Xitong Liu

Xitong Liu (劉 熙同)

Email:  me [AT] xtliu.com
Phone:  (302) 533-8128
Office:  111 8th Ave, New York, NY 10011
Web:  xtliu.com

Short Bio

I am a software engineer at Google NYC, working on improving the quality of question answering. I obtained my Ph.D. from Department of Electrical and Computer Engineering at University of Delaware. I am fortunate to have Dr. Hui Fang as my advisor. I received my Bachelor degree from Huazhong University of Science of Technology.

Research

My main research interest is Information Retrieval, particularly on how to leverage entities (e.g., person, location, organization, concept, etc.) to improve search accuracy. My research topics include:

  • Web Search: Entities have been realized to exist extensively across the Web, and the recent advances of entity recognition techniques make it feasible to mine entities from Web-scale data. We utilize entities in Web search queries and documents as bridge to model the relevance in an entity-based latent space in which additional semantic content could be captured, and therefore it could improve the performance of several state-of-the-art retrieval models significantly.
  • Enterprise Search: Enterprise data is organized in both unstructured and structured format, centering around the employees, products and other entities of the enterprise. Our research leverages the integrated information of entities from unstructured and structured data to deliver better search results.
  • Entity Search: Query log analyses have revealed that a substantial portion of Web search queries target at entities. In stead of retrieving documents in traditional search paradigm, our research explores the entity relations in both unstructured and structured data to retrieve relevant entities to fulfill user's information need directly, therefore saving user's efforts.

Part of my research is supported by HP Labs Innovation Research Program.

Refereed Publications

  • Xitong Liu and Hui Fang. Latent Entity Space: A Novel Retrieval Approach for Entity-Bearing Queries. In Information Retrieval, Volume 18, Issue 6, December 2015, pages 473-503. [PDF] [Bibtex] [Link] [Dataset]
  • Xitong Liu, Hui Fang and Deng Cai. Towards Less Biased Web Search. In Proceedings of the 5th ACM International Conference on the Theory of Information Retrieval, pages 373-376, 2015. [PDF] [Bibtex] [Dataset]
  • Yue Wang, Xitong Liu and Hui Fang. A Study of Concept-based Weighting Regularization for Medical Records Search. In Proceedings of the 52nd Annual Meeting of the Association for Computational Linguistics, pages 603–612, 2014. [PDF]
  • Xitong Liu, Fei Chen, Hui Fang and Min Wang. Exploiting Entity Relationship for Query Expansion in Enterprise Search. In Information Retrieval, Volume 17, Issue 3, June 2014, pages 265-294. [PDF] [Bibtex] [Link]
  • Christopher Boston, Hui Fang, Sandra Carberry, Hao Wu and Xitong Liu. Wikimantic: Toward Eective Disambiguation and Expansion of Queries. In Data & Knowledge Engineering, Volume 90, March 2014, pages 22–37. [Link]
  • Xitong Liu, Wei Zheng and Hui Fang. An Exploration of Ranking Models and Feedback Method for Related Entity Finding. In Information Processing & Management, Volume 49, Issue 5, September 2013, pages 995–1007. [Link] [Bibtex]
  • Xitong Liu, Hui Fang, Fei Chen and Min Wang. Entity Centric Query Expansion for Enterprise Search. In Proceedings of the 21st ACM International Conference on Information and Knowledge Management, pages 1955-1959, 2012. [PDF] [Bibtex]
  • Xitong Liu, Hui Fang, Cong-Lei Yao and Min Wang. Finding Relevant Information of Certain Types from Enterprise Data. In Proceedings of the 20th ACM International Conference on Information and Knowledge Management, pages 47-56, 2011. [PDF] [Bibtex]

Workshop and Technical Reports

  • Xitong Liu, Peilin Yang and Hui Fang. Entity Came to Rescue - Leveraging Entities to Minimize Risks in Web Search. In online Proceedings of the 23rd Text REtrieval Conference (TREC'14), Gaithersburg, Maryland, 2014. (Top 1 group in Web Track, oral presentation). [PDF] [Bibtex] [Dataset]
  • Xitong Liu, Peilin Yang and Hui Fang. EntEXPO: An Interactive Search System for Entity-Bearing Queries. In 36th European Conference on Information Retrieval, 2013. (Demo paper) [PDF]
  • Xitong Liu, Jerry Darko and Hui Fang. A Related Entity based Approach for Knowledge Base Acceleration. In online Proceedings of the 22nd Text REtrieval Conference (TREC'13), Gaithersburg, Maryland, 2013. (Top 3 group in KBA Track) [PDF].
  • Xitong Liu and Hui Fang. Leveraging Related Entities for Knowledge Base Acceleration. In 4th International Workshop on Web-scale Knowledge Representation, Retrieval, and Reasoning (Web-KR), 2013. [PDF] [Bibtex]
  • Xitong Liu and Hui Fang. Entity Profile based Approach in Automatic Knowledge Finding. In online Proceedings of the 21st Text REtrieval Conference (TREC'12), Gaithersburg, Maryland, 2012. (Top 2 group in KBA Track, oral presentation) [PDF] [Bibtex].
  • Xitong Liu, Cong-Lei Yao and Hui Fang. A Study of Semantic Search in SemSearch 2011. In 4th International Workshop on Semantic Search, 2011. [PDF]
  • Xitong Liu and Hui Fang. A Study of Entity Search in Semantic Search Workshop. In 3rd International Workshop on Semantic Search, 2010. [PDF]

Honors and Awards

Industry Experience

  • Jun. 2014 - Sep. 2014, Research Intern at Microsoft Research/Bing on entity carousel. [Shipped to production. Example: American computer scientists] Mentor: Dr. Hao Hu.
  • Jun. 2011 - Aug. 2011, Research Intern at HP Labs China on Context-Aware Enterprise Search. Mentor: Dr. Fei Chen.

Patents

  • Performing a search based on entity-related criteria, PCT/US2012/061034, WO 2014062192 A1.

Therefore, since we are surrounded by such a great cloud of witnesses, let us throw off everything that hinders and the sin that so easily entangles, and let us run with perseverance the race marked out for us.

Hebrews 12:1

Last updated: April 2016