The Importance of Prior Probabilities for Entry Page Search

W. Kraaij; T.H.W. Westerveld; Djoerd Hiemstra

doi:10.1145/564376.564383

The Importance of Prior Probabilities for Entry Page Search

W. Kraaij, T.H.W. Westerveld, Djoerd Hiemstra

Databases (Former)

Research output: Contribution to conference › Paper › peer-review

13 Downloads (Pure)

Abstract

An important class of searches on the world-wide-web has the goal to find an entry page (homepage) of an organisation. Entry page search is quite different from Ad Hoc search. Indeed a plain Ad Hoc system performs disappointingly. We explored three non-content features of web pages: page length, number of incoming links and URL form. Especially the URL form proved to be a good predictor. Using URL form priors we found over 70% of all entry pages at rank 1, and up to 89% in the top 10. Non-content features can easily be embedded in a language model framework as a prior probability.

Original language	Undefined
Pages	27-34
Number of pages	8
DOIs	https://doi.org/10.1145/564376.564383
Publication status	Published - Aug 2002
Event	25th Annual International ACM/SIGIR Conference on Research and Development in Information Retrieval, SIGIR 2002 - Tampere, Finland Duration: 11 Aug 2002 → 15 Aug 2002 Conference number: 25 http://sigir.org/sigir2002/

Conference

Conference	25th Annual International ACM/SIGIR Conference on Research and Development in Information Retrieval, SIGIR 2002
Abbreviated title	SIGIR
Country/Territory	Finland
City	Tampere
Period	11/08/02 → 15/08/02
Internet address	http://sigir.org/sigir2002/

Keywords

DB-IR: INFORMATION RETRIEVAL
EWI-7274
IR-63507

Access to Document

10.1145/564376.564383

Cite this

@conference{382a944c2406404db1de527cadab4953,

title = "The Importance of Prior Probabilities for Entry Page Search",

abstract = "An important class of searches on the world-wide-web has the goal to find an entry page (homepage) of an organisation. Entry page search is quite different from Ad Hoc search. Indeed a plain Ad Hoc system performs disappointingly. We explored three non-content features of web pages: page length, number of incoming links and URL form. Especially the URL form proved to be a good predictor. Using URL form priors we found over 70\% of all entry pages at rank 1, and up to 89\% in the top 10. Non-content features can easily be embedded in a language model framework as a prior probability.",

keywords = "DB-IR: INFORMATION RETRIEVAL, EWI-7274, IR-63507",

author = "W. Kraaij and T.H.W. Westerveld and Djoerd Hiemstra",

note = "Imported from EWI/DB PMS [db-utwente:inpr:0000003211]; 25th Annual International ACM/SIGIR Conference on Research and Development in Information Retrieval, SIGIR 2002 ; Conference date: 11-08-2002 Through 15-08-2002",

year = "2002",

month = aug,

doi = "10.1145/564376.564383",

language = "Undefined",

pages = "27--34",

url = "http://sigir.org/sigir2002/",

}

TY - CONF

T1 - The Importance of Prior Probabilities for Entry Page Search

AU - Kraaij, W.

AU - Westerveld, T.H.W.

AU - Hiemstra, Djoerd

N1 - Conference code: 25

PY - 2002/8

Y1 - 2002/8

N2 - An important class of searches on the world-wide-web has the goal to find an entry page (homepage) of an organisation. Entry page search is quite different from Ad Hoc search. Indeed a plain Ad Hoc system performs disappointingly. We explored three non-content features of web pages: page length, number of incoming links and URL form. Especially the URL form proved to be a good predictor. Using URL form priors we found over 70% of all entry pages at rank 1, and up to 89% in the top 10. Non-content features can easily be embedded in a language model framework as a prior probability.

AB - An important class of searches on the world-wide-web has the goal to find an entry page (homepage) of an organisation. Entry page search is quite different from Ad Hoc search. Indeed a plain Ad Hoc system performs disappointingly. We explored three non-content features of web pages: page length, number of incoming links and URL form. Especially the URL form proved to be a good predictor. Using URL form priors we found over 70% of all entry pages at rank 1, and up to 89% in the top 10. Non-content features can easily be embedded in a language model framework as a prior probability.

KW - DB-IR: INFORMATION RETRIEVAL

KW - EWI-7274

KW - IR-63507

U2 - 10.1145/564376.564383

DO - 10.1145/564376.564383

M3 - Paper

SP - 27

EP - 34

T2 - 25th Annual International ACM/SIGIR Conference on Research and Development in Information Retrieval, SIGIR 2002

Y2 - 11 August 2002 through 15 August 2002

ER -