Extending a Hybrid Tag-Based Recommender System with Personalization

Frederico Durao, Peter Dolog

Research output: Contribution to journalConference article in JournalResearchpeer-review

25 Citations (Scopus)

Abstract

Tagging activity has been recently identified as a potential source of knowledge about personal interests, preferences, goals, and other attributes known from user models. Tags themselves can be therefore used for finding personalized recommendations of items. This paper proposes a semantic extension for a hybrid tag-based recommender system, which suggests similar Web pages based on the similarity of their tags. The semantic extension aims at discovering tag relations which are not considered in basic syntax similarity. With the goal of generating more semantically grounded recommendations, the proposal extends a hybrid tag-based recommender system with a semantic factor, which looks for tag relations in different semantic sources. In order to evaluate the benefits acquired with the semantic extension, we have compared the new findings with results from a previous experiment involving 38 people from 12 countries using data from del.icio.us.
Original languageEnglish
JournalACM Symposium on Applied Computing,
Pages (from-to)1723-1727
DOIs
Publication statusPublished - 2010

Fingerprint

Recommender systems
Semantics
Websites
Experiments

Bibliographical note

Symposium on Applied Computing
Proceedings of the 2010 ACM Symposium on Applied Computing table of contents
Sierre, Switzerland
SESSION: Information access and retrieval track

Cite this

@inproceedings{a06dc068694c41a285f108b6cc6ac97b,
title = "Extending a Hybrid Tag-Based Recommender System with Personalization",
abstract = "Tagging activity has been recently identified as a potential source of knowledge about personal interests, preferences, goals, and other attributes known from user models. Tags themselves can be therefore used for finding personalized recommendations of items. This paper proposes a semantic extension for a hybrid tag-based recommender system, which suggests similar Web pages based on the similarity of their tags. The semantic extension aims at discovering tag relations which are not considered in basic syntax similarity. With the goal of generating more semantically grounded recommendations, the proposal extends a hybrid tag-based recommender system with a semantic factor, which looks for tag relations in different semantic sources. In order to evaluate the benefits acquired with the semantic extension, we have compared the new findings with results from a previous experiment involving 38 people from 12 countries using data from del.icio.us.",
author = "Frederico Durao and Peter Dolog",
note = "Symposium on Applied Computing Proceedings of the 2010 ACM Symposium on Applied Computing table of contents Sierre, Switzerland SESSION: Information access and retrieval track",
year = "2010",
doi = "10.1145/1774088.1774457",
language = "English",
pages = "1723--1727",
journal = "ACM Symposium on Applied Computing,",

}

Extending a Hybrid Tag-Based Recommender System with Personalization. / Durao, Frederico; Dolog, Peter.

In: ACM Symposium on Applied Computing, 2010, p. 1723-1727.

Research output: Contribution to journalConference article in JournalResearchpeer-review

TY - GEN

T1 - Extending a Hybrid Tag-Based Recommender System with Personalization

AU - Durao, Frederico

AU - Dolog, Peter

N1 - Symposium on Applied Computing Proceedings of the 2010 ACM Symposium on Applied Computing table of contents Sierre, Switzerland SESSION: Information access and retrieval track

PY - 2010

Y1 - 2010

N2 - Tagging activity has been recently identified as a potential source of knowledge about personal interests, preferences, goals, and other attributes known from user models. Tags themselves can be therefore used for finding personalized recommendations of items. This paper proposes a semantic extension for a hybrid tag-based recommender system, which suggests similar Web pages based on the similarity of their tags. The semantic extension aims at discovering tag relations which are not considered in basic syntax similarity. With the goal of generating more semantically grounded recommendations, the proposal extends a hybrid tag-based recommender system with a semantic factor, which looks for tag relations in different semantic sources. In order to evaluate the benefits acquired with the semantic extension, we have compared the new findings with results from a previous experiment involving 38 people from 12 countries using data from del.icio.us.

AB - Tagging activity has been recently identified as a potential source of knowledge about personal interests, preferences, goals, and other attributes known from user models. Tags themselves can be therefore used for finding personalized recommendations of items. This paper proposes a semantic extension for a hybrid tag-based recommender system, which suggests similar Web pages based on the similarity of their tags. The semantic extension aims at discovering tag relations which are not considered in basic syntax similarity. With the goal of generating more semantically grounded recommendations, the proposal extends a hybrid tag-based recommender system with a semantic factor, which looks for tag relations in different semantic sources. In order to evaluate the benefits acquired with the semantic extension, we have compared the new findings with results from a previous experiment involving 38 people from 12 countries using data from del.icio.us.

U2 - 10.1145/1774088.1774457

DO - 10.1145/1774088.1774457

M3 - Conference article in Journal

SP - 1723

EP - 1727

JO - ACM Symposium on Applied Computing,

JF - ACM Symposium on Applied Computing,

ER -