Query suggestions using Query-Flow graphs

Paolo Boldi, Francesco Bonchi, Carlos Castillo, Debora Donato, Sebastiano Vigna

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

98 Citations (Scopus)

Abstract

The query-low graph [Boldi et al., CIKM 2008] is an aggregated representation of the latent querying behavior contained in a query log. Intuitively, in the query-flow graph a directed edge from query qi to query qj means that the two queries are likely to be part of the same search mission. Any path over the query-low graph may be seen as a possible search task, whose likelihood is given by the strength of the edges along the path. An edge (qi; qj) is also labelled with some information: e.g., the probability that user moves from qi to qj, or the type of the transition, for instance, the fact that qj is a specialization of qi. In this paper we propose, and experimentally study, query recommendations based on short random walks on the query-flow graph. Our experiments show that these methods can match in precision, and often improve, recommendations based on query-click graphs, without using users' clicks. Our experiments also show that it is important to consider transition-type labels on edges for havi good quality recommendations. Finally, one feature that we had in mind while devising our methods was that of providing diverse sets of recommendations: the experimentation that we conducted provides encouraging results in this sense.

Original languageEnglish
Title of host publicationProceedings of Workshop on Web Search Click Data, WSCD'09
Pages56-63
Number of pages8
DOIs
Publication statusPublished - 14 Jul 2009
Externally publishedYes
EventWorkshop on Web Search Click Data, WSCD'09 - Barcelona, Spain
Duration: 9 Feb 20099 Feb 2009

Other

OtherWorkshop on Web Search Click Data, WSCD'09
CountrySpain
CityBarcelona
Period9/2/099/2/09

Fingerprint

Flow graphs
Labels
Experiments

ASJC Scopus subject areas

  • Information Systems
  • Software

Cite this

Boldi, P., Bonchi, F., Castillo, C., Donato, D., & Vigna, S. (2009). Query suggestions using Query-Flow graphs. In Proceedings of Workshop on Web Search Click Data, WSCD'09 (pp. 56-63) https://doi.org/10.1145/1507509.1507518

Query suggestions using Query-Flow graphs. / Boldi, Paolo; Bonchi, Francesco; Castillo, Carlos; Donato, Debora; Vigna, Sebastiano.

Proceedings of Workshop on Web Search Click Data, WSCD'09. 2009. p. 56-63.

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

Boldi, P, Bonchi, F, Castillo, C, Donato, D & Vigna, S 2009, Query suggestions using Query-Flow graphs. in Proceedings of Workshop on Web Search Click Data, WSCD'09. pp. 56-63, Workshop on Web Search Click Data, WSCD'09, Barcelona, Spain, 9/2/09. https://doi.org/10.1145/1507509.1507518
Boldi P, Bonchi F, Castillo C, Donato D, Vigna S. Query suggestions using Query-Flow graphs. In Proceedings of Workshop on Web Search Click Data, WSCD'09. 2009. p. 56-63 https://doi.org/10.1145/1507509.1507518
Boldi, Paolo ; Bonchi, Francesco ; Castillo, Carlos ; Donato, Debora ; Vigna, Sebastiano. / Query suggestions using Query-Flow graphs. Proceedings of Workshop on Web Search Click Data, WSCD'09. 2009. pp. 56-63
@inproceedings{d68584d3292344b9b7749dd70ffe52dc,
title = "Query suggestions using Query-Flow graphs",
abstract = "The query-low graph [Boldi et al., CIKM 2008] is an aggregated representation of the latent querying behavior contained in a query log. Intuitively, in the query-flow graph a directed edge from query qi to query qj means that the two queries are likely to be part of the same search mission. Any path over the query-low graph may be seen as a possible search task, whose likelihood is given by the strength of the edges along the path. An edge (qi; qj) is also labelled with some information: e.g., the probability that user moves from qi to qj, or the type of the transition, for instance, the fact that qj is a specialization of qi. In this paper we propose, and experimentally study, query recommendations based on short random walks on the query-flow graph. Our experiments show that these methods can match in precision, and often improve, recommendations based on query-click graphs, without using users' clicks. Our experiments also show that it is important to consider transition-type labels on edges for havi good quality recommendations. Finally, one feature that we had in mind while devising our methods was that of providing diverse sets of recommendations: the experimentation that we conducted provides encouraging results in this sense.",
author = "Paolo Boldi and Francesco Bonchi and Carlos Castillo and Debora Donato and Sebastiano Vigna",
year = "2009",
month = "7",
day = "14",
doi = "10.1145/1507509.1507518",
language = "English",
isbn = "9781605584348",
pages = "56--63",
booktitle = "Proceedings of Workshop on Web Search Click Data, WSCD'09",

}

TY - GEN

T1 - Query suggestions using Query-Flow graphs

AU - Boldi, Paolo

AU - Bonchi, Francesco

AU - Castillo, Carlos

AU - Donato, Debora

AU - Vigna, Sebastiano

PY - 2009/7/14

Y1 - 2009/7/14

N2 - The query-low graph [Boldi et al., CIKM 2008] is an aggregated representation of the latent querying behavior contained in a query log. Intuitively, in the query-flow graph a directed edge from query qi to query qj means that the two queries are likely to be part of the same search mission. Any path over the query-low graph may be seen as a possible search task, whose likelihood is given by the strength of the edges along the path. An edge (qi; qj) is also labelled with some information: e.g., the probability that user moves from qi to qj, or the type of the transition, for instance, the fact that qj is a specialization of qi. In this paper we propose, and experimentally study, query recommendations based on short random walks on the query-flow graph. Our experiments show that these methods can match in precision, and often improve, recommendations based on query-click graphs, without using users' clicks. Our experiments also show that it is important to consider transition-type labels on edges for havi good quality recommendations. Finally, one feature that we had in mind while devising our methods was that of providing diverse sets of recommendations: the experimentation that we conducted provides encouraging results in this sense.

AB - The query-low graph [Boldi et al., CIKM 2008] is an aggregated representation of the latent querying behavior contained in a query log. Intuitively, in the query-flow graph a directed edge from query qi to query qj means that the two queries are likely to be part of the same search mission. Any path over the query-low graph may be seen as a possible search task, whose likelihood is given by the strength of the edges along the path. An edge (qi; qj) is also labelled with some information: e.g., the probability that user moves from qi to qj, or the type of the transition, for instance, the fact that qj is a specialization of qi. In this paper we propose, and experimentally study, query recommendations based on short random walks on the query-flow graph. Our experiments show that these methods can match in precision, and often improve, recommendations based on query-click graphs, without using users' clicks. Our experiments also show that it is important to consider transition-type labels on edges for havi good quality recommendations. Finally, one feature that we had in mind while devising our methods was that of providing diverse sets of recommendations: the experimentation that we conducted provides encouraging results in this sense.

UR - http://www.scopus.com/inward/record.url?scp=67650093744&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=67650093744&partnerID=8YFLogxK

U2 - 10.1145/1507509.1507518

DO - 10.1145/1507509.1507518

M3 - Conference contribution

SN - 9781605584348

SP - 56

EP - 63

BT - Proceedings of Workshop on Web Search Click Data, WSCD'09

ER -