A Unified Contextual Bandit Framework for Long- and Short-Term Recommendations

Publikation: Beiträge in SammelwerkenAufsätze in KonferenzbändenForschungbegutachtet

Standard

A Unified Contextual Bandit Framework for Long- and Short-Term Recommendations. / Tavakol, Maryam; Brefeld, Ulf.
Machine Learning and Knowledge Discovery in Databases: European Conference, ECML PKDD 2017 Skopje, Macedonia, September 18 – 22, 2017; Proceedings, Part II. Hrsg. / Michelangelo Ceci; Jaakko Hollmen; Ljupco Todorovski; Celine Vens; Saso Dzeroski. Band 2 Springer Verlag, 2017. S. 269-284 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Band 10535 LNAI).

Publikation: Beiträge in SammelwerkenAufsätze in KonferenzbändenForschungbegutachtet

Harvard

Tavakol, M & Brefeld, U 2017, A Unified Contextual Bandit Framework for Long- and Short-Term Recommendations. in M Ceci, J Hollmen, L Todorovski, C Vens & S Dzeroski (Hrsg.), Machine Learning and Knowledge Discovery in Databases: European Conference, ECML PKDD 2017 Skopje, Macedonia, September 18 – 22, 2017; Proceedings, Part II. Bd. 2, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), Bd. 10535 LNAI, Springer Verlag, S. 269-284, THE EUROPEAN CONFERENCE ON MACHINE LEARNING & PRINCIPLES AND PRACTICE OF KNOWLEDGE DISCOVERY IN DATABASES 2017, Skopje, Mazedonien, ehemalige jugoslawische Republik, 18.09.17. https://doi.org/10.1007/978-3-319-71246-8_17

APA

Tavakol, M., & Brefeld, U. (2017). A Unified Contextual Bandit Framework for Long- and Short-Term Recommendations. In M. Ceci, J. Hollmen, L. Todorovski, C. Vens, & S. Dzeroski (Hrsg.), Machine Learning and Knowledge Discovery in Databases: European Conference, ECML PKDD 2017 Skopje, Macedonia, September 18 – 22, 2017; Proceedings, Part II (Band 2, S. 269-284). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Band 10535 LNAI). Springer Verlag. https://doi.org/10.1007/978-3-319-71246-8_17

Vancouver

Tavakol M, Brefeld U. A Unified Contextual Bandit Framework for Long- and Short-Term Recommendations. in Ceci M, Hollmen J, Todorovski L, Vens C, Dzeroski S, Hrsg., Machine Learning and Knowledge Discovery in Databases: European Conference, ECML PKDD 2017 Skopje, Macedonia, September 18 – 22, 2017; Proceedings, Part II. Band 2. Springer Verlag. 2017. S. 269-284. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-319-71246-8_17

Bibtex

@inbook{0f45dbe652f64731ab63d83ca20dbbef,
title = "A Unified Contextual Bandit Framework for Long- and Short-Term Recommendations",
abstract = "We present a unified contextual bandit framework for recommendation problems that is able to capture long- and short-term interests of users. The model is devised in dual space and the derivation is consequentially carried out using Fenchel-Legrende conjugates and thus leverages to a wide range of tasks and settings.We detail two instantiations for regression and classification scenarios and obtain well-known algorithms for these special cases. The resulting general and unified framework allows for quickly adapting contextual bandits to different applications at-hand. The empirical study demonstrates that the proposed long- and short-term framework outperforms both, short-term and long-term models on data. Moreover, a tweak of the combined model proves beneficial in cold start problems.",
keywords = "Informatics, Machine learning, Recommender systems, Contextual bandits, Recommendation, Dual optimization, Personalization",
author = "Maryam Tavakol and Ulf Brefeld",
year = "2017",
month = dec,
day = "30",
doi = "10.1007/978-3-319-71246-8_17",
language = "English",
isbn = "978-3-319-71245-1",
volume = "2",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer Verlag",
pages = "269--284",
editor = "Michelangelo Ceci and Jaakko Hollmen and Ljupco Todorovski and Celine Vens and Saso Dzeroski",
booktitle = "Machine Learning and Knowledge Discovery in Databases",
address = "Germany",
note = "THE EUROPEAN CONFERENCE ON MACHINE LEARNING & PRINCIPLES AND PRACTICE OF KNOWLEDGE DISCOVERY IN DATABASES 2017, ECML PKDD 2017 ; Conference date: 18-09-2017 Through 22-09-2017",
url = "http://ecmlpkdd2017.ijs.si/",

}

RIS

TY - CHAP

T1 - A Unified Contextual Bandit Framework for Long- and Short-Term Recommendations

AU - Tavakol, Maryam

AU - Brefeld, Ulf

N1 - Conference code: 27

PY - 2017/12/30

Y1 - 2017/12/30

N2 - We present a unified contextual bandit framework for recommendation problems that is able to capture long- and short-term interests of users. The model is devised in dual space and the derivation is consequentially carried out using Fenchel-Legrende conjugates and thus leverages to a wide range of tasks and settings.We detail two instantiations for regression and classification scenarios and obtain well-known algorithms for these special cases. The resulting general and unified framework allows for quickly adapting contextual bandits to different applications at-hand. The empirical study demonstrates that the proposed long- and short-term framework outperforms both, short-term and long-term models on data. Moreover, a tweak of the combined model proves beneficial in cold start problems.

AB - We present a unified contextual bandit framework for recommendation problems that is able to capture long- and short-term interests of users. The model is devised in dual space and the derivation is consequentially carried out using Fenchel-Legrende conjugates and thus leverages to a wide range of tasks and settings.We detail two instantiations for regression and classification scenarios and obtain well-known algorithms for these special cases. The resulting general and unified framework allows for quickly adapting contextual bandits to different applications at-hand. The empirical study demonstrates that the proposed long- and short-term framework outperforms both, short-term and long-term models on data. Moreover, a tweak of the combined model proves beneficial in cold start problems.

KW - Informatics

KW - Machine learning

KW - Recommender systems

KW - Contextual bandits

KW - Recommendation

KW - Dual optimization

KW - Personalization

UR - http://ecmlpkdd2017.ijs.si/papers/paperID239.pdf

UR - http://www.scopus.com/inward/record.url?scp=85040222425&partnerID=8YFLogxK

U2 - 10.1007/978-3-319-71246-8_17

DO - 10.1007/978-3-319-71246-8_17

M3 - Article in conference proceedings

SN - 978-3-319-71245-1

VL - 2

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

SP - 269

EP - 284

BT - Machine Learning and Knowledge Discovery in Databases

A2 - Ceci, Michelangelo

A2 - Hollmen, Jaakko

A2 - Todorovski, Ljupco

A2 - Vens, Celine

A2 - Dzeroski, Saso

PB - Springer Verlag

T2 - THE EUROPEAN CONFERENCE ON MACHINE LEARNING & PRINCIPLES AND PRACTICE OF KNOWLEDGE DISCOVERY IN DATABASES 2017

Y2 - 18 September 2017 through 22 September 2017

ER -

DOI

Zuletzt angesehen

Publikationen

  1. The science-policy interface on ecosystems and people
  2. Influence of Long-Lasting Static Stretching on Maximal Strength, Muscle Thickness and Flexibility
  3. Industrial Application Study and Development of a Control Strategy for an Ice Clamping Device
  4. Biodiversity buffers pollination from changes in environmental conditions
  5. EURODEMO - Improving the uptake of efficient soil and groundwater remediation technologies
  6. Das Schreiben, das Interpretieren, die Tatsachen
  7. Brain Drain
  8. L'agenda 21 locale
  9. International Master’s Programme in Sustainable Development and Management
  10. Critical evaluation of commonly used methods to determine the concordance between sonography and magnetic resonance imaging: A comparative study
  11. Embedded Self-Managing Modes of Organizing
  12. Love in Paramyth
  13. Firm size and the use of export intermediaries.
  14. A sensitive microsystem as biosensor for cell growth monitoring and antibiotic testing
  15. Frank Fischer/Herbert Gottweis (Hg.) The Argumentative Turn Revisited.
  16. Schreiben Englisch
  17. Tourismuswissenschaft
  18. Groundwater abstraction for irrigation and its impacts on low flows in a watershed in Northwest Germany
  19. How can we bring together empiricists and modellers in functional biodiversity research?
  20. Mimikbasierte Emotionserfassung anhand von dynamischen Flächen und Streckenveränderung
  21. The new US horizontal merger guidelines
  22. The global perspective of education for sustainable development
  23. Mildes Luthertum?
  24. Translating European labor relations practices to the United States through global framework agreements?
  25. The impact of growth markets in the downstream sector - the parameters for connectivity and services: Beyond outer space law
  26. Autonomie der Migration
  27. Artist Placement Group
  28. Vergütung, variable
  29. When status differences are illegitimate, groups' needs diverge
  30. Mehrsprachigkeit in der Grundschule
  31. Is innovative firm behavior correlated with age and gender composition of the workforce ?
  32. Unterrichtsqualität an Hamburger Grundschulen
  33. Pharmaceuticals in the Environment
  34. Interferences and Events
  35. Anton Schnack: Werke in zwei Bänden
  36. "Die Lüneburger Heide goes digital"
  37. Synapses in the Network
  38. Vergleichende Regionalismusforschung
  39. Standardized Tests Fail to Assess the Effects of Antibiotics on Environmental Bacteria
  40. Productivity premia for many modes of internationalization.
  41. Resilience and coastal governance