Fast Mining of Complex Time-Stamped Events

Hanghang Tong, Yasushi Sakurai, Eliassi Rad Tina, Christos Faloutsos

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

10 Citations (Scopus)

Abstract

Given a collection of complex, time-stamped events, how do we find patterns and anomalies? Events could be meetings with one or more persons with one or more agenda items at zero or more locations (e.g., teleconferences), or they could be publications with authors, keywords, publishers, etc. In such settings, we want to solve the following problems: (1) find time stamps that look similar to each other and group them; (2) find anomalies; (3) provide interpretations of the clusters and anomalies by annotating them; (4) automatically find the right time-granularity in which to do analysis. Moreover, we want fast, scalable algorithms for all these problems. We address the above challenges through two main ideas. The first (T3) is to turn the problem into a graph analysis problem, by carefully treating each time stamp as a node in a graph. This viewpoint brings to bear the vast machinery of graph analysis methods (PageRank, graph partitioning, proximity analysis, and CenterPiece Subgraphs, to name a few). Thus, T3 can automatically group the time stamps into meaningful clusters and spot anomalies. Moreover, it can select representative events/persons/locations for each cluster and each anomaly, as their interpretations. The second idea (MT3) is to use temporal multi-resolution analysis (e.g., minutes, hours, days). We show that MT3 can quickly derive results from finer-to-coarser resolutions, achieving up to 2 orders of magnitude speedups. We verify the effectiveness as well as efficiency of T3 and MT3 on several real datasets.

Original languageEnglish (US)
Title of host publicationInternational Conference on Information and Knowledge Management, Proceedings
Pages759-767
Number of pages9
DOIs
StatePublished - 2008
Externally publishedYes
Event17th ACM Conference on Information and Knowledge Management, CIKM'08 - Napa Valley, CA, United States
Duration: Oct 26 2008Oct 30 2008

Other

Other17th ACM Conference on Information and Knowledge Management, CIKM'08
CountryUnited States
CityNapa Valley, CA
Period10/26/0810/30/08

Fingerprint

Anomaly
Graph
Key words
Multiresolution analysis
Agenda
Node
Proximity
Machinery
PageRank
Partitioning

Keywords

  • Graph mining
  • Multi-resolution analysis
  • Scalability

ASJC Scopus subject areas

  • Business, Management and Accounting(all)
  • Decision Sciences(all)

Cite this

Tong, H., Sakurai, Y., Tina, E. R., & Faloutsos, C. (2008). Fast Mining of Complex Time-Stamped Events. In International Conference on Information and Knowledge Management, Proceedings (pp. 759-767) https://doi.org/10.1145/1458082.1458184

Fast Mining of Complex Time-Stamped Events. / Tong, Hanghang; Sakurai, Yasushi; Tina, Eliassi Rad; Faloutsos, Christos.

International Conference on Information and Knowledge Management, Proceedings. 2008. p. 759-767.

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

Tong, H, Sakurai, Y, Tina, ER & Faloutsos, C 2008, Fast Mining of Complex Time-Stamped Events. in International Conference on Information and Knowledge Management, Proceedings. pp. 759-767, 17th ACM Conference on Information and Knowledge Management, CIKM'08, Napa Valley, CA, United States, 10/26/08. https://doi.org/10.1145/1458082.1458184
Tong H, Sakurai Y, Tina ER, Faloutsos C. Fast Mining of Complex Time-Stamped Events. In International Conference on Information and Knowledge Management, Proceedings. 2008. p. 759-767 https://doi.org/10.1145/1458082.1458184
Tong, Hanghang ; Sakurai, Yasushi ; Tina, Eliassi Rad ; Faloutsos, Christos. / Fast Mining of Complex Time-Stamped Events. International Conference on Information and Knowledge Management, Proceedings. 2008. pp. 759-767
@inproceedings{7818d93f4c2d47a7bd2166b0e8af4228,
title = "Fast Mining of Complex Time-Stamped Events",
abstract = "Given a collection of complex, time-stamped events, how do we find patterns and anomalies? Events could be meetings with one or more persons with one or more agenda items at zero or more locations (e.g., teleconferences), or they could be publications with authors, keywords, publishers, etc. In such settings, we want to solve the following problems: (1) find time stamps that look similar to each other and group them; (2) find anomalies; (3) provide interpretations of the clusters and anomalies by annotating them; (4) automatically find the right time-granularity in which to do analysis. Moreover, we want fast, scalable algorithms for all these problems. We address the above challenges through two main ideas. The first (T3) is to turn the problem into a graph analysis problem, by carefully treating each time stamp as a node in a graph. This viewpoint brings to bear the vast machinery of graph analysis methods (PageRank, graph partitioning, proximity analysis, and CenterPiece Subgraphs, to name a few). Thus, T3 can automatically group the time stamps into meaningful clusters and spot anomalies. Moreover, it can select representative events/persons/locations for each cluster and each anomaly, as their interpretations. The second idea (MT3) is to use temporal multi-resolution analysis (e.g., minutes, hours, days). We show that MT3 can quickly derive results from finer-to-coarser resolutions, achieving up to 2 orders of magnitude speedups. We verify the effectiveness as well as efficiency of T3 and MT3 on several real datasets.",
keywords = "Graph mining, Multi-resolution analysis, Scalability",
author = "Hanghang Tong and Yasushi Sakurai and Tina, {Eliassi Rad} and Christos Faloutsos",
year = "2008",
doi = "10.1145/1458082.1458184",
language = "English (US)",
isbn = "9781595939913",
pages = "759--767",
booktitle = "International Conference on Information and Knowledge Management, Proceedings",

}

TY - GEN

T1 - Fast Mining of Complex Time-Stamped Events

AU - Tong, Hanghang

AU - Sakurai, Yasushi

AU - Tina, Eliassi Rad

AU - Faloutsos, Christos

PY - 2008

Y1 - 2008

N2 - Given a collection of complex, time-stamped events, how do we find patterns and anomalies? Events could be meetings with one or more persons with one or more agenda items at zero or more locations (e.g., teleconferences), or they could be publications with authors, keywords, publishers, etc. In such settings, we want to solve the following problems: (1) find time stamps that look similar to each other and group them; (2) find anomalies; (3) provide interpretations of the clusters and anomalies by annotating them; (4) automatically find the right time-granularity in which to do analysis. Moreover, we want fast, scalable algorithms for all these problems. We address the above challenges through two main ideas. The first (T3) is to turn the problem into a graph analysis problem, by carefully treating each time stamp as a node in a graph. This viewpoint brings to bear the vast machinery of graph analysis methods (PageRank, graph partitioning, proximity analysis, and CenterPiece Subgraphs, to name a few). Thus, T3 can automatically group the time stamps into meaningful clusters and spot anomalies. Moreover, it can select representative events/persons/locations for each cluster and each anomaly, as their interpretations. The second idea (MT3) is to use temporal multi-resolution analysis (e.g., minutes, hours, days). We show that MT3 can quickly derive results from finer-to-coarser resolutions, achieving up to 2 orders of magnitude speedups. We verify the effectiveness as well as efficiency of T3 and MT3 on several real datasets.

AB - Given a collection of complex, time-stamped events, how do we find patterns and anomalies? Events could be meetings with one or more persons with one or more agenda items at zero or more locations (e.g., teleconferences), or they could be publications with authors, keywords, publishers, etc. In such settings, we want to solve the following problems: (1) find time stamps that look similar to each other and group them; (2) find anomalies; (3) provide interpretations of the clusters and anomalies by annotating them; (4) automatically find the right time-granularity in which to do analysis. Moreover, we want fast, scalable algorithms for all these problems. We address the above challenges through two main ideas. The first (T3) is to turn the problem into a graph analysis problem, by carefully treating each time stamp as a node in a graph. This viewpoint brings to bear the vast machinery of graph analysis methods (PageRank, graph partitioning, proximity analysis, and CenterPiece Subgraphs, to name a few). Thus, T3 can automatically group the time stamps into meaningful clusters and spot anomalies. Moreover, it can select representative events/persons/locations for each cluster and each anomaly, as their interpretations. The second idea (MT3) is to use temporal multi-resolution analysis (e.g., minutes, hours, days). We show that MT3 can quickly derive results from finer-to-coarser resolutions, achieving up to 2 orders of magnitude speedups. We verify the effectiveness as well as efficiency of T3 and MT3 on several real datasets.

KW - Graph mining

KW - Multi-resolution analysis

KW - Scalability

UR - http://www.scopus.com/inward/record.url?scp=70349239367&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=70349239367&partnerID=8YFLogxK

U2 - 10.1145/1458082.1458184

DO - 10.1145/1458082.1458184

M3 - Conference contribution

SN - 9781595939913

SP - 759

EP - 767

BT - International Conference on Information and Knowledge Management, Proceedings

ER -