Concept-aware ranking

Teaching an old graph new moves

Colin DeLong, Sandeep Mane, Jaideep Srivastava

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

4 Citations (Scopus)

Abstract

In ranking algorithms for web graphs, such as PageRank and HITS, the lack of attention to concepts/topics representing web page content causes problems such as topic drift and mutually reinforcing relationships between hosts. This paper proposes a novel approach to expand the Web graph to incorporate conceptual information encoded by links (anchor text) between web pages. Using web graph link structure and conceptual information associated with each web page (automatically extracted from anchor text of inlinks), a new graph is defined where each node represents a unique pair of a web page and concept associated with that web page, and an edge represents an explicit or implicit link between two such nodes. This graph captures inter-concept relationships, which is then utilized by ranking algorithms. Our experimental results show that such an approach improves accuracy (e.g., first X precision) by retrieving links which are more authoritative given a user's context.

Original languageEnglish
Title of host publicationProceedings - IEEE International Conference on Data Mining, ICDM
Pages80-85
Number of pages6
Publication statusPublished - 2006
Externally publishedYes
Event6th IEEE International Conference on Data Mining - Workshops, ICDM 2006 - Hong Kong
Duration: 18 Dec 200618 Dec 2006

Other

Other6th IEEE International Conference on Data Mining - Workshops, ICDM 2006
CityHong Kong
Period18/12/0618/12/06

Fingerprint

Websites
Teaching
Anchors
World Wide Web

Keywords

  • Concept-aware ranking
  • Concept-page graph
  • Concepts
  • Implicit links
  • Topic drift

ASJC Scopus subject areas

  • Engineering(all)

Cite this

DeLong, C., Mane, S., & Srivastava, J. (2006). Concept-aware ranking: Teaching an old graph new moves. In Proceedings - IEEE International Conference on Data Mining, ICDM (pp. 80-85). [4063603]

Concept-aware ranking : Teaching an old graph new moves. / DeLong, Colin; Mane, Sandeep; Srivastava, Jaideep.

Proceedings - IEEE International Conference on Data Mining, ICDM. 2006. p. 80-85 4063603.

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

DeLong, C, Mane, S & Srivastava, J 2006, Concept-aware ranking: Teaching an old graph new moves. in Proceedings - IEEE International Conference on Data Mining, ICDM., 4063603, pp. 80-85, 6th IEEE International Conference on Data Mining - Workshops, ICDM 2006, Hong Kong, 18/12/06.
DeLong C, Mane S, Srivastava J. Concept-aware ranking: Teaching an old graph new moves. In Proceedings - IEEE International Conference on Data Mining, ICDM. 2006. p. 80-85. 4063603
DeLong, Colin ; Mane, Sandeep ; Srivastava, Jaideep. / Concept-aware ranking : Teaching an old graph new moves. Proceedings - IEEE International Conference on Data Mining, ICDM. 2006. pp. 80-85
@inproceedings{c34711bee1c345aeaeee1ead627d062f,
title = "Concept-aware ranking: Teaching an old graph new moves",
abstract = "In ranking algorithms for web graphs, such as PageRank and HITS, the lack of attention to concepts/topics representing web page content causes problems such as topic drift and mutually reinforcing relationships between hosts. This paper proposes a novel approach to expand the Web graph to incorporate conceptual information encoded by links (anchor text) between web pages. Using web graph link structure and conceptual information associated with each web page (automatically extracted from anchor text of inlinks), a new graph is defined where each node represents a unique pair of a web page and concept associated with that web page, and an edge represents an explicit or implicit link between two such nodes. This graph captures inter-concept relationships, which is then utilized by ranking algorithms. Our experimental results show that such an approach improves accuracy (e.g., first X precision) by retrieving links which are more authoritative given a user's context.",
keywords = "Concept-aware ranking, Concept-page graph, Concepts, Implicit links, Topic drift",
author = "Colin DeLong and Sandeep Mane and Jaideep Srivastava",
year = "2006",
language = "English",
isbn = "0769527027",
pages = "80--85",
booktitle = "Proceedings - IEEE International Conference on Data Mining, ICDM",

}

TY - GEN

T1 - Concept-aware ranking

T2 - Teaching an old graph new moves

AU - DeLong, Colin

AU - Mane, Sandeep

AU - Srivastava, Jaideep

PY - 2006

Y1 - 2006

N2 - In ranking algorithms for web graphs, such as PageRank and HITS, the lack of attention to concepts/topics representing web page content causes problems such as topic drift and mutually reinforcing relationships between hosts. This paper proposes a novel approach to expand the Web graph to incorporate conceptual information encoded by links (anchor text) between web pages. Using web graph link structure and conceptual information associated with each web page (automatically extracted from anchor text of inlinks), a new graph is defined where each node represents a unique pair of a web page and concept associated with that web page, and an edge represents an explicit or implicit link between two such nodes. This graph captures inter-concept relationships, which is then utilized by ranking algorithms. Our experimental results show that such an approach improves accuracy (e.g., first X precision) by retrieving links which are more authoritative given a user's context.

AB - In ranking algorithms for web graphs, such as PageRank and HITS, the lack of attention to concepts/topics representing web page content causes problems such as topic drift and mutually reinforcing relationships between hosts. This paper proposes a novel approach to expand the Web graph to incorporate conceptual information encoded by links (anchor text) between web pages. Using web graph link structure and conceptual information associated with each web page (automatically extracted from anchor text of inlinks), a new graph is defined where each node represents a unique pair of a web page and concept associated with that web page, and an edge represents an explicit or implicit link between two such nodes. This graph captures inter-concept relationships, which is then utilized by ranking algorithms. Our experimental results show that such an approach improves accuracy (e.g., first X precision) by retrieving links which are more authoritative given a user's context.

KW - Concept-aware ranking

KW - Concept-page graph

KW - Concepts

KW - Implicit links

KW - Topic drift

UR - http://www.scopus.com/inward/record.url?scp=67650547175&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=67650547175&partnerID=8YFLogxK

M3 - Conference contribution

SN - 0769527027

SN - 9780769527024

SP - 80

EP - 85

BT - Proceedings - IEEE International Conference on Data Mining, ICDM

ER -