LTR-expand: query expansion model based on learning to rank association rules
- PDF / 1,313,683 Bytes
- 26 Pages / 439.642 x 666.49 pts Page_size
- 7 Downloads / 198 Views
LTR-expand: query expansion model based on learning to rank association rules Ahlem Bouziri1 · Chiraz Latiri2
· Eric Gaussier3
Received: 11 August 2019 / Revised: 25 November 2019 / Accepted: 10 February 2020 / © Springer Science+Business Media, LLC, part of Springer Nature 2020
Abstract Query Expansion (QE) is widely applied to improve the retrieval performance of ad-hoc search, using different techniques and several data sources to find expansion terms. In Information Retrieval literature, selecting expansion terms remains a challenging task that relies on the extraction of term relationships. In this paper, we propose a new learning to rankbased query expansion model. The main idea behind is that, given a query and the set of its related ARs, our model ranks these ARs according to their relevance score regarding to this query and then selects the most suitable ones to be used in the QE process. Experiments are conducted on three test collections, namely: CLEF2003, TREC-Robust and TREC-Microblog, including long, hard and short queries. Results showed that the retrieval performance can be significantly improved when the ARs ranking method is used compared to other state of the art expansion models, especially for hard and long queries. Keywords Formal Concept Analysis (FCA) · Association Rule (AR) · Generic basis · Learning to rank · SVM rank · Query Expansion (QE)
1 Introduction In a classical Information Retrieval (IR) model, keywords submitted by the user are matched against the collection index to find the documents containing those keywords, which are Ahlem Bouziri
[email protected] Chiraz Latiri [email protected] Eric Gaussier [email protected] 1
ENSI-STICODE, University of Manouba, Tunis, Tunisia
2
LIPAH Research Laboratory, Faculty of Sciences of Tunis, Tunis EL Manar University, Tunis, Tunisia
3
Research Laboratory LIG, AMA Group, University Joseph Fourier-Grenoble, Grenoble, France
Journal of Intelligent Information Systems
then sorted by various methods. Given that user queries are usually ambiguous, this simple retrieval model is often prone to errors and omissions. In order to reduce the term mismatch problem and achieve better retrieval effectiveness, automatic Query Expansion (QE) techniques reformulate the original query by adding new terms that are related to the original query terms, even though they have not been explicitly mentioned by the user. Many Query Expansion (QE) techniques and algorithms have been developed in the last decades. An interesting survey on QE was given in Carpineto and Romano (2012) and Houle et al. (2017). As explained in Carpineto and Romano (2012), the expanded query is the result of four sequential steps, namely: (1) the preprocessing of data sources, (2) the generation and ranking of candidate expansion terms, (3) the selection of expansion terms and (4) query expansion. In this paper, we focus on techniques used to generate candidate terms, which we will simply refer to as terms, according to the strength of their relationship with the origina
Data Loading...
