Topic based language models for ad hoc information retrieval

L. Azzopardi, M. Girolami, C.J. van Rijsbergen

Research output: Chapter in Book/Report/Conference proceedingConference contribution book

34 Citations (Scopus)

Abstract

We propose a topic based approach to language modelling for ad-hoc Information Retrieval (IR). Many smoothed estimators used for the multinomial query model in IR rely upon the estimated background collection probabilities. In this paper, we propose a topic based language modelling approach, that uses a more informative prior based on the topical content of a document. In our experiments, the proposed model provides comparable IR performance to the standard models, but when combined in a two stage language model, it outperforms all other estimated models.

LanguageEnglish
Title of host publication2004 IEEE International Joint Conference on Neural Networks
Place of PublicationPiscataway, NJ
PublisherIEEE
Pages3281-3286
Number of pages6
ISBN (Print)0780383591
DOIs
Publication statusPublished - 17 Jan 2005
Event2004 IEEE International Joint Conference on Neural Networks - Budapest, Hungary
Duration: 25 Jul 200429 Jul 2004

Conference

Conference2004 IEEE International Joint Conference on Neural Networks
CountryHungary
CityBudapest
Period25/07/0429/07/04

Fingerprint

Information retrieval
information retrieval
language
experiment
performance
Experiments

Keywords

  • language modelling
  • Information retrieval
  • two stage language model

Cite this

Azzopardi, L., Girolami, M., & van Rijsbergen, C. J. (2005). Topic based language models for ad hoc information retrieval. In 2004 IEEE International Joint Conference on Neural Networks (pp. 3281-3286). Piscataway, NJ: IEEE. https://doi.org/10.1109/IJCNN.2004.1381205
Azzopardi, L. ; Girolami, M. ; van Rijsbergen, C.J. / Topic based language models for ad hoc information retrieval. 2004 IEEE International Joint Conference on Neural Networks. Piscataway, NJ : IEEE, 2005. pp. 3281-3286
@inproceedings{8b99325a24f74c86977399772d58b34c,
title = "Topic based language models for ad hoc information retrieval",
abstract = "We propose a topic based approach to language modelling for ad-hoc Information Retrieval (IR). Many smoothed estimators used for the multinomial query model in IR rely upon the estimated background collection probabilities. In this paper, we propose a topic based language modelling approach, that uses a more informative prior based on the topical content of a document. In our experiments, the proposed model provides comparable IR performance to the standard models, but when combined in a two stage language model, it outperforms all other estimated models.",
keywords = "language modelling, Information retrieval, two stage language model",
author = "L. Azzopardi and M. Girolami and {van Rijsbergen}, C.J.",
year = "2005",
month = "1",
day = "17",
doi = "10.1109/IJCNN.2004.1381205",
language = "English",
isbn = "0780383591",
pages = "3281--3286",
booktitle = "2004 IEEE International Joint Conference on Neural Networks",
publisher = "IEEE",

}

Azzopardi, L, Girolami, M & van Rijsbergen, CJ 2005, Topic based language models for ad hoc information retrieval. in 2004 IEEE International Joint Conference on Neural Networks. IEEE, Piscataway, NJ, pp. 3281-3286, 2004 IEEE International Joint Conference on Neural Networks, Budapest, Hungary, 25/07/04. https://doi.org/10.1109/IJCNN.2004.1381205

Topic based language models for ad hoc information retrieval. / Azzopardi, L.; Girolami, M.; van Rijsbergen, C.J.

2004 IEEE International Joint Conference on Neural Networks. Piscataway, NJ : IEEE, 2005. p. 3281-3286.

Research output: Chapter in Book/Report/Conference proceedingConference contribution book

TY - GEN

T1 - Topic based language models for ad hoc information retrieval

AU - Azzopardi, L.

AU - Girolami, M.

AU - van Rijsbergen, C.J.

PY - 2005/1/17

Y1 - 2005/1/17

N2 - We propose a topic based approach to language modelling for ad-hoc Information Retrieval (IR). Many smoothed estimators used for the multinomial query model in IR rely upon the estimated background collection probabilities. In this paper, we propose a topic based language modelling approach, that uses a more informative prior based on the topical content of a document. In our experiments, the proposed model provides comparable IR performance to the standard models, but when combined in a two stage language model, it outperforms all other estimated models.

AB - We propose a topic based approach to language modelling for ad-hoc Information Retrieval (IR). Many smoothed estimators used for the multinomial query model in IR rely upon the estimated background collection probabilities. In this paper, we propose a topic based language modelling approach, that uses a more informative prior based on the topical content of a document. In our experiments, the proposed model provides comparable IR performance to the standard models, but when combined in a two stage language model, it outperforms all other estimated models.

KW - language modelling

KW - Information retrieval

KW - two stage language model

UR - http://www.scopus.com/inward/record.url?scp=10944265361&partnerID=8YFLogxK

UR - https://ieeexplore.ieee.org/xpl/mostRecentIssue.jsp?punumber=9486

U2 - 10.1109/IJCNN.2004.1381205

DO - 10.1109/IJCNN.2004.1381205

M3 - Conference contribution book

SN - 0780383591

SP - 3281

EP - 3286

BT - 2004 IEEE International Joint Conference on Neural Networks

PB - IEEE

CY - Piscataway, NJ

ER -

Azzopardi L, Girolami M, van Rijsbergen CJ. Topic based language models for ad hoc information retrieval. In 2004 IEEE International Joint Conference on Neural Networks. Piscataway, NJ: IEEE. 2005. p. 3281-3286 https://doi.org/10.1109/IJCNN.2004.1381205