A survey on learning to rank for information retrieval

PhD Qualifying Examination


Title: "A survey on learning to rank for information retrieval"

by

Mr. Chi-Wai Cheung


Abstract:

Ranking is a central problem in information retrieval. There are many 
conventional ranking models in the literature that are based on 
statistical and probability theory, but they have some drawbacks. For 
example, they require difficult parameter tuning and hence are hard to be 
adopted in different applications. Learning to rank aims at learning the 
ranking function automatically from a training dataset and it has become a 
hot research topic in information retrieval and machine learning. In the 
literature, the learning to rank methods were categorized into three 
categories, namely pointwise, pairwise and listwise methods. In this 
survey, we cover several learning to rank methods in these three 
categories. Some challenges of learning to rank are discussed and some 
possible future research directions are suggested.


Date:                   Tuesday, 17 August 2010

Time:                   2:00pm - 4:00pm

Venue:                  Room 5566
                         lifts 27/28

Committee Members:      Prof. Dik-Lun Lee (Supervisor)
                         Dr. Lei Chen (Chairperson)
                         Dr. Raymond Wong
 			Prof. Dit-Yan Yeung


**** ALL are Welcome ****