Exploring tweets normalization and query time sensitivity for twitter search

Zhongyu Wei, Wei Gao, Lanjun Zhou, Binyang Li, Kam Fai Wong

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

Abstract

This paper presents our work for the Realtime Adhoc Task of TREC 2011 Microblog Track. Microblog texts like tweets are generally characterized by the inclusion of a large proportion of irregular expressions, such as ill- formed words, which can lead to significant mismatch between query terms and tweets. In addition. Twitter queries are distinguished from Web queries with many unique characteristics, one of which reflects the clearly distinct temporal aspects of Twitter search behavior. In this study, we deal with the first problem by normalizing tweet texts and the second by capturing the temporal characteristics of a topic. We divided topics into two categories: time-sensitive and time-insensitive. For the time-sensitive ones, we introduce a decay factor to adjust the relevance score of results according to the expected date of the topical event to happen, and then re-rank the search results. Experiments demonstrate that our methods are significantly better than baseline and outperform the medium of all runs.

Original languageEnglish
Title of host publicationNIST Special Publication
Publication statusPublished - 2011
Event20th Text REtrieval Conference, TREC 2011 - Gaithersburg, MD, United States
Duration: 15 Nov 201118 Nov 2011

Other

Other20th Text REtrieval Conference, TREC 2011
CountryUnited States
CityGaithersburg, MD
Period15/11/1118/11/11

Fingerprint

Experiments

ASJC Scopus subject areas

  • Engineering(all)

Cite this

Wei, Z., Gao, W., Zhou, L., Li, B., & Wong, K. F. (2011). Exploring tweets normalization and query time sensitivity for twitter search. In NIST Special Publication

Exploring tweets normalization and query time sensitivity for twitter search. / Wei, Zhongyu; Gao, Wei; Zhou, Lanjun; Li, Binyang; Wong, Kam Fai.

NIST Special Publication. 2011.

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

Wei, Z, Gao, W, Zhou, L, Li, B & Wong, KF 2011, Exploring tweets normalization and query time sensitivity for twitter search. in NIST Special Publication. 20th Text REtrieval Conference, TREC 2011, Gaithersburg, MD, United States, 15/11/11.
Wei Z, Gao W, Zhou L, Li B, Wong KF. Exploring tweets normalization and query time sensitivity for twitter search. In NIST Special Publication. 2011
Wei, Zhongyu ; Gao, Wei ; Zhou, Lanjun ; Li, Binyang ; Wong, Kam Fai. / Exploring tweets normalization and query time sensitivity for twitter search. NIST Special Publication. 2011.
@inproceedings{2925c2a513274d87b195f8dbd6775da6,
title = "Exploring tweets normalization and query time sensitivity for twitter search",
abstract = "This paper presents our work for the Realtime Adhoc Task of TREC 2011 Microblog Track. Microblog texts like tweets are generally characterized by the inclusion of a large proportion of irregular expressions, such as ill- formed words, which can lead to significant mismatch between query terms and tweets. In addition. Twitter queries are distinguished from Web queries with many unique characteristics, one of which reflects the clearly distinct temporal aspects of Twitter search behavior. In this study, we deal with the first problem by normalizing tweet texts and the second by capturing the temporal characteristics of a topic. We divided topics into two categories: time-sensitive and time-insensitive. For the time-sensitive ones, we introduce a decay factor to adjust the relevance score of results according to the expected date of the topical event to happen, and then re-rank the search results. Experiments demonstrate that our methods are significantly better than baseline and outperform the medium of all runs.",
author = "Zhongyu Wei and Wei Gao and Lanjun Zhou and Binyang Li and Wong, {Kam Fai}",
year = "2011",
language = "English",
booktitle = "NIST Special Publication",

}

TY - GEN

T1 - Exploring tweets normalization and query time sensitivity for twitter search

AU - Wei, Zhongyu

AU - Gao, Wei

AU - Zhou, Lanjun

AU - Li, Binyang

AU - Wong, Kam Fai

PY - 2011

Y1 - 2011

N2 - This paper presents our work for the Realtime Adhoc Task of TREC 2011 Microblog Track. Microblog texts like tweets are generally characterized by the inclusion of a large proportion of irregular expressions, such as ill- formed words, which can lead to significant mismatch between query terms and tweets. In addition. Twitter queries are distinguished from Web queries with many unique characteristics, one of which reflects the clearly distinct temporal aspects of Twitter search behavior. In this study, we deal with the first problem by normalizing tweet texts and the second by capturing the temporal characteristics of a topic. We divided topics into two categories: time-sensitive and time-insensitive. For the time-sensitive ones, we introduce a decay factor to adjust the relevance score of results according to the expected date of the topical event to happen, and then re-rank the search results. Experiments demonstrate that our methods are significantly better than baseline and outperform the medium of all runs.

AB - This paper presents our work for the Realtime Adhoc Task of TREC 2011 Microblog Track. Microblog texts like tweets are generally characterized by the inclusion of a large proportion of irregular expressions, such as ill- formed words, which can lead to significant mismatch between query terms and tweets. In addition. Twitter queries are distinguished from Web queries with many unique characteristics, one of which reflects the clearly distinct temporal aspects of Twitter search behavior. In this study, we deal with the first problem by normalizing tweet texts and the second by capturing the temporal characteristics of a topic. We divided topics into two categories: time-sensitive and time-insensitive. For the time-sensitive ones, we introduce a decay factor to adjust the relevance score of results according to the expected date of the topical event to happen, and then re-rank the search results. Experiments demonstrate that our methods are significantly better than baseline and outperform the medium of all runs.

UR - http://www.scopus.com/inward/record.url?scp=84873563850&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84873563850&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:84873563850

BT - NIST Special Publication

ER -