Fast online learning to recommend a diverse set from big data

Mahmuda Rahman, Jae C Oh

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

3 Scopus citations

Abstract

Building a recommendation system to withstand the rapid change in items’ relevance to users is a challenge requiring continual optimization. In a Big Data scenario, it becomes a harder problem, in which users get substantially diverse in their tastes. We propose an algorithm that is based on the UBC1 bandit algorithm to cover a large variety of users. To enhance UCB1, we designed a new rewarding scheme to encourage the bandits to choose items that satisfy a large number of users. Our approach takes account of the correlation among the items preferred by different types of users, in effect, increasing the coverage of the recommendation set efficiently. Our method performs better than existing techniques such as Ranked Bandits [8] and Independent Bandits [6] in terms of satisfying diverse types of users.

Original languageEnglish (US)
Title of host publicationLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
PublisherSpringer Verlag
Pages361-370
Number of pages10
Volume9101
ISBN (Print)9783319190655
DOIs
StatePublished - 2015
Event28th International Conference on Industrial, Engineering and Other Applications of Applied Intelligent Systems, IEA/AIE 2015 - Seoul, Korea, Republic of
Duration: Jun 10 2015Jun 12 2015

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume9101
ISSN (Print)03029743
ISSN (Electronic)16113349

Other

Other28th International Conference on Industrial, Engineering and Other Applications of Applied Intelligent Systems, IEA/AIE 2015
CountryKorea, Republic of
CitySeoul
Period6/10/156/12/15

Keywords

  • Diversity
  • Multi armed bandit
  • Online learning
  • Recommendation system
  • Upper confidence bound

ASJC Scopus subject areas

  • Computer Science(all)
  • Theoretical Computer Science

Fingerprint Dive into the research topics of 'Fast online learning to recommend a diverse set from big data'. Together they form a unique fingerprint.

  • Cite this

    Rahman, M., & Oh, J. C. (2015). Fast online learning to recommend a diverse set from big data. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 9101, pp. 361-370). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 9101). Springer Verlag. https://doi.org/10.1007/978-3-319-19066-2_35