Document-topic hierarchies from document graphs

Tim Weninger; Yonatan Bisk; Jiawei Han

doi:10.1145/2396761.2396843

Document-topic hierarchies from document graphs

Tim Weninger, Yonatan Bisk, Jiawei Han

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Abstract

Topic taxonomies present a multi-level view of a document collection, where general topics live towards the top of the taxonomy and more specific topics live towards the bottom. Topic taxonomies allow users to quickly drill down into their topic of interest to find documents. We show that hierarchies of documents, where documents live at the inner nodes of the hierarchy-tree can also be inferred by combining document text with inter-document links. We present a Bayesian generative model by which an explicit hierarchy of documents is created. Experiments on three document-graph data sets shows that the generated document hierarchies are able to fit the observed data, and that the levels in the constructed document hierarchy represent practical groupings.

Original language	English (US)
Title of host publication	CIKM 2012 - Proceedings of the 21st ACM International Conference on Information and Knowledge Management
Publisher	Association for Computing Machinery
Pages	635-644
Number of pages	10
ISBN (Print)	9781450311564
DOIs	https://doi.org/10.1145/2396761.2396843
State	Published - Jan 1 2012
Event	21st ACM International Conference on Information and Knowledge Management, CIKM 2012 - Maui, HI, United States Duration: Oct 29 2012 → Nov 2 2012

Publication series

Name	ACM International Conference Proceeding Series

Other

Other	21st ACM International Conference on Information and Knowledge Management, CIKM 2012
Country	United States
City	Maui, HI
Period	10/29/12 → 11/2/12

Keywords

bayesian generative models
hierarchical clustering
model evaluation
topic models

ASJC Scopus subject areas

Software
Human-Computer Interaction
Computer Vision and Pattern Recognition
Computer Networks and Communications

Access to Document

10.1145/2396761.2396843

Fingerprint Dive into the research topics of 'Document-topic hierarchies from document graphs'. Together they form a unique fingerprint.

Cite this

Document-topic hierarchies from document graphs. / Weninger, Tim; Bisk, Yonatan; Han, Jiawei.

CIKM 2012 - Proceedings of the 21st ACM International Conference on Information and Knowledge Management. Association for Computing Machinery, 2012. p. 635-644 (ACM International Conference Proceeding Series).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Weninger, T, Bisk, Y & Han, J 2012, Document-topic hierarchies from document graphs. in CIKM 2012 - Proceedings of the 21st ACM International Conference on Information and Knowledge Management. ACM International Conference Proceeding Series, Association for Computing Machinery, pp. 635-644, 21st ACM International Conference on Information and Knowledge Management, CIKM 2012, Maui, HI, United States, 10/29/12. https://doi.org/10.1145/2396761.2396843

@inproceedings{cf6a4d94ad2d4949837d637487a39d17,

title = "Document-topic hierarchies from document graphs",

abstract = "Topic taxonomies present a multi-level view of a document collection, where general topics live towards the top of the taxonomy and more specific topics live towards the bottom. Topic taxonomies allow users to quickly drill down into their topic of interest to find documents. We show that hierarchies of documents, where documents live at the inner nodes of the hierarchy-tree can also be inferred by combining document text with inter-document links. We present a Bayesian generative model by which an explicit hierarchy of documents is created. Experiments on three document-graph data sets shows that the generated document hierarchies are able to fit the observed data, and that the levels in the constructed document hierarchy represent practical groupings.",

keywords = "bayesian generative models, hierarchical clustering, model evaluation, topic models",

author = "Tim Weninger and Yonatan Bisk and Jiawei Han",

year = "2012",

month = jan,

day = "1",

doi = "10.1145/2396761.2396843",

language = "English (US)",

isbn = "9781450311564",

series = "ACM International Conference Proceeding Series",

publisher = "Association for Computing Machinery",

pages = "635--644",

booktitle = "CIKM 2012 - Proceedings of the 21st ACM International Conference on Information and Knowledge Management",

note = "21st ACM International Conference on Information and Knowledge Management, CIKM 2012 ; Conference date: 29-10-2012 Through 02-11-2012",

}

TY - GEN

T1 - Document-topic hierarchies from document graphs

AU - Weninger, Tim

AU - Bisk, Yonatan

AU - Han, Jiawei

PY - 2012/1/1

Y1 - 2012/1/1

N2 - Topic taxonomies present a multi-level view of a document collection, where general topics live towards the top of the taxonomy and more specific topics live towards the bottom. Topic taxonomies allow users to quickly drill down into their topic of interest to find documents. We show that hierarchies of documents, where documents live at the inner nodes of the hierarchy-tree can also be inferred by combining document text with inter-document links. We present a Bayesian generative model by which an explicit hierarchy of documents is created. Experiments on three document-graph data sets shows that the generated document hierarchies are able to fit the observed data, and that the levels in the constructed document hierarchy represent practical groupings.

AB - Topic taxonomies present a multi-level view of a document collection, where general topics live towards the top of the taxonomy and more specific topics live towards the bottom. Topic taxonomies allow users to quickly drill down into their topic of interest to find documents. We show that hierarchies of documents, where documents live at the inner nodes of the hierarchy-tree can also be inferred by combining document text with inter-document links. We present a Bayesian generative model by which an explicit hierarchy of documents is created. Experiments on three document-graph data sets shows that the generated document hierarchies are able to fit the observed data, and that the levels in the constructed document hierarchy represent practical groupings.

KW - bayesian generative models

KW - hierarchical clustering

KW - model evaluation

KW - topic models

UR - http://www.scopus.com/inward/record.url?scp=84871084846&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84871084846&partnerID=8YFLogxK

U2 - 10.1145/2396761.2396843

DO - 10.1145/2396761.2396843

M3 - Conference contribution

AN - SCOPUS:84871084846

SN - 9781450311564

T3 - ACM International Conference Proceeding Series

SP - 635

EP - 644

BT - CIKM 2012 - Proceedings of the 21st ACM International Conference on Information and Knowledge Management

PB - Association for Computing Machinery

T2 - 21st ACM International Conference on Information and Knowledge Management, CIKM 2012

Y2 - 29 October 2012 through 2 November 2012

ER -