Detecting Hotspot Information Using Multi-Attribute Based Topic Model.
Microblogging as a kind of social network has become more and more important in our daily lives. Enormous amounts of information are produced and shared on a daily basis. Detecting hot topics in the mountains of information can help people get to the essential information more quickly. However, due...
Main Authors: | , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
Public Library of Science (PLoS)
2015-01-01
|
Series: | PLoS ONE |
Online Access: | http://europepmc.org/articles/PMC4619720?pdf=render |
id |
doaj-405ee9a19ce74543b51748a72e9876c9 |
---|---|
record_format |
Article |
spelling |
doaj-405ee9a19ce74543b51748a72e9876c92020-11-25T01:52:38ZengPublic Library of Science (PLoS)PLoS ONE1932-62032015-01-011010e014053910.1371/journal.pone.0140539Detecting Hotspot Information Using Multi-Attribute Based Topic Model.Jing WangLi LiFeng TanYing ZhuWeisi FengMicroblogging as a kind of social network has become more and more important in our daily lives. Enormous amounts of information are produced and shared on a daily basis. Detecting hot topics in the mountains of information can help people get to the essential information more quickly. However, due to short and sparse features, a large number of meaningless tweets and other characteristics of microblogs, traditional topic detection methods are often ineffective in detecting hot topics. In this paper, we propose a new topic model named multi-attribute latent dirichlet allocation (MA-LDA), in which the time and hashtag attributes of microblogs are incorporated into LDA model. By introducing time attribute, MA-LDA model can decide whether a word should appear in hot topics or not. Meanwhile, compared with the traditional LDA model, applying hashtag attribute in MA-LDA model gives the core words an artificially high ranking in results meaning the expressiveness of outcomes can be improved. Empirical evaluations on real data sets demonstrate that our method is able to detect hot topics more accurately and efficiently compared with several baselines. Our method provides strong evidence of the importance of the temporal factor in extracting hot topics.http://europepmc.org/articles/PMC4619720?pdf=render |
collection |
DOAJ |
language |
English |
format |
Article |
sources |
DOAJ |
author |
Jing Wang Li Li Feng Tan Ying Zhu Weisi Feng |
spellingShingle |
Jing Wang Li Li Feng Tan Ying Zhu Weisi Feng Detecting Hotspot Information Using Multi-Attribute Based Topic Model. PLoS ONE |
author_facet |
Jing Wang Li Li Feng Tan Ying Zhu Weisi Feng |
author_sort |
Jing Wang |
title |
Detecting Hotspot Information Using Multi-Attribute Based Topic Model. |
title_short |
Detecting Hotspot Information Using Multi-Attribute Based Topic Model. |
title_full |
Detecting Hotspot Information Using Multi-Attribute Based Topic Model. |
title_fullStr |
Detecting Hotspot Information Using Multi-Attribute Based Topic Model. |
title_full_unstemmed |
Detecting Hotspot Information Using Multi-Attribute Based Topic Model. |
title_sort |
detecting hotspot information using multi-attribute based topic model. |
publisher |
Public Library of Science (PLoS) |
series |
PLoS ONE |
issn |
1932-6203 |
publishDate |
2015-01-01 |
description |
Microblogging as a kind of social network has become more and more important in our daily lives. Enormous amounts of information are produced and shared on a daily basis. Detecting hot topics in the mountains of information can help people get to the essential information more quickly. However, due to short and sparse features, a large number of meaningless tweets and other characteristics of microblogs, traditional topic detection methods are often ineffective in detecting hot topics. In this paper, we propose a new topic model named multi-attribute latent dirichlet allocation (MA-LDA), in which the time and hashtag attributes of microblogs are incorporated into LDA model. By introducing time attribute, MA-LDA model can decide whether a word should appear in hot topics or not. Meanwhile, compared with the traditional LDA model, applying hashtag attribute in MA-LDA model gives the core words an artificially high ranking in results meaning the expressiveness of outcomes can be improved. Empirical evaluations on real data sets demonstrate that our method is able to detect hot topics more accurately and efficiently compared with several baselines. Our method provides strong evidence of the importance of the temporal factor in extracting hot topics. |
url |
http://europepmc.org/articles/PMC4619720?pdf=render |
work_keys_str_mv |
AT jingwang detectinghotspotinformationusingmultiattributebasedtopicmodel AT lili detectinghotspotinformationusingmultiattributebasedtopicmodel AT fengtan detectinghotspotinformationusingmultiattributebasedtopicmodel AT yingzhu detectinghotspotinformationusingmultiattributebasedtopicmodel AT weisifeng detectinghotspotinformationusingmultiattributebasedtopicmodel |
_version_ |
1724994049662779392 |