Projects

ExplainAble Recommendation and Search (EARS)

    Explainable Recommendation and Search refers to the (personalized) recommendation and search algorithms that not only provide the user with the search and recommendation results, but also let the user know why such results are provided, i.e., they try to address the problem of "why" in recommendation and search systems. Explainable recommendation and search algorithms devise interpretable models and generates intuitive explanations for users. The explainations can be provided in various forms, such as word clouds, sentences, visual images, or diagrams, etc, which help to improve the effectiveness, efficiency, persuasiveness, and user satisfaction of recommendation and search systems.

    Related Publications:
  • Explainable Recommendation: A Survey and New Perspectives. Yongfeng Zhang and Xu Chen. arXiv Preprint 2018. arXiv:1804.11192. [PDF]
  • Yongfeng Zhang, Guokun Lai, Min Zhang, Yi Zhang, Yiqun Liu and Shaoping Ma. Explicit Factor Models for Explainable Recommendation based on Phrase-level Sentiment Analysis. In Proceedings of the 37th Annual International ACM SIGIR Conference on Research and Development on Information Retrieval (SIGIR 2014), July 6 - 11, 2014, Gold Coast, Australia. [PDF]
  • Xu Chen, Yongfeng Zhang, Hongteng Xu, Yixin Cao, Zheng Qin, and Hongyuan Zha. Visually Explainable Recommendation. Preprint. [PDF]
  • Xu Chen, Zheng Qin, Yongfeng Zhang, and Tao Xu. Learning to Rank Features for Recommendation over Multiple Categories. In Proceedings of the 39th Annual International ACM SIGIR Conference on Research and Development on Information Retrieval (SIGIR 2016), July 17 - 21, 2016, Pisa, Italy. [PDF]
  • Xu Chen, Hongteng Xu, Yongfeng Zhang, Yixin Cao, Hongyuan Zha, Zheng Qin and Jiaxi Tang. Sequential Recommendation with User Memory Networks. In Proceedings of the 11th ACM International Conference on Web Search and Data Mining (WSDM 2018), February 5 - 9, 2018, Los Angeles, California, USA. [PDF]
  • Yongfeng Zhang. Explainable Recommendation: Theory and Applications. arXiv Preprint 2017. arXiv:1708.06409. [PDF]
  • Yongfeng Zhang. Incorporating Phrase-level Sentiment Analysis on Textual Reviews for Personalized Recommendation. In Proceedings of the 8th International Conference on Web Search and Data Mining (WSDM 2015), Feb. 2 - 6, 2015, Shanghai, China. [PDF]
  • Yongfeng Zhang. Browser-Oriented Universal Cross-Site Recommendation and Explanation based on User Browsing Logs. In Proceedings of the 8th ACM Conference Series on Recommender Systems (RecSys 2014), Oct. 6 - 10, 2014, Foster City, Silicon Valley, USA. [PDF]
  • Yongfeng Zhang, Min Zhang, Yi Zhang, Guokun Lai, Yiqun Liu, Honghui Zhang, Shaoping Ma. Daily-Aware Personalized Recommendation based on Feature-Level Time Series Analysis. In Proceedings of the 24th International World Wide Web Conference (WWW 2015), May 18 - 22, 2015, Florence, Italy. [PDF]

Economics of Recommendation and Search

    With the continuous shifting of human activities from offline to online, the Web is no longer just a platform for information sharing and transmission, but a huge online economy where various products or services are distributed from producers to consumers. As a result, a fundamentally important role of the Web economy is Online Resource Allocation (ORA) from producers to consumers, such as product allocation in E-commerce, job allocation in freelancing platforms, and driver resource allocation in P2P riding services. Since users have the freedom to choose, such allocations are not provided in a forced manner, but usually in forms of personalized recommendation or search. Economic Recommendation aims at divising recommendation and resource allocation algorithms for the Web economy based on principled economic theories or intuitions, so as to achieve targeted goals in online resource allocation, such as intelligent marketing, welfare distribution between consumers and producers, improving economic efficiency, cost reduction in terms of price, time, location, etc.

    Related Publications:
  • Yongfeng Zhang, Qi Zhao, Yi Zhang, Daniel Friedman, Min Zhang, Yiqun Liu, and Shaoping Ma. Economic Recommendation with Surplus Maximization. In Proceedings of the 25th International World Wide Web Conference (WWW 2016), April 11 - 15, 2016, Montreal, Canada. [PDF]
  • Qi Zhao, Yongfeng Zhang, Yi Zhang, and Daniel Friedman. Multi-Product Utility Maximization for Economic Recommendation. In Proceedings of the 10th International Conference on Web Search and Data Mining (WSDM 2017), February 6 - 10, 2017, Cambridge, UK. [PDF]
  • Yongfeng Zhang, Yi Zhang and Daniel Friedman. Economic Recommendation based on Pareto Efficient Resource Allocation. Science Center Berlin for Social Research Discussion Papers, Wissenschaftszentrum Berlin für Sozialforschung. [PDF]
  • Xiao Lin, Min Zhang, Yongfeng Zhang, Zhaoquan Gu, Yiqun Liu, Shaoping Ma. Fairness-Aware Group Recommendation with Pareto Efficiency. In Proceedings of the 11th ACM Conference on Recommender Systems (RecSys 2017), August 27 - 31, 2017, Como, Italy. [PDF]
  • Xiao Lin, Min Zhang, Yongfeng Zhang, Yiqun Liu, and Shaoping Ma. Boosting Moving Average Reversion Strategy for Online Portfolio Selection: A Meta-Learning Approach. In Proceedings of the 22nd International Conference on Database Systems for Advanced Applications (DASFAA 2017), March 27 - 30, 2017, Suzhou, China. [PDF]

Product Search

    Product search in e-commerce exhibits unique challenges compared with general web search. Because user preferences may be different even on the same type of products, the product search task can be highly personalized in nature. Different from document search, products are described by a hybridization of structured and unstructured data, including product description, image, user ratings/reviews, structured knowledge base, etc. Product search is also a relatively more difficult task because we not only care about user clicks but also user purchase behavoirs, and purchase behaviors cost a lot more than simply clicking a result and are thus more sparse. The product search task is also closely related to the system profits in e-commerce by prodiving satisfactory results that can attract user purchase.

    Related Publications:
  • Qingyao Ai, Yongfeng Zhang, Keping Bi, Xu Chen, and W. Bruce Croft. Learning a Hierarchical Embedding Model for Personalized Product Search. In Proceedings of the 40th International ACM SIGIR Conference on Research and Development in Information Retrieval (SIGIR 2017), August 7 - 11, 2017, Tokyo, Japan. [PDF]

Deep Learning for Recommendation and Search

    Our research on deep learning for recommendationo mostly focus on representation learnining for recommendation. To do so, we incoporate multimodal and heterogeneous information sources for user personalization and item recommendataion, and the tasks include top-N recommendation, sequential recommendation, multimedia recommendation, etc. Our Joint Representation Learning (JRL) model achieved 2~3 times of improvment against traditional shallow models for top-N recommendation (N=10), in terms of NDCG, Precision, Recall, and Hit-Ratio, on several product domains of a standard Amazon dataset.

    Related Publications:
  • Yongfeng Zhang, Qingyao Ai, Xu Chen, and W. Bruce Croft. Joint Representation Learning for Top-N Recommendation with Heterogenous Information Sources. In Proceedings of the 26th ACM International Conference on Information and Knowledge Management (CIKM 2017), November 6 - 10, 2017, Singapore. [PDF]
  • Xu Chen, Hongteng Xu, Yongfeng Zhang, Yixin Cao, Hongyuan Zha, Zheng Qin and Jiaxi Tang. Sequential Recommendation with User Memory Networks. In Proceedings of the 11th ACM International Conference on Web Search and Data Mining (WSDM 2018), February 5 - 9, 2018, Los Angeles, California, USA. [PDF]
  • Xu Chen, Yongfeng Zhang, Qingyao Ai, Hongteng Xu, Junchi Yan, Zheng Qin. Personalized Key Frame Recommendation. In Proceedings of the 40th International ACM SIGIR Conference on Research and Development in Information Retrieval (SIGIR 2017), August 7-11, 2017, Tokyo, Japan. [PDF]
  • Xu Chen, Yongfeng Zhang, Hongteng Xu, Yixin Cao, Zheng Qin, and Hongyuan Zha. Visually Explainable Recommendation. Preprint. [PDF]
  • Xu Chen, Wayne Xin Zhao, Yongfeng Zhang, Zheng Qin, Wenwen Ye. A Collaborative Neural Model for Rating Prediction by Leveraging User Reviews and Product Images. In Proceedings of the 13th Asia Information Retrieval Societies Conference (AIRS 2017), November 22 - 24, 2017, Jeju Island, Korea. Best Paper Award. [PDF]

Phrase-level Sentiment Analysis

    Extracting sentimental elements from textual corpora serves as the basis for a lot of higher-level research tasks such as recommendation, search, document summarization, public opinion analysis, intelligent marketing, etc. By integrating statistical and machine learning approaches, we developed a phrase-level sentiment analysis toolkit that extracts 'feature-opinion-sentiment' triplets to construct a context-sensitive sentiment lexicon from large scale user textual reviews. For example, in the product domain of mobile phone, sampled triplets include 'picture-clear-positive', 'battery life-short-negative', 'quality-high-positive', or 'noise-high-negative'. It further contains a module that is capable of matching the triplets contained in a piece of review, and another module that contructs feature-level profiles for online items according to the user reviews made towards a specific product. For more details about the toolkit and its application, please checkout the "Sentires: Phrase-level Sentiment Analysis toolkit" under the Software of my homepage.

    Related Publications:
  • Yongfeng Zhang, Haochen Zhang, Min Zhang, Yiqun Liu and Shaoping Ma. Do Users Rate or Review? Boost Phrase-level Sentiment Labeling with Review-level Sentiment Classification. In Proceedings of the 37th Annual International ACM SIGIR Conference on Research and Development on Information Retrieval (SIGIR 2014) (short paper), July 6 - 11, 2014, Gold Coast, Australia. [PDF][poster]
  • Yunzhi Tan, Yongfeng Zhang, Min Zhang, Yiqun Liu and Shaoping Ma. A Unified Framework for Emotional Elements Extraction based on Finite State Matching Machine. In Proceedings of the 2nd CCF Conference on Natural Language Processing and Chinese Computing (NLP&CC 2013), Nov. 15 - 19, 2013, Chongqing, China. [PDF]
  • Yongfeng Zhang, Min Zhang, Yiqun Liu and Shaoping Ma. Boost Phrase-level Polarity Labelling with Review-level Sentiment Classification. arXiv Preprint 2015. arXiv:1502.03322. [PDF]
  • Yongfeng Zhang, Guokun Lai, Min Zhang, Yi Zhang, Yiqun Liu and Shaoping Ma. Explicit Factor Models for Explainable Recommendation based on Phrase-level Sentiment Analysis. In Proceedings of the 37th Annual International ACM SIGIR Conference on Research and Development on Information Retrieval (SIGIR 2014), July 6 - 11, 2014, Gold Coast, Australia. [PDF]
  • Xu Chen, Zheng Qin, Yongfeng Zhang, and Tao Xu. Learning to Rank Features for Recommendation over Multiple Categories. In Proceedings of the 39th Annual International ACM SIGIR Conference on Research and Development on Information Retrieval (SIGIR 2016), July 17 - 21, 2016, Pisa, Italy. [PDF]
  • Yongfeng Zhang. Incorporating Phrase-level Sentiment Analysis on Textual Reviews for Personalized Recommendation. The 8th International Conference on Web Search and Data Mining (WSDM 2015), Feb. 2 - 6, 2015, Shanghai, China. [PDF]
  • Yongfeng Zhang, Min Zhang, Yi Zhang, Guokun Lai, Yiqun Liu, Honghui Zhang, Shaoping Ma. Daily-Aware Personalized Recommendation based on Feature-Level Time Series Analysis. In Proceedings of the 24th International World Wide Web Conference (WWW 2015), May 18 - 22, 2015, Florence, Italy. [PDF]