Header image  
Tutorials: Nov 13, 2011, Shanghai, China
Main Conference: Nov 14-17, 2011, Shanghai, China
Workshops: Nov 18, 2011, Hangzhou, China
line decor
   Final program of ICONIP2011 and book of abstracts are available now.  
line decor

Learning to Rank


Hang Li
Microsoft Research Asia, China


In this tutorial, I will conduct a survey on learning to rank, by explaining the fundamental problems, existing approaches, and future work of learning to rank. The objective of this tutorial is to introduce this new machine learning technology to the scientists, engineers, educators, and students in neural information processing and to seek new opportunities of applying it to neural information processing and other related fields.


Learning to rank refers to machine learning techniques for training the model in a ranking task. Learning to rank is useful for many applications in Information Retrieval, Natural Language Processing, and Data Mining. Intensive studies have been conducted on the problem and significant progress has been made.

Various ranking problems in Information Retrieval and Natural Language Processing can be formalized as two basic ranking tasks, namely ranking creation (or simply ranking) and ranking aggregation. In ranking creation, given a request, one wants to generate a ranking list of offerings based on the features derived from the request and the offerings. In ranking aggregation, given a request, as well as a number of ranking lists of offerings, one wants to generate a new ranking list of the offerings. Ranking creation (or ranking) is the major problem in learning to rank. It is usually formalized as a supervised learning task. Detailed explanations on learning for ranking creation and ranking aggregation will be given in this tutorial, including training and testing, evaluation, feature creation, and major approaches.

Many methods have been proposed for ranking creation. The methods can be categorized as the pointwise, pairwise, and listwise approaches according to the loss functions they employ. They can also be categorized according to the techniques they employ, such as the SVM based, Boosting SVM, Neural Network based approaches. Several popular learning to rank methods will be explained in details in this tutorial. These include OC SVM, Ranking SVM, IR SVM, SVM MAP, PermuRank, and ListNet.

Furthermore, several example applications of learning to rank will be described, including web search, collaborative filtering, and re-ranking in machine translation. Ongoing and future research directions for learning to rank will be introduced. Discussions on potential applications of learning to rank to neural information processing will be made.

Target Audience:

This tutorial is geared toward professionals who are interested in machine learning technologies for ranking. It is assumed that the audience has certain knowledge on statistics and machine learning.


        Hang Li is senior researcher and research manager at Microsoft Research Asia. He is also adjunct professors at Peking University, Nanjing University, Xi’an Jiaotong University, and Nankai University. His research areas include information retrieval, natural language processing, statistical machine learning, and data mining. He graduated from Kyoto University in 1988 and earned his PhD from the University of Tokyo in 1998. He worked at the NEC lab in Japan during 1991 and 2001. He joined Microsoft Research Asia in 2001 and has been working there until present. Hang has about 100 publications at top international journals and conferences, including SIGIR, WWW, WSDM, ACL, EMNLP, ICML, NIPS, and SIGKDD. He and his colleagues' papers received the SIGKDD'08 best application paper award and the SIGIR?8 best student paper award. Hang has also been working on the development of several products. These include Microsoft SQL Server 2005, Microsoft Office 2007 and Office 2010, Microsoft Live Search 2008,Microsoft Bing 2009 and Bing 2010. He has also been very active in the research communities and severed or is serving the top conferences and journals. For example, in 2011, he is PC co-chair of WSDM'11; area chairs of SIGIR'11, AAAI'11, NIPS'11; PC members of WWW'11, ACL-HLT'11, SIGKDD'11, ICDM'11, EMNLP'11; editorial board members of Journal of the American Society for Information Science and Journal of Computer Science & Technology. Hang Li is one of the researchers actively working on learning to rank. A monograph on learning to rank written by him has been published recently.

Copyright© 2010-2011 International Conference on Neural Information Processing. All rights reserved.