Jianling Wang, Kaize Ding, Ziwei Zhu, Yin Zhang, James Caverlee
Recommendation systems typically rely on the interactions between a crowd of ordinary users and items, ignoring the fact that many real-world communities are notably influenced by a small group of key opinion leaders, whose feedback on items wields outsize influence. With important positions in the community (e.g. have a large number of followers), their elite opinions are able to diffuse to the community and further impact what items we buy, what media we consume, and how we interact with online platforms. Hence, this paper investigates how to develop a novel recommendation system by explicitly capturing the influence from key opinion leaders to the whole community. Centering around opinion elicitation and diffusion, we propose an end-to-end Graph-based neural model - GoRec. Specifically, to preserve the multi-relations between key opinion leaders and items, GoRec elicits the opinions from key opinion leaders with a translation-based embedding method. Moreover, GoRec adopts the idea of Graph Neural Networks to model the elite opinion diffusion process for improved recommendation. Through experiments on Goodreads and Epinions, the proposed model outperforms state-of-the-art approaches by 10.75% and 9.28% on average in Top-K item recommendation.
{"title":"Key Opinion Leaders in Recommendation Systems: Opinion Elicitation and Diffusion","authors":"Jianling Wang, Kaize Ding, Ziwei Zhu, Yin Zhang, James Caverlee","doi":"10.1145/3336191.3371826","DOIUrl":"https://doi.org/10.1145/3336191.3371826","url":null,"abstract":"Recommendation systems typically rely on the interactions between a crowd of ordinary users and items, ignoring the fact that many real-world communities are notably influenced by a small group of key opinion leaders, whose feedback on items wields outsize influence. With important positions in the community (e.g. have a large number of followers), their elite opinions are able to diffuse to the community and further impact what items we buy, what media we consume, and how we interact with online platforms. Hence, this paper investigates how to develop a novel recommendation system by explicitly capturing the influence from key opinion leaders to the whole community. Centering around opinion elicitation and diffusion, we propose an end-to-end Graph-based neural model - GoRec. Specifically, to preserve the multi-relations between key opinion leaders and items, GoRec elicits the opinions from key opinion leaders with a translation-based embedding method. Moreover, GoRec adopts the idea of Graph Neural Networks to model the elite opinion diffusion process for improved recommendation. Through experiments on Goodreads and Epinions, the proposed model outperforms state-of-the-art approaches by 10.75% and 9.28% on average in Top-K item recommendation.","PeriodicalId":319008,"journal":{"name":"Proceedings of the 13th International Conference on Web Search and Data Mining","volume":"50 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":"122512015","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 tutorial, we present a portion of unique industry experience in efficient data labelling via crowdsourcing shared by both leading researchers and engineers from Yandex. We will make an introduction to data labelling via public crowdsourcing marketplaces and will present key components of efficient label collection. This will be followed by a practice session, where participants will choose one of the real label collection tasks, experiment with selecting settings for the labelling process, and launch their label collection project on Yandex.Toloka, one of the largest crowdsourcing marketplaces. The projects will be run on real crowds within the tutorial session. Finally, participants will receive a feedback about their projects and practical advice to make them more efficient. We expect that our tutorial will address an audience with a wide range of background and interests. We do not require specific prerequisite knowledge or skills. We invite beginners, advanced specialists, and researchers to learn how to efficiently collect labelled data.
{"title":"Practice of Efficient Data Collection via Crowdsourcing: Aggregation, Incremental Relabelling, and Pricing","authors":"Alexey Drutsa, Valentina Fedorova, Dmitry Ustalov, Olga Megorskaya, Evfrosiniya Zerminova, Daria Baidakova","doi":"10.1145/3336191.3371875","DOIUrl":"https://doi.org/10.1145/3336191.3371875","url":null,"abstract":"In this tutorial, we present a portion of unique industry experience in efficient data labelling via crowdsourcing shared by both leading researchers and engineers from Yandex. We will make an introduction to data labelling via public crowdsourcing marketplaces and will present key components of efficient label collection. This will be followed by a practice session, where participants will choose one of the real label collection tasks, experiment with selecting settings for the labelling process, and launch their label collection project on Yandex.Toloka, one of the largest crowdsourcing marketplaces. The projects will be run on real crowds within the tutorial session. Finally, participants will receive a feedback about their projects and practical advice to make them more efficient. We expect that our tutorial will address an audience with a wide range of background and interests. We do not require specific prerequisite knowledge or skills. We invite beginners, advanced specialists, and researchers to learn how to efficiently collect labelled data.","PeriodicalId":319008,"journal":{"name":"Proceedings of the 13th International Conference on Web Search and Data Mining","volume":"56 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":"126668955","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 fashion domain is a magnet for computer vision. New vision problems are emerging in step with the fashion industry's rapid evolution towards an online, social, and personalized business. Style models, trend forecasting, interactive search, and recommendation all require visual understanding with rich detail and subtlety. As a result, research in this area is poised to have great influence on how people shop, how the fashion industry analyzes its enterprise, and how we model the cultural trends revealed by what people wear. In this talk, I will present our work over the last few years developing computer vision methods for fashion. To begin, we explore how to discover styles from Web photos, learning how people assemble their outfits and the latent themes they share. Leveraging such styles, we show how to infer compatibility of new garments, optimize personalized mix-and-match capsule wardrobes, suggest minimal edits to make an outfit more fashionable, and recommend clothing that flatters diverse human body shapes. Next, turning to the world stage, we investigate fashion forecasting and influence. Given photos of fashion products, we learn to forecast what looks and styles will be popular in the future. We further boost those forecasts by modeling the spatio-temporal style influences between 44 major world cities. Throughout, by learning models from unlabeled Web photos, our approaches sidestep subjective manual annotations in favor of direct observations of what people choose to wear.
{"title":"Computer Vision for Fashion: From Individual Recommendations to World-wide Trends","authors":"K. Grauman","doi":"10.1145/3336191.3372192","DOIUrl":"https://doi.org/10.1145/3336191.3372192","url":null,"abstract":"The fashion domain is a magnet for computer vision. New vision problems are emerging in step with the fashion industry's rapid evolution towards an online, social, and personalized business. Style models, trend forecasting, interactive search, and recommendation all require visual understanding with rich detail and subtlety. As a result, research in this area is poised to have great influence on how people shop, how the fashion industry analyzes its enterprise, and how we model the cultural trends revealed by what people wear. In this talk, I will present our work over the last few years developing computer vision methods for fashion. To begin, we explore how to discover styles from Web photos, learning how people assemble their outfits and the latent themes they share. Leveraging such styles, we show how to infer compatibility of new garments, optimize personalized mix-and-match capsule wardrobes, suggest minimal edits to make an outfit more fashionable, and recommend clothing that flatters diverse human body shapes. Next, turning to the world stage, we investigate fashion forecasting and influence. Given photos of fashion products, we learn to forecast what looks and styles will be popular in the future. We further boost those forecasts by modeling the spatio-temporal style influences between 44 major world cities. Throughout, by learning models from unlabeled Web photos, our approaches sidestep subjective manual annotations in favor of direct observations of what people choose to wear.","PeriodicalId":319008,"journal":{"name":"Proceedings of the 13th International Conference on Web Search and Data Mining","volume":"19 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":"128133514","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 importance of the distribution of ratings on recommender systems (RS) is well-recognized. And yet, recommendation approaches based on latent factor models and recently introduced neural variants (e.g., NCF) optimize for the head of these distributions, potentially leading to large estimation errors for tail ratings. These errors in tail ratings that are far from the mean predicted rating fall out of a uni-modal assumption underlying these popular models, as we show in this paper. We propose to improve the estimation of tail ratings by extending traditional single latent representations (e.g., an item is represented by a single latent vector) with new multi-latent representations for better modeling these tail ratings. We show how to incorporate these multi-latent representations in an end-to-end neural prediction model that is designed to better reflect the underlying ratings distributions of items. Through experiments over six datasets, we find the proposed model leads to a significant improvement in RMSE versus a suite of benchmark methods. We also find that the predictions for the most polarized items are improved by more than 15%.
{"title":"Improving the Estimation of Tail Ratings in Recommender System with Multi-Latent Representations","authors":"Xing Zhao, Ziwei Zhu, Yin Zhang, James Caverlee","doi":"10.1145/3336191.3371810","DOIUrl":"https://doi.org/10.1145/3336191.3371810","url":null,"abstract":"The importance of the distribution of ratings on recommender systems (RS) is well-recognized. And yet, recommendation approaches based on latent factor models and recently introduced neural variants (e.g., NCF) optimize for the head of these distributions, potentially leading to large estimation errors for tail ratings. These errors in tail ratings that are far from the mean predicted rating fall out of a uni-modal assumption underlying these popular models, as we show in this paper. We propose to improve the estimation of tail ratings by extending traditional single latent representations (e.g., an item is represented by a single latent vector) with new multi-latent representations for better modeling these tail ratings. We show how to incorporate these multi-latent representations in an end-to-end neural prediction model that is designed to better reflect the underlying ratings distributions of items. Through experiments over six datasets, we find the proposed model leads to a significant improvement in RMSE versus a suite of benchmark methods. We also find that the predictions for the most polarized items are improved by more than 15%.","PeriodicalId":319008,"journal":{"name":"Proceedings of the 13th International Conference on Web Search and Data Mining","volume":"58 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":"121600475","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}
Accelerating Deep Convolutional Neural Networks (CNNs) has recently received ever-increasing research focus. Among various approaches proposed in the literature, filter pruning has been regarded as a promising solution, which is due to its advantage in significant speedup and memory reduction of both network model and intermediate feature maps. Previous works utilized "smaller-norm-less-important" criterion to prune filters with smaller ࡁp-norm values by pruning and retraining alternately. This trends to narrow the model capacity for the following reasons: (1) Violent pruning. Previous works adopt a violent strategy in which all filters are simultaneously pruned, which leaving the room to retain model accuracy limited. (2) Filter degradation. Previous works simply set the pruned filter to 0 and retrained it alterately, which easily led to the loss of learning ability of filters. To solve this problem, we propose a novel filter pruning method, namely Incremental Filter Pruning via Random Walk (IFPRW). IFPRW solves the problem of violent pruning by incremental method and Filter degradation by means of random walk. When applied to two image classification benchmarks, the usefulness and strength of IFPRW is validated. Notably, on CIFAR-10, IFPRW reduces more than 46% FLOPs on ResNet-110 with even 0.28% relative accuracy improvement. Moreover, on ILSVRC-2012, IFPRW reduces more than 54% FLOPs on ResNet-101 with only 0.7% top-5 accurcacy drop. which proving that IFPRW outperforms the state-of-the-art filter pruning methods.
{"title":"Incremental Filter Pruning via Random Walk for Accelerating Deep Convolutional Neural Networks","authors":"Qinghua Li, Cuiping Li, Hong Chen","doi":"10.1145/3336191.3371849","DOIUrl":"https://doi.org/10.1145/3336191.3371849","url":null,"abstract":"Accelerating Deep Convolutional Neural Networks (CNNs) has recently received ever-increasing research focus. Among various approaches proposed in the literature, filter pruning has been regarded as a promising solution, which is due to its advantage in significant speedup and memory reduction of both network model and intermediate feature maps. Previous works utilized \"smaller-norm-less-important\" criterion to prune filters with smaller ࡁp-norm values by pruning and retraining alternately. This trends to narrow the model capacity for the following reasons: (1) Violent pruning. Previous works adopt a violent strategy in which all filters are simultaneously pruned, which leaving the room to retain model accuracy limited. (2) Filter degradation. Previous works simply set the pruned filter to 0 and retrained it alterately, which easily led to the loss of learning ability of filters. To solve this problem, we propose a novel filter pruning method, namely Incremental Filter Pruning via Random Walk (IFPRW). IFPRW solves the problem of violent pruning by incremental method and Filter degradation by means of random walk. When applied to two image classification benchmarks, the usefulness and strength of IFPRW is validated. Notably, on CIFAR-10, IFPRW reduces more than 46% FLOPs on ResNet-110 with even 0.28% relative accuracy improvement. Moreover, on ILSVRC-2012, IFPRW reduces more than 54% FLOPs on ResNet-101 with only 0.7% top-5 accurcacy drop. which proving that IFPRW outperforms the state-of-the-art filter pruning methods.","PeriodicalId":319008,"journal":{"name":"Proceedings of the 13th International Conference on Web Search and Data Mining","volume":"20 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":"125252521","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}
Academic homepages are an important source for learning researchers' profiles. Recognising person names and publications in academic homepages are two fundamental tasks for understanding the identities of the homepages and collaboration networks of the researchers. Existing studies have tackled person name recognition and publication recognition separately. We observe that these two tasks are correlated since person names and publications often co-occur. Further, there are strong position patterns for the occurrence of person names and publications. With these observations, we propose a novel deep learning model consisting of two main modules, an alternatingly updated memory module which exploits the knowledge and correlation from both tasks, and a position-aware memory module which captures the patterns of where in a homepage names and publications appear. Empirical results show that our proposed model outperforms the state-of-the-art publication recognition model by 3.64% in F1 score and outperforms the state-of-the-art person name recognition model by 2.06% in F1 score. Ablation studies and visualisation confirm the effectiveness of the proposed modules.
{"title":"Joint Recognition of Names and Publications in Academic Homepages","authors":"Yimeng Dai, Jianzhong Qi, Rui Zhang","doi":"10.1145/3336191.3371771","DOIUrl":"https://doi.org/10.1145/3336191.3371771","url":null,"abstract":"Academic homepages are an important source for learning researchers' profiles. Recognising person names and publications in academic homepages are two fundamental tasks for understanding the identities of the homepages and collaboration networks of the researchers. Existing studies have tackled person name recognition and publication recognition separately. We observe that these two tasks are correlated since person names and publications often co-occur. Further, there are strong position patterns for the occurrence of person names and publications. With these observations, we propose a novel deep learning model consisting of two main modules, an alternatingly updated memory module which exploits the knowledge and correlation from both tasks, and a position-aware memory module which captures the patterns of where in a homepage names and publications appear. Empirical results show that our proposed model outperforms the state-of-the-art publication recognition model by 3.64% in F1 score and outperforms the state-of-the-art person name recognition model by 2.06% in F1 score. Ablation studies and visualisation confirm the effectiveness of the proposed modules.","PeriodicalId":319008,"journal":{"name":"Proceedings of the 13th International Conference on Web Search and Data Mining","volume":"11 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":"114235774","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}
Network embedding, that aims to learn low-dimensional vector representation of nodes such that the network structure is preserved, has gained significant research attention in recent years. However, most state-of-the-art network embedding methods are computationally expensive and hence unsuitable for representing nodes in billion-scale networks. In this paper, we present LouvainNE, a hierarchical clustering approach to network embedding. Precisely, we employ Louvain, an extremely fast and accurate community detection method, to build a hierarchy of successively smaller subgraphs. We obtain representations of individual nodes in the original graph at different levels of the hierarchy, then we aggregate these representations to learn the final embedding vectors. Our theoretical analysis shows that our proposed algorithm has quasi-linear run-time and memory complexity. Our extensive experimental evaluation, carried out on multiple real-world networks of different scales, demonstrates both (i) the scalability of our proposed approach that can handle graphs containing tens of billions of edges, as well as (ii) its effectiveness in performing downstream network mining tasks such as network reconstruction and node classification.
{"title":"LouvainNE","authors":"Ayan Kumar Bhowmick, Koushik Meneni, Maximilien Danisch, J. Guillaume, Bivas Mitra","doi":"10.1145/3336191.3371800","DOIUrl":"https://doi.org/10.1145/3336191.3371800","url":null,"abstract":"Network embedding, that aims to learn low-dimensional vector representation of nodes such that the network structure is preserved, has gained significant research attention in recent years. However, most state-of-the-art network embedding methods are computationally expensive and hence unsuitable for representing nodes in billion-scale networks. In this paper, we present LouvainNE, a hierarchical clustering approach to network embedding. Precisely, we employ Louvain, an extremely fast and accurate community detection method, to build a hierarchy of successively smaller subgraphs. We obtain representations of individual nodes in the original graph at different levels of the hierarchy, then we aggregate these representations to learn the final embedding vectors. Our theoretical analysis shows that our proposed algorithm has quasi-linear run-time and memory complexity. Our extensive experimental evaluation, carried out on multiple real-world networks of different scales, demonstrates both (i) the scalability of our proposed approach that can handle graphs containing tens of billions of edges, as well as (ii) its effectiveness in performing downstream network mining tasks such as network reconstruction and node classification.","PeriodicalId":319008,"journal":{"name":"Proceedings of the 13th International Conference on Web Search and Data Mining","volume":"19 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":"114259025","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}
Da-Cheng Juan, Chun-Ta Lu, Zhuguo Li, Futang Peng, Aleksei Timofeev, Yi-Ting Chen, Y. Gao, Tom Duerig, A. Tomkins, Sujith Ravi
"How to learn image embeddings that capture fine-grained semantics based on the instance of an image?" "Is it possible for such embeddings to further understand image semantics closer to humans' perception?" In this paper, we present, Graph-Regularized Image Semantic Embedding (Graph-RISE), a web-scale neural graph learning framework deployed at Google, which allows us to train image embeddings to discriminate an unprecedented O(40M) ultra-fine-grained semantic labels. The proposed Graph-RISE outperforms state-of-the-art image embedding algorithms on several evaluation tasks, including kNN search and triplet ranking: the accuracy is improved by approximately 2X on the ImageNet dataset and by more than 5X on the iNaturalist dataset. Qualitatively, image retrieval from one billion images based on the proposed Graph-RISE effectively captures semantics and, compared to the state-of-the-art, differentiates nuances at levels that are closer to human-perception.
{"title":"Ultra Fine-Grained Image Semantic Embedding","authors":"Da-Cheng Juan, Chun-Ta Lu, Zhuguo Li, Futang Peng, Aleksei Timofeev, Yi-Ting Chen, Y. Gao, Tom Duerig, A. Tomkins, Sujith Ravi","doi":"10.1145/3336191.3371784","DOIUrl":"https://doi.org/10.1145/3336191.3371784","url":null,"abstract":"\"How to learn image embeddings that capture fine-grained semantics based on the instance of an image?\" \"Is it possible for such embeddings to further understand image semantics closer to humans' perception?\" In this paper, we present, Graph-Regularized Image Semantic Embedding (Graph-RISE), a web-scale neural graph learning framework deployed at Google, which allows us to train image embeddings to discriminate an unprecedented O(40M) ultra-fine-grained semantic labels. The proposed Graph-RISE outperforms state-of-the-art image embedding algorithms on several evaluation tasks, including kNN search and triplet ranking: the accuracy is improved by approximately 2X on the ImageNet dataset and by more than 5X on the iNaturalist dataset. Qualitatively, image retrieval from one billion images based on the proposed Graph-RISE effectively captures semantics and, compared to the state-of-the-art, differentiates nuances at levels that are closer to human-perception.","PeriodicalId":319008,"journal":{"name":"Proceedings of the 13th International Conference on Web Search and Data Mining","volume":"13 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":"123892651","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}
Lixin Zou, Long Xia, Pan Du, Zhuo Zhang, Ting Bai, Weidong Liu, J. Nie, Dawei Yin
Applying reinforcement learning (RL) in recommender systems is attractive but costly due to the constraint of the interaction with real customers, where performing online policy learning through interacting with real customers usually harms customer experiences. A practical alternative is to build a recommender agent offline from logged data, whereas directly using logged data offline leads to the problem of selection bias between logging policy and the recommendation policy. The existing direct offline learning algorithms, such as Monte Carlo methods and temporal difference methods are either computationally expensive or unstable on convergence. To address these issues, we propose Pseudo Dyna-Q (PDQ). In PDQ, instead of interacting with real customers, we resort to a customer simulator, referred to as the World Model, which is designed to simulate the environment and handle the selection bias of logged data. During policy improvement, the World Model is constantly updated and optimized adaptively, according to the current recommendation policy. This way, the proposed PDQ not only avoids the instability of convergence and high computation cost of existing approaches but also provides unlimited interactions without involving real customers. Moreover, a proved upper bound of empirical error of reward function guarantees that the learned offline policy has lower bias and variance. Extensive experiments demonstrated the advantages of PDQ on two real-world datasets against state-of-the-arts methods.
{"title":"Pseudo Dyna-Q: A Reinforcement Learning Framework for Interactive Recommendation","authors":"Lixin Zou, Long Xia, Pan Du, Zhuo Zhang, Ting Bai, Weidong Liu, J. Nie, Dawei Yin","doi":"10.1145/3336191.3371801","DOIUrl":"https://doi.org/10.1145/3336191.3371801","url":null,"abstract":"Applying reinforcement learning (RL) in recommender systems is attractive but costly due to the constraint of the interaction with real customers, where performing online policy learning through interacting with real customers usually harms customer experiences. A practical alternative is to build a recommender agent offline from logged data, whereas directly using logged data offline leads to the problem of selection bias between logging policy and the recommendation policy. The existing direct offline learning algorithms, such as Monte Carlo methods and temporal difference methods are either computationally expensive or unstable on convergence. To address these issues, we propose Pseudo Dyna-Q (PDQ). In PDQ, instead of interacting with real customers, we resort to a customer simulator, referred to as the World Model, which is designed to simulate the environment and handle the selection bias of logged data. During policy improvement, the World Model is constantly updated and optimized adaptively, according to the current recommendation policy. This way, the proposed PDQ not only avoids the instability of convergence and high computation cost of existing approaches but also provides unlimited interactions without involving real customers. Moreover, a proved upper bound of empirical error of reward function guarantees that the learned offline policy has lower bias and variance. Extensive experiments demonstrated the advantages of PDQ on two real-world datasets against state-of-the-arts methods.","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":"115474817","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}
User intent understanding is a crucial step in designing both conversational agents and search engines. Detecting or inferring user intent is challenging, since the user utterances or queries can be short, ambiguous, and contextually dependent. To address these research challenges, my thesis work focuses on: 1) Utterance topic and intent classification for conversational agents 2) Query intent mining and classification for Web search engines, focusing on the e-commerce domain. To address the first topic, I proposed novel models to incorporate entity information and conversation-context clues to predict both topic and intent of the user's utterances. For the second research topic, I plan to extend the existing state of the art methods in Web search intent prediction to the e-commerce domain, via: 1) Developing a joint learning model to predict search queries' intents and the product categories associated with them, 2) Discovering new hidden users' intents. All the models will be evaluated on the real queries available from a major e-commerce site search engine. The results from these studies can be leveraged to improve performance of various tasks such as natural language understanding, query scoping, query suggestion, and ranking, resulting in an enriched user experience.
{"title":"User Intent Inference for Web Search and Conversational Agents","authors":"Ali Ahmadvand","doi":"10.1145/3336191.3372187","DOIUrl":"https://doi.org/10.1145/3336191.3372187","url":null,"abstract":"User intent understanding is a crucial step in designing both conversational agents and search engines. Detecting or inferring user intent is challenging, since the user utterances or queries can be short, ambiguous, and contextually dependent. To address these research challenges, my thesis work focuses on: 1) Utterance topic and intent classification for conversational agents 2) Query intent mining and classification for Web search engines, focusing on the e-commerce domain. To address the first topic, I proposed novel models to incorporate entity information and conversation-context clues to predict both topic and intent of the user's utterances. For the second research topic, I plan to extend the existing state of the art methods in Web search intent prediction to the e-commerce domain, via: 1) Developing a joint learning model to predict search queries' intents and the product categories associated with them, 2) Discovering new hidden users' intents. All the models will be evaluated on the real queries available from a major e-commerce site search engine. The results from these studies can be leveraged to improve performance of various tasks such as natural language understanding, query scoping, query suggestion, and ranking, resulting in an enriched user experience.","PeriodicalId":319008,"journal":{"name":"Proceedings of the 13th International Conference on Web Search and Data Mining","volume":"102 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":"115594873","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：