Using english information in non-english web search

Wei Gao, John Blitzer, Ming Zhou

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

9 Citations (Scopus)

Abstract

The leading web search engines have spent a decade building highly specialized ranking functions for English web pages. One of the reasons these ranking functions are effective is that they are designed around features such as PageRank, automatic query and domain taxonomies, and click-through information, etc. Unfortunately, many of these features are absent or altered in other languages. In this work, we show how to exploit these English features for a subset of Chinese queries which we call linguistically non-local (LNL). LNL Chinese queries have a minimally ambiguous English translation which also functions as a good English query. We first show how to identify pairs of Chinese LNL queries and their English counterparts from Chinese and English query logs. Then we show how to effectively exploit these pairs to improve Chinese relevance ranking. Our improved relevance ranker proceeds by (1) translating a query into English, (2) computing a cross-lingual relational graph between the Chinese and English documents, and (3) employing the relational ranking method of Qin et al. [15] to rank the Chinese documents. Our technique gives consistent improvements over a state-of-theart Chinese mono-lingual ranker on web search data from the Microsoft Live China search engine.

Original languageEnglish
Title of host publicationInternational Conference on Information and Knowledge Management, Proceedings
Pages17-24
Number of pages8
DOIs
Publication statusPublished - 1 Dec 2008
Externally publishedYes
Event2nd ACM Workshop on Improving Non English Web Searching, iNEWS'08, Co-located with the 17th ACM Conference on Information and Knowledge Management, CIKM 2008 - Napa Valley, CA, United States
Duration: 30 Oct 200830 Oct 2008

Other

Other2nd ACM Workshop on Improving Non English Web Searching, iNEWS'08, Co-located with the 17th ACM Conference on Information and Knowledge Management, CIKM 2008
CountryUnited States
CityNapa Valley, CA
Period30/10/0830/10/08

Fingerprint

Web search
Query
Search engine
Ranking function
Ranking
Clickthrough
PageRank
Graph
Microsoft
Query logs
China
World Wide Web
Taxonomy
Language

Keywords

  • Cross-lingual similarity metrics
  • Learning-to-rank
  • Non-English web search
  • Query translation

ASJC Scopus subject areas

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

Cite this

Gao, W., Blitzer, J., & Zhou, M. (2008). Using english information in non-english web search. In International Conference on Information and Knowledge Management, Proceedings (pp. 17-24) https://doi.org/10.1145/1460027.1460031

Using english information in non-english web search. / Gao, Wei; Blitzer, John; Zhou, Ming.

International Conference on Information and Knowledge Management, Proceedings. 2008. p. 17-24.

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

Gao, W, Blitzer, J & Zhou, M 2008, Using english information in non-english web search. in International Conference on Information and Knowledge Management, Proceedings. pp. 17-24, 2nd ACM Workshop on Improving Non English Web Searching, iNEWS'08, Co-located with the 17th ACM Conference on Information and Knowledge Management, CIKM 2008, Napa Valley, CA, United States, 30/10/08. https://doi.org/10.1145/1460027.1460031
Gao W, Blitzer J, Zhou M. Using english information in non-english web search. In International Conference on Information and Knowledge Management, Proceedings. 2008. p. 17-24 https://doi.org/10.1145/1460027.1460031
Gao, Wei ; Blitzer, John ; Zhou, Ming. / Using english information in non-english web search. International Conference on Information and Knowledge Management, Proceedings. 2008. pp. 17-24
@inproceedings{4bbd024d1d7446e29c650940ba6c46d4,
title = "Using english information in non-english web search",
abstract = "The leading web search engines have spent a decade building highly specialized ranking functions for English web pages. One of the reasons these ranking functions are effective is that they are designed around features such as PageRank, automatic query and domain taxonomies, and click-through information, etc. Unfortunately, many of these features are absent or altered in other languages. In this work, we show how to exploit these English features for a subset of Chinese queries which we call linguistically non-local (LNL). LNL Chinese queries have a minimally ambiguous English translation which also functions as a good English query. We first show how to identify pairs of Chinese LNL queries and their English counterparts from Chinese and English query logs. Then we show how to effectively exploit these pairs to improve Chinese relevance ranking. Our improved relevance ranker proceeds by (1) translating a query into English, (2) computing a cross-lingual relational graph between the Chinese and English documents, and (3) employing the relational ranking method of Qin et al. [15] to rank the Chinese documents. Our technique gives consistent improvements over a state-of-theart Chinese mono-lingual ranker on web search data from the Microsoft Live China search engine.",
keywords = "Cross-lingual similarity metrics, Learning-to-rank, Non-English web search, Query translation",
author = "Wei Gao and John Blitzer and Ming Zhou",
year = "2008",
month = "12",
day = "1",
doi = "10.1145/1460027.1460031",
language = "English",
isbn = "9781605584164",
pages = "17--24",
booktitle = "International Conference on Information and Knowledge Management, Proceedings",

}

TY - GEN

T1 - Using english information in non-english web search

AU - Gao, Wei

AU - Blitzer, John

AU - Zhou, Ming

PY - 2008/12/1

Y1 - 2008/12/1

N2 - The leading web search engines have spent a decade building highly specialized ranking functions for English web pages. One of the reasons these ranking functions are effective is that they are designed around features such as PageRank, automatic query and domain taxonomies, and click-through information, etc. Unfortunately, many of these features are absent or altered in other languages. In this work, we show how to exploit these English features for a subset of Chinese queries which we call linguistically non-local (LNL). LNL Chinese queries have a minimally ambiguous English translation which also functions as a good English query. We first show how to identify pairs of Chinese LNL queries and their English counterparts from Chinese and English query logs. Then we show how to effectively exploit these pairs to improve Chinese relevance ranking. Our improved relevance ranker proceeds by (1) translating a query into English, (2) computing a cross-lingual relational graph between the Chinese and English documents, and (3) employing the relational ranking method of Qin et al. [15] to rank the Chinese documents. Our technique gives consistent improvements over a state-of-theart Chinese mono-lingual ranker on web search data from the Microsoft Live China search engine.

AB - The leading web search engines have spent a decade building highly specialized ranking functions for English web pages. One of the reasons these ranking functions are effective is that they are designed around features such as PageRank, automatic query and domain taxonomies, and click-through information, etc. Unfortunately, many of these features are absent or altered in other languages. In this work, we show how to exploit these English features for a subset of Chinese queries which we call linguistically non-local (LNL). LNL Chinese queries have a minimally ambiguous English translation which also functions as a good English query. We first show how to identify pairs of Chinese LNL queries and their English counterparts from Chinese and English query logs. Then we show how to effectively exploit these pairs to improve Chinese relevance ranking. Our improved relevance ranker proceeds by (1) translating a query into English, (2) computing a cross-lingual relational graph between the Chinese and English documents, and (3) employing the relational ranking method of Qin et al. [15] to rank the Chinese documents. Our technique gives consistent improvements over a state-of-theart Chinese mono-lingual ranker on web search data from the Microsoft Live China search engine.

KW - Cross-lingual similarity metrics

KW - Learning-to-rank

KW - Non-English web search

KW - Query translation

UR - http://www.scopus.com/inward/record.url?scp=70450280889&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=70450280889&partnerID=8YFLogxK

U2 - 10.1145/1460027.1460031

DO - 10.1145/1460027.1460031

M3 - Conference contribution

SN - 9781605584164

SP - 17

EP - 24

BT - International Conference on Information and Knowledge Management, Proceedings

ER -