Fast Identification of Topic Burst Patterns Based on Temporal Clustering

Zhuoyang Xu, Mizuho Iwaihara

研究成果: Conference contribution

抄録

Temporal text mining is widely used in summarization and tracking of evolutionary topic trends. In online collaborative systems like Wikipedia, edit history of each article is stored as revisions. Topics of articles or categories grow and fade over time and retain evolutionary information in edit history. This paper studies a particular temporal text mining task: quickly finding burst patterns of topics from phrases extracted from edit history of Wikipedia articles. We first extract several candidate phrases from edit history by specific features and build time series with edit frequency. Temporal clustering of burst patterns of phrases reveals bursts of topics. However, distance measure for temporal clustering, such as dynamic time warping (DTW), is often costly. In this paper, we propose segmented DTW which decomposes time series into proper segments and computes DTW distance within segments separately. Our segmented DTW shows reasonable speed up over DTW, while the proposed method can identify interesting evolutionary topic burst patterns effectively. Research so far can be applied in domains like trend tracking, temporal relatedness of phrases and popular topic discovery.

元の言語English
ホスト出版物のタイトルProceedings - 2018 7th International Congress on Advanced Applied Informatics, IIAI-AAI 2018
出版者Institute of Electrical and Electronics Engineers Inc.
ページ548-553
ページ数6
ISBN(電子版)9781538674475
DOI
出版物ステータスPublished - 2019 4 16
イベント7th International Congress on Advanced Applied Informatics, IIAI-AAI 2018 - Yonago, Japan
継続期間: 2018 7 82018 7 13

出版物シリーズ

名前Proceedings - 2018 7th International Congress on Advanced Applied Informatics, IIAI-AAI 2018

Conference

Conference7th International Congress on Advanced Applied Informatics, IIAI-AAI 2018
Japan
Yonago
期間18/7/818/7/13

Fingerprint

Wikipedia
history
Time series
time series
trend
time
Clustering
candidacy
Warping
Evolutionary
Text mining
Summarization
Distance measure
Collaborative systems

ASJC Scopus subject areas

  • Computer Networks and Communications
  • Communication
  • Information Systems
  • Information Systems and Management
  • Education

これを引用

Xu, Z., & Iwaihara, M. (2019). Fast Identification of Topic Burst Patterns Based on Temporal Clustering. : Proceedings - 2018 7th International Congress on Advanced Applied Informatics, IIAI-AAI 2018 (pp. 548-553). [8693094] (Proceedings - 2018 7th International Congress on Advanced Applied Informatics, IIAI-AAI 2018). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/IIAI-AAI.2018.00117

Fast Identification of Topic Burst Patterns Based on Temporal Clustering. / Xu, Zhuoyang; Iwaihara, Mizuho.

Proceedings - 2018 7th International Congress on Advanced Applied Informatics, IIAI-AAI 2018. Institute of Electrical and Electronics Engineers Inc., 2019. p. 548-553 8693094 (Proceedings - 2018 7th International Congress on Advanced Applied Informatics, IIAI-AAI 2018).

研究成果: Conference contribution

Xu, Z & Iwaihara, M 2019, Fast Identification of Topic Burst Patterns Based on Temporal Clustering. : Proceedings - 2018 7th International Congress on Advanced Applied Informatics, IIAI-AAI 2018., 8693094, Proceedings - 2018 7th International Congress on Advanced Applied Informatics, IIAI-AAI 2018, Institute of Electrical and Electronics Engineers Inc., pp. 548-553, 7th International Congress on Advanced Applied Informatics, IIAI-AAI 2018, Yonago, Japan, 18/7/8. https://doi.org/10.1109/IIAI-AAI.2018.00117
Xu Z, Iwaihara M. Fast Identification of Topic Burst Patterns Based on Temporal Clustering. : Proceedings - 2018 7th International Congress on Advanced Applied Informatics, IIAI-AAI 2018. Institute of Electrical and Electronics Engineers Inc. 2019. p. 548-553. 8693094. (Proceedings - 2018 7th International Congress on Advanced Applied Informatics, IIAI-AAI 2018). https://doi.org/10.1109/IIAI-AAI.2018.00117
Xu, Zhuoyang ; Iwaihara, Mizuho. / Fast Identification of Topic Burst Patterns Based on Temporal Clustering. Proceedings - 2018 7th International Congress on Advanced Applied Informatics, IIAI-AAI 2018. Institute of Electrical and Electronics Engineers Inc., 2019. pp. 548-553 (Proceedings - 2018 7th International Congress on Advanced Applied Informatics, IIAI-AAI 2018).
@inproceedings{55c98e4a1b3b444ea34cf236625be178,
title = "Fast Identification of Topic Burst Patterns Based on Temporal Clustering",
abstract = "Temporal text mining is widely used in summarization and tracking of evolutionary topic trends. In online collaborative systems like Wikipedia, edit history of each article is stored as revisions. Topics of articles or categories grow and fade over time and retain evolutionary information in edit history. This paper studies a particular temporal text mining task: quickly finding burst patterns of topics from phrases extracted from edit history of Wikipedia articles. We first extract several candidate phrases from edit history by specific features and build time series with edit frequency. Temporal clustering of burst patterns of phrases reveals bursts of topics. However, distance measure for temporal clustering, such as dynamic time warping (DTW), is often costly. In this paper, we propose segmented DTW which decomposes time series into proper segments and computes DTW distance within segments separately. Our segmented DTW shows reasonable speed up over DTW, while the proposed method can identify interesting evolutionary topic burst patterns effectively. Research so far can be applied in domains like trend tracking, temporal relatedness of phrases and popular topic discovery.",
keywords = "Burst detection, Dynamic time warp, Temporal clustering, Topic evolution",
author = "Zhuoyang Xu and Mizuho Iwaihara",
year = "2019",
month = "4",
day = "16",
doi = "10.1109/IIAI-AAI.2018.00117",
language = "English",
series = "Proceedings - 2018 7th International Congress on Advanced Applied Informatics, IIAI-AAI 2018",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
pages = "548--553",
booktitle = "Proceedings - 2018 7th International Congress on Advanced Applied Informatics, IIAI-AAI 2018",

}

TY - GEN

T1 - Fast Identification of Topic Burst Patterns Based on Temporal Clustering

AU - Xu, Zhuoyang

AU - Iwaihara, Mizuho

PY - 2019/4/16

Y1 - 2019/4/16

N2 - Temporal text mining is widely used in summarization and tracking of evolutionary topic trends. In online collaborative systems like Wikipedia, edit history of each article is stored as revisions. Topics of articles or categories grow and fade over time and retain evolutionary information in edit history. This paper studies a particular temporal text mining task: quickly finding burst patterns of topics from phrases extracted from edit history of Wikipedia articles. We first extract several candidate phrases from edit history by specific features and build time series with edit frequency. Temporal clustering of burst patterns of phrases reveals bursts of topics. However, distance measure for temporal clustering, such as dynamic time warping (DTW), is often costly. In this paper, we propose segmented DTW which decomposes time series into proper segments and computes DTW distance within segments separately. Our segmented DTW shows reasonable speed up over DTW, while the proposed method can identify interesting evolutionary topic burst patterns effectively. Research so far can be applied in domains like trend tracking, temporal relatedness of phrases and popular topic discovery.

AB - Temporal text mining is widely used in summarization and tracking of evolutionary topic trends. In online collaborative systems like Wikipedia, edit history of each article is stored as revisions. Topics of articles or categories grow and fade over time and retain evolutionary information in edit history. This paper studies a particular temporal text mining task: quickly finding burst patterns of topics from phrases extracted from edit history of Wikipedia articles. We first extract several candidate phrases from edit history by specific features and build time series with edit frequency. Temporal clustering of burst patterns of phrases reveals bursts of topics. However, distance measure for temporal clustering, such as dynamic time warping (DTW), is often costly. In this paper, we propose segmented DTW which decomposes time series into proper segments and computes DTW distance within segments separately. Our segmented DTW shows reasonable speed up over DTW, while the proposed method can identify interesting evolutionary topic burst patterns effectively. Research so far can be applied in domains like trend tracking, temporal relatedness of phrases and popular topic discovery.

KW - Burst detection

KW - Dynamic time warp

KW - Temporal clustering

KW - Topic evolution

UR - http://www.scopus.com/inward/record.url?scp=85065197398&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85065197398&partnerID=8YFLogxK

U2 - 10.1109/IIAI-AAI.2018.00117

DO - 10.1109/IIAI-AAI.2018.00117

M3 - Conference contribution

AN - SCOPUS:85065197398

T3 - Proceedings - 2018 7th International Congress on Advanced Applied Informatics, IIAI-AAI 2018

SP - 548

EP - 553

BT - Proceedings - 2018 7th International Congress on Advanced Applied Informatics, IIAI-AAI 2018

PB - Institute of Electrical and Electronics Engineers Inc.

ER -