictors assess the quality of matchers in the absence of an exact match. We propose MCD (Match Competitor Deviation), a new diversity-based predictor that compares the strength of a matcher confidence in the correspondence of a concept pair with respect to other correspondences that involve either concept. We also propose to use MCD as a regulator to optimally control a balance between Precision and Recall and use it towards 1:1 matching by combining it with a similarity measure that is based on solving a maximum weight bipartite graph matching (MWBM). Optimizing the combined measure is known to be an NP-Hard problem. Therefore, we propose CEM, an approximation to an optimal match by efficiently scanning multiple possible matches, using rare event estimation. Using a thorough empirical study over several benchmark real-world datasets, we show that MCD outperforms other state-of-the-art predictor and that CEM significantly outperform existing matchers.
{"title":"From Diversity-based Prediction to Better Ontology & Schema Matching","authors":"A. Gal, Haggai Roitman, Tomer Sagi","doi":"10.1145/2872427.2882999","DOIUrl":"https://doi.org/10.1145/2872427.2882999","url":null,"abstract":"Ontology & schema matching predictors assess the quality of matchers in the absence of an exact match. We propose MCD (Match Competitor Deviation), a new diversity-based predictor that compares the strength of a matcher confidence in the correspondence of a concept pair with respect to other correspondences that involve either concept. We also propose to use MCD as a regulator to optimally control a balance between Precision and Recall and use it towards 1:1 matching by combining it with a similarity measure that is based on solving a maximum weight bipartite graph matching (MWBM). Optimizing the combined measure is known to be an NP-Hard problem. Therefore, we propose CEM, an approximation to an optimal match by efficiently scanning multiple possible matches, using rare event estimation. Using a thorough empirical study over several benchmark real-world datasets, we show that MCD outperforms other state-of-the-art predictor and that CEM significantly outperform existing matchers.","PeriodicalId":20455,"journal":{"name":"Proceedings of the 25th International Conference on World Wide Web","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2016-04-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"89204154","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}
Alexey Borisov, I. Markov, M. de Rijke, P. Serdyukov
Understanding user browsing behavior in web search is key to improving web search effectiveness. Many click models have been proposed to explain or predict user clicks on search engine results. They are based on the probabilistic graphical model (PGM) framework, in which user behavior is represented as a sequence of observable and hidden events. The PGM framework provides a mathematically solid way to reason about a set of events given some information about other events. But the structure of the dependencies between the events has to be set manually. Different click models use different hand-crafted sets of dependencies. We propose an alternative based on the idea of distributed representations: to represent the user's information need and the information available to the user with a vector state. The components of the vector state are learned to represent concepts that are useful for modeling user behavior. And user behavior is modeled as a sequence of vector states associated with a query session: the vector state is initialized with a query, and then iteratively updated based on information about interactions with the search engine results. This approach allows us to directly understand user browsing behavior from click-through data, i.e., without the need for a predefined set of rules as is customary for PGM-based click models. We illustrate our approach using a set of neural click models. Our experimental results show that the neural click model that uses the same training data as traditional PGM-based click models, has better performance on the click prediction task (i.e., predicting user click on search engine results) and the relevance prediction task (i.e., ranking documents by their relevance to a query). An analysis of the best performing neural click model shows that it learns similar concepts to those used in traditional click models, and that it also learns other concepts that cannot be designed manually.
{"title":"A Neural Click Model for Web Search","authors":"Alexey Borisov, I. Markov, M. de Rijke, P. Serdyukov","doi":"10.1145/2872427.2883033","DOIUrl":"https://doi.org/10.1145/2872427.2883033","url":null,"abstract":"Understanding user browsing behavior in web search is key to improving web search effectiveness. Many click models have been proposed to explain or predict user clicks on search engine results. They are based on the probabilistic graphical model (PGM) framework, in which user behavior is represented as a sequence of observable and hidden events. The PGM framework provides a mathematically solid way to reason about a set of events given some information about other events. But the structure of the dependencies between the events has to be set manually. Different click models use different hand-crafted sets of dependencies. We propose an alternative based on the idea of distributed representations: to represent the user's information need and the information available to the user with a vector state. The components of the vector state are learned to represent concepts that are useful for modeling user behavior. And user behavior is modeled as a sequence of vector states associated with a query session: the vector state is initialized with a query, and then iteratively updated based on information about interactions with the search engine results. This approach allows us to directly understand user browsing behavior from click-through data, i.e., without the need for a predefined set of rules as is customary for PGM-based click models. We illustrate our approach using a set of neural click models. Our experimental results show that the neural click model that uses the same training data as traditional PGM-based click models, has better performance on the click prediction task (i.e., predicting user click on search engine results) and the relevance prediction task (i.e., ranking documents by their relevance to a query). An analysis of the best performing neural click model shows that it learns similar concepts to those used in traditional click models, and that it also learns other concepts that cannot be designed manually.","PeriodicalId":20455,"journal":{"name":"Proceedings of the 25th International Conference on World Wide Web","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2016-04-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"74980255","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}
Web search has been a reactive scenario for decades which often starts by users issuing queries. By studying the user behavior in search engine logs, we have discovered that many of the search tasks such as stock-price checking, news reading exhibit strong repeated patterns from day to day. In addition, users exhibit even stronger repetition on mobile devices. This provides us chances to perform proactive recommendations without user issuing queries. In this work, we aim at discovering and characterizing these types of tasks so that we can automatically predict when and what types of tasks will be repeated by the users in the future, through analyzing search logs from a commercial Web search engine and user interaction logs from a mobile App that offers proactive recommendations. We first introduce a set of novel features that can accurately capture task repetition. We then propose a novel deep learning framework that learns user preferences and makes automatic predictions. Our framework is capable of learning both user-independent global models as well as catering personalized models via model adaptation. The model we developed significantly outperforms other state-of-the-art predictive models by large margins. We also demonstrate the power of our model and features through an application to improve the recommendation quality of the mobile App. Results indicate a significant relevance improvement over the current production system.
{"title":"Query-Less: Predicting Task Repetition for NextGen Proactive Search and Recommendation Engines","authors":"Yang Song, Qi Guo","doi":"10.1145/2872427.2883020","DOIUrl":"https://doi.org/10.1145/2872427.2883020","url":null,"abstract":"Web search has been a reactive scenario for decades which often starts by users issuing queries. By studying the user behavior in search engine logs, we have discovered that many of the search tasks such as stock-price checking, news reading exhibit strong repeated patterns from day to day. In addition, users exhibit even stronger repetition on mobile devices. This provides us chances to perform proactive recommendations without user issuing queries. In this work, we aim at discovering and characterizing these types of tasks so that we can automatically predict when and what types of tasks will be repeated by the users in the future, through analyzing search logs from a commercial Web search engine and user interaction logs from a mobile App that offers proactive recommendations. We first introduce a set of novel features that can accurately capture task repetition. We then propose a novel deep learning framework that learns user preferences and makes automatic predictions. Our framework is capable of learning both user-independent global models as well as catering personalized models via model adaptation. The model we developed significantly outperforms other state-of-the-art predictive models by large margins. We also demonstrate the power of our model and features through an application to improve the recommendation quality of the mobile App. Results indicate a significant relevance improvement over the current production system.","PeriodicalId":20455,"journal":{"name":"Proceedings of the 25th International Conference on World Wide Web","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2016-04-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"75867564","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 typically leverage two types of signals to effectively recommend items to users: user activities and content matching between user and item profiles, and recommendation models in literature are usually categorized into collaborative filtering models, content-based models and hybrid models. In practice, when rich profiles about users and items are available, and user activities are sparse (cold-start), effective content matching signals become much more important in the relevance of the recommendation. The de-facto method to measure similarity between two pieces of text is computing the cosine similarity of the two bags of words, and each word is weighted by TF (term frequency within the document) x IDF (inverted document frequency of the word within the corpus). In general sense, TF can represent any local weighting scheme of the word within each document, and IDF can represent any global weighting scheme of the word across the corpus. In this paper, we focus on the latter, i.e., optimizing the global term weights, for a particular recommendation domain by leveraging supervised approaches. The intuition is that some frequent words (lower IDF, e.g. ``database'') can be essential and predictive for relevant recommendation, while some rare words (higher IDF, e.g. the name of a small company) could have less predictive power. Given plenty of observed activities between users and items as training data, we should be able to learn better domain-specific global term weights, which can further improve the relevance of recommendation. We propose a unified method that can simultaneously learn the weights of multiple content matching signals, as well as global term weights for specific recommendation tasks. Our method is efficient to handle large-scale training data generated by production recommender systems. And experiments on LinkedIn job recommendation data justify the effectiveness of our approach.
{"title":"Learning Global Term Weights for Content-based Recommender Systems","authors":"Yupeng Gu, Bo Zhao, D. Hardtke, Yizhou Sun","doi":"10.1145/2872427.2883069","DOIUrl":"https://doi.org/10.1145/2872427.2883069","url":null,"abstract":"Recommender systems typically leverage two types of signals to effectively recommend items to users: user activities and content matching between user and item profiles, and recommendation models in literature are usually categorized into collaborative filtering models, content-based models and hybrid models. In practice, when rich profiles about users and items are available, and user activities are sparse (cold-start), effective content matching signals become much more important in the relevance of the recommendation. The de-facto method to measure similarity between two pieces of text is computing the cosine similarity of the two bags of words, and each word is weighted by TF (term frequency within the document) x IDF (inverted document frequency of the word within the corpus). In general sense, TF can represent any local weighting scheme of the word within each document, and IDF can represent any global weighting scheme of the word across the corpus. In this paper, we focus on the latter, i.e., optimizing the global term weights, for a particular recommendation domain by leveraging supervised approaches. The intuition is that some frequent words (lower IDF, e.g. ``database'') can be essential and predictive for relevant recommendation, while some rare words (higher IDF, e.g. the name of a small company) could have less predictive power. Given plenty of observed activities between users and items as training data, we should be able to learn better domain-specific global term weights, which can further improve the relevance of recommendation. We propose a unified method that can simultaneously learn the weights of multiple content matching signals, as well as global term weights for specific recommendation tasks. Our method is efficient to handle large-scale training data generated by production recommender systems. And experiments on LinkedIn job recommendation data justify the effectiveness of our approach.","PeriodicalId":20455,"journal":{"name":"Proceedings of the 25th International Conference on World Wide Web","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2016-04-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"73176231","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}
Liangda Li, Hongbo Deng, Yunlong He, Anlei Dong, Yi Chang, H. Zha
Search tasks in users' query sequences are dynamic and interconnected. The formulation of search tasks can be influenced by multiple latent factors such as user characteristics, product features and search interactions, which makes search task identification a challenging problem. In this paper, we propose an unsupervised approach to identify search tasks via topic membership along with topic transition probabilities, thus it becomes possible to interpret how user's search intent emerges and evolves over time. Moreover, a novel hidden semi-Markov model is introduced to model topic transitions by considering not only the semantic information of queries but also the latent search factors originated from user search behaviors. A variational inference algorithm is developed to identify remarkable search behavior patterns, typical topic transition tracks, and the topic membership of each query from query logs. The learned topic transition tracks and the inferred topic memberships enable us to identify both small search tasks, where a user searches the same topic, and big search tasks, where a user searches a series of related topics. We extensively evaluate the proposed approach and compare with several state-of-the-art search task identification methods on both synthetic and real-world query log data, and experimental results illustrate the effectiveness of our proposed model.
{"title":"Behavior Driven Topic Transition for Search Task Identification","authors":"Liangda Li, Hongbo Deng, Yunlong He, Anlei Dong, Yi Chang, H. Zha","doi":"10.1145/2872427.2883047","DOIUrl":"https://doi.org/10.1145/2872427.2883047","url":null,"abstract":"Search tasks in users' query sequences are dynamic and interconnected. The formulation of search tasks can be influenced by multiple latent factors such as user characteristics, product features and search interactions, which makes search task identification a challenging problem. In this paper, we propose an unsupervised approach to identify search tasks via topic membership along with topic transition probabilities, thus it becomes possible to interpret how user's search intent emerges and evolves over time. Moreover, a novel hidden semi-Markov model is introduced to model topic transitions by considering not only the semantic information of queries but also the latent search factors originated from user search behaviors. A variational inference algorithm is developed to identify remarkable search behavior patterns, typical topic transition tracks, and the topic membership of each query from query logs. The learned topic transition tracks and the inferred topic memberships enable us to identify both small search tasks, where a user searches the same topic, and big search tasks, where a user searches a series of related topics. We extensively evaluate the proposed approach and compare with several state-of-the-art search task identification methods on both synthetic and real-world query log data, and experimental results illustrate the effectiveness of our proposed model.","PeriodicalId":20455,"journal":{"name":"Proceedings of the 25th International Conference on World Wide Web","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2016-04-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"76394780","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 web search, latent semantic models have been proposed to bridge the lexical gap between queries and documents that is due to the fact that searchers and content creators often use different vocabularies and language styles to express the same concept. Modern search engines simply use the outputs of latent semantic models as features for a so-called global ranker. We argue that this is not optimal, because a single value output by a latent semantic model may be insufficient to describe all aspects of the model's prediction, and thus some information captured by the model is not used effectively by the search engine. To increase the effectiveness of latent semantic models in web search, we propose to create metafeatures-feature vectors that describe the structure of the model's prediction for a given query-document pair and pass them to the global ranker along with the models? scores. We provide simple guidelines to represent the latent semantic model's prediction with more than a single number, and illustrate these guidelines using several latent semantic models. We test the impact of the proposed metafeatures on a web document ranking task using four latent semantic models. Our experiments show that (1) through the use of metafeatures, the performance of each individual latent semantic model can be improved by 10.2% and 4.2% in NDCG scores at truncation levels 1 and 10; and (2) through the use of metafeatures, the performance of a combination of latent semantic models can be improved by 7.6% and 3.8% in NDCG scores at truncation levels 1 and 10, respectively.
{"title":"Using Metafeatures to Increase the Effectiveness of Latent Semantic Models in Web Search","authors":"Alexey Borisov, P. Serdyukov, M. de Rijke","doi":"10.1145/2872427.2882987","DOIUrl":"https://doi.org/10.1145/2872427.2882987","url":null,"abstract":"In web search, latent semantic models have been proposed to bridge the lexical gap between queries and documents that is due to the fact that searchers and content creators often use different vocabularies and language styles to express the same concept. Modern search engines simply use the outputs of latent semantic models as features for a so-called global ranker. We argue that this is not optimal, because a single value output by a latent semantic model may be insufficient to describe all aspects of the model's prediction, and thus some information captured by the model is not used effectively by the search engine. To increase the effectiveness of latent semantic models in web search, we propose to create metafeatures-feature vectors that describe the structure of the model's prediction for a given query-document pair and pass them to the global ranker along with the models? scores. We provide simple guidelines to represent the latent semantic model's prediction with more than a single number, and illustrate these guidelines using several latent semantic models. We test the impact of the proposed metafeatures on a web document ranking task using four latent semantic models. Our experiments show that (1) through the use of metafeatures, the performance of each individual latent semantic model can be improved by 10.2% and 4.2% in NDCG scores at truncation levels 1 and 10; and (2) through the use of metafeatures, the performance of a combination of latent semantic models can be improved by 7.6% and 3.8% in NDCG scores at truncation levels 1 and 10, respectively.","PeriodicalId":20455,"journal":{"name":"Proceedings of the 25th International Conference on World Wide Web","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2016-04-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"74295775","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：