Risk-aware recommender systems

Djallel Bouneffouf; Amel Bouzeghoub; Alda Lopes Gancarski

doi:10.1007/978-3-642-42054-2_8

Risk-aware recommender systems

Djallel Bouneffouf
, Amel Bouzeghoub
, Alda Lopes Gancarski

CNRS UMR 5157 SAMOVAR

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

Abstract

Context-Aware Recommender Systems can naturally be modelled as an exploration/exploitation trade-off (exr/exp) problem, where the system has to choose between maximizing its expected rewards dealing with its current knowledge (exploitation) and learning more about the unknown user's preferences to improve its knowledge (exploration). This problem has been addressed by the reinforcement learning community but they do not consider the risk level of the current user's situation, where it may be dangerous to recommend items the user may not desire in her current situation if the risk level is high. We introduce in this paper an algorithm named R-UCB that considers the risk level of the user's situation to adaptively balance between exr and exp. The detailed analysis of the experimental results reveals several important discoveries in the exr/exp behaviour.

Original language	English
Title of host publication	Neural Information Processing - 20th International Conference, ICONIP 2013, Proceedings
Pages	57-65
Number of pages	9
Edition	PART 1
DOIs	https://doi.org/10.1007/978-3-642-42054-2_8
Publication status	Published - 1 Dec 2013
Event	20th International Conference on Neural Information Processing, ICONIP 2013 - Daegu, Korea, Republic of Duration: 3 Nov 2013 → 7 Nov 2013

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Number	PART 1
Volume	8226 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	20th International Conference on Neural Information Processing, ICONIP 2013
Country/Territory	Korea, Republic of
City	Daegu
Period	3/11/13 → 7/11/13

Access to Document

10.1007/978-3-642-42054-2_8

Cite this

Bouneffouf, D., Bouzeghoub, A., & Gancarski, A. L. (2013). Risk-aware recommender systems. In Neural Information Processing - 20th International Conference, ICONIP 2013, Proceedings (PART 1 ed., pp. 57-65). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 8226 LNCS, No. PART 1). https://doi.org/10.1007/978-3-642-42054-2_8

@inproceedings{77f48ad9783a4abc8ef843941b8af7a0,

title = "Risk-aware recommender systems",

abstract = "Context-Aware Recommender Systems can naturally be modelled as an exploration/exploitation trade-off (exr/exp) problem, where the system has to choose between maximizing its expected rewards dealing with its current knowledge (exploitation) and learning more about the unknown user's preferences to improve its knowledge (exploration). This problem has been addressed by the reinforcement learning community but they do not consider the risk level of the current user's situation, where it may be dangerous to recommend items the user may not desire in her current situation if the risk level is high. We introduce in this paper an algorithm named R-UCB that considers the risk level of the user's situation to adaptively balance between exr and exp. The detailed analysis of the experimental results reveals several important discoveries in the exr/exp behaviour.",

author = "Djallel Bouneffouf and Amel Bouzeghoub and Gancarski, \{Alda Lopes\}",

year = "2013",

month = dec,

day = "1",

doi = "10.1007/978-3-642-42054-2\_8",

language = "English",

isbn = "9783642420535",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

number = "PART 1",

pages = "57--65",

booktitle = "Neural Information Processing - 20th International Conference, ICONIP 2013, Proceedings",

edition = "PART 1",

note = "20th International Conference on Neural Information Processing, ICONIP 2013 ; Conference date: 03-11-2013 Through 07-11-2013",

}

Bouneffouf, D, Bouzeghoub, A & Gancarski, AL 2013, Risk-aware recommender systems. in Neural Information Processing - 20th International Conference, ICONIP 2013, Proceedings. PART 1 edn, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), no. PART 1, vol. 8226 LNCS, pp. 57-65, 20th International Conference on Neural Information Processing, ICONIP 2013, Daegu, Korea, Republic of, 3/11/13. https://doi.org/10.1007/978-3-642-42054-2_8

Risk-aware recommender systems. / Bouneffouf, Djallel; Bouzeghoub, Amel; Gancarski, Alda Lopes.
Neural Information Processing - 20th International Conference, ICONIP 2013, Proceedings. PART 1. ed. 2013. p. 57-65 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 8226 LNCS, No. PART 1).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - Risk-aware recommender systems

AU - Bouneffouf, Djallel

AU - Bouzeghoub, Amel

AU - Gancarski, Alda Lopes

PY - 2013/12/1

Y1 - 2013/12/1

N2 - Context-Aware Recommender Systems can naturally be modelled as an exploration/exploitation trade-off (exr/exp) problem, where the system has to choose between maximizing its expected rewards dealing with its current knowledge (exploitation) and learning more about the unknown user's preferences to improve its knowledge (exploration). This problem has been addressed by the reinforcement learning community but they do not consider the risk level of the current user's situation, where it may be dangerous to recommend items the user may not desire in her current situation if the risk level is high. We introduce in this paper an algorithm named R-UCB that considers the risk level of the user's situation to adaptively balance between exr and exp. The detailed analysis of the experimental results reveals several important discoveries in the exr/exp behaviour.

AB - Context-Aware Recommender Systems can naturally be modelled as an exploration/exploitation trade-off (exr/exp) problem, where the system has to choose between maximizing its expected rewards dealing with its current knowledge (exploitation) and learning more about the unknown user's preferences to improve its knowledge (exploration). This problem has been addressed by the reinforcement learning community but they do not consider the risk level of the current user's situation, where it may be dangerous to recommend items the user may not desire in her current situation if the risk level is high. We introduce in this paper an algorithm named R-UCB that considers the risk level of the user's situation to adaptively balance between exr and exp. The detailed analysis of the experimental results reveals several important discoveries in the exr/exp behaviour.

U2 - 10.1007/978-3-642-42054-2_8

DO - 10.1007/978-3-642-42054-2_8

M3 - Conference contribution

AN - SCOPUS:84893424636

SN - 9783642420535

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 57

EP - 65

BT - Neural Information Processing - 20th International Conference, ICONIP 2013, Proceedings

T2 - 20th International Conference on Neural Information Processing, ICONIP 2013

Y2 - 3 November 2013 through 7 November 2013

ER -

Risk-aware recommender systems

Abstract

Publication series

Conference

Access to Document

Fingerprint

Cite this