ListBM: a learning-to-rank method for XML keyword search

Authors:
Ning Gao;Zhi-Hong Deng;Yong-Qing Xiang;Yu Hang
Affiliations:
Key Laboratory of Machine Perception, Ministry of Education, School of Electronic Engineering and Computer Science, Peking University;Key Laboratory of Machine Perception, Ministry of Education, School of Electronic Engineering and Computer Science, Peking University;Key Laboratory of Machine Perception, Ministry of Education, School of Electronic Engineering and Computer Science, Peking University;Key Laboratory of Machine Perception, Ministry of Education, School of Electronic Engineering and Computer Science, Peking University
Venue:
INEX'09 Proceedings of the Focused retrieval and evaluation, and 8th international conference on Initiative for the evaluation of XML retrieval
Year:
2009

Citing 6
Cited 0

Searching XML documents via XML fragments

Proceedings of the 26th annual international ACM SIGIR conference on Research and development in informaion retrieval
An efficient boosting algorithm for combining preferences

The Journal of Machine Learning Research
An efficient and versatile query engine for TopX search

VLDB '05 Proceedings of the 31st international conference on Very large data bases
A support vector method for optimizing average precision

SIGIR '07 Proceedings of the 30th annual international ACM SIGIR conference on Research and development in information retrieval
Listwise approach to learning to rank: theory and algorithm

Proceedings of the 25th international conference on Machine learning
University of Waterloo at INEX 2008: Adhoc, Book, and Link-the-Wiki Tracks

Advances in Focused Retrieval

Quantified Score

Hi-index	0.00

Visualization

Abstract

This paper describes Peking University's approach to the Ad Hoc Track. In our first participation, results for all four tasks were submitted: the Best In Context, the Focused, the Relevance In Context and the Thorough. Based on retrieval method Okapi BM25, we implement two different ranking methods Normal BM25 and Learning BM25 according to different parameter settings. Specially, the parameters used in Learning BM25 are learnt by a new learning method called List BM. The evaluation result shows that Learning BM25 is able to beat Normal BM25 in most tasks.