From preference into decision making: Modeling user interactions in recommender systems

Qian Zhao; Martijn C. Willemsen; Gediminas Adomavicius; F. Maxwell Harper; Joseph A. Konstan

doi:10.1145/3298689.3347065

From preference into decision making: Modeling user interactions in recommender systems

Qian Zhao, Martijn C. Willemsen, Gediminas Adomavicius, F. Maxwell Harper, Joseph A. Konstan

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

4 Scopus citations

Abstract

User-system interaction in recommender systems involves three aspects: temporal browsing (viewing recommendation lists and/or searching/fltering), action (performing actions on recommended items, e.g., clicking, consuming) and inaction (neglecting or skipping recommended items). Modern recommenders build machine learning models from recordings of such user interaction with the system, and in doing so they commonly make certain assumptions (e.g., pairwise preference orders, independent or competitive probabilistic choices, etc.). In this paper, we set out to study the efects of these assumptions along three dimensions in eight diferent single models and three associated hybrid models on a user browsing data set collected from a real-world recommender system application. We further design a novel model based on recurrent neural networks and multi-task learning, inspired by Decision Field Theory, a model of human decision making. We report on precision, recall, and MAP, fnding that this new model outperforms the others.

Original language	English (US)
Title of host publication	RecSys 2019 - 13th ACM Conference on Recommender Systems
Publisher	Association for Computing Machinery, Inc
Pages	29-33
Number of pages	5
ISBN (Electronic)	9781450362436
DOIs	https://doi.org/10.1145/3298689.3347065
State	Published - Sep 10 2019
Event	13th ACM Conference on Recommender Systems, RecSys 2019 - Copenhagen, Denmark Duration: Sep 16 2019 → Sep 20 2019

Publication series

Name	RecSys 2019 - 13th ACM Conference on Recommender Systems

Conference

Conference	13th ACM Conference on Recommender Systems, RecSys 2019
Country/Territory	Denmark
City	Copenhagen
Period	9/16/19 → 9/20/19

Bibliographical note

Funding Information:
This work was supported by the National Science Foundation under grant IIS-1319382. The frst author was also supported by the Doctoral Dissertation Fellowship, 2016-17, by the Graduate School at the University of Minnesota.

Publisher Copyright:
© 2019 Association for Computing Machinery.

Keywords

Decision feld theory
Decision making
Recurrent neural networks

Access

10.1145/3298689.3347065

OpenUrl availability

Full text

Cite this

Zhao, Q., Willemsen, M. C., Adomavicius, G., Maxwell Harper, F., & Konstan, J. A. (2019). From preference into decision making: Modeling user interactions in recommender systems. In RecSys 2019 - 13th ACM Conference on Recommender Systems (pp. 29-33). (RecSys 2019 - 13th ACM Conference on Recommender Systems). Association for Computing Machinery, Inc. https://doi.org/10.1145/3298689.3347065

From preference into decision making: Modeling user interactions in recommender systems. / Zhao, Qian; Willemsen, Martijn C.; Adomavicius, Gediminas et al.
RecSys 2019 - 13th ACM Conference on Recommender Systems. Association for Computing Machinery, Inc, 2019. p. 29-33 (RecSys 2019 - 13th ACM Conference on Recommender Systems).

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

Zhao, Q, Willemsen, MC, Adomavicius, G, Maxwell Harper, F & Konstan, JA 2019, From preference into decision making: Modeling user interactions in recommender systems. in RecSys 2019 - 13th ACM Conference on Recommender Systems. RecSys 2019 - 13th ACM Conference on Recommender Systems, Association for Computing Machinery, Inc, pp. 29-33, 13th ACM Conference on Recommender Systems, RecSys 2019, Copenhagen, Denmark, 9/16/19. https://doi.org/10.1145/3298689.3347065

@inproceedings{3490811d06494f50aca7e6bdae2d1f5c,

title = "From preference into decision making: Modeling user interactions in recommender systems",

abstract = "User-system interaction in recommender systems involves three aspects: temporal browsing (viewing recommendation lists and/or searching/fltering), action (performing actions on recommended items, e.g., clicking, consuming) and inaction (neglecting or skipping recommended items). Modern recommenders build machine learning models from recordings of such user interaction with the system, and in doing so they commonly make certain assumptions (e.g., pairwise preference orders, independent or competitive probabilistic choices, etc.). In this paper, we set out to study the efects of these assumptions along three dimensions in eight diferent single models and three associated hybrid models on a user browsing data set collected from a real-world recommender system application. We further design a novel model based on recurrent neural networks and multi-task learning, inspired by Decision Field Theory, a model of human decision making. We report on precision, recall, and MAP, fnding that this new model outperforms the others.",

keywords = "Decision feld theory, Decision making, Recurrent neural networks",

author = "Qian Zhao and Willemsen, {Martijn C.} and Gediminas Adomavicius and {Maxwell Harper}, F. and Konstan, {Joseph A.}",

note = "Funding Information: This work was supported by the National Science Foundation under grant IIS-1319382. The frst author was also supported by the Doctoral Dissertation Fellowship, 2016-17, by the Graduate School at the University of Minnesota. Publisher Copyright: {\textcopyright} 2019 Association for Computing Machinery.; 13th ACM Conference on Recommender Systems, RecSys 2019 ; Conference date: 16-09-2019 Through 20-09-2019",

year = "2019",

month = sep,

day = "10",

doi = "10.1145/3298689.3347065",

language = "English (US)",

series = "RecSys 2019 - 13th ACM Conference on Recommender Systems",

publisher = "Association for Computing Machinery, Inc",

pages = "29--33",

booktitle = "RecSys 2019 - 13th ACM Conference on Recommender Systems",

}

TY - GEN

T1 - From preference into decision making

T2 - 13th ACM Conference on Recommender Systems, RecSys 2019

AU - Zhao, Qian

AU - Willemsen, Martijn C.

AU - Adomavicius, Gediminas

AU - Maxwell Harper, F.

AU - Konstan, Joseph A.

N1 - Funding Information: This work was supported by the National Science Foundation under grant IIS-1319382. The frst author was also supported by the Doctoral Dissertation Fellowship, 2016-17, by the Graduate School at the University of Minnesota. Publisher Copyright: © 2019 Association for Computing Machinery.

PY - 2019/9/10

Y1 - 2019/9/10

N2 - User-system interaction in recommender systems involves three aspects: temporal browsing (viewing recommendation lists and/or searching/fltering), action (performing actions on recommended items, e.g., clicking, consuming) and inaction (neglecting or skipping recommended items). Modern recommenders build machine learning models from recordings of such user interaction with the system, and in doing so they commonly make certain assumptions (e.g., pairwise preference orders, independent or competitive probabilistic choices, etc.). In this paper, we set out to study the efects of these assumptions along three dimensions in eight diferent single models and three associated hybrid models on a user browsing data set collected from a real-world recommender system application. We further design a novel model based on recurrent neural networks and multi-task learning, inspired by Decision Field Theory, a model of human decision making. We report on precision, recall, and MAP, fnding that this new model outperforms the others.

AB - User-system interaction in recommender systems involves three aspects: temporal browsing (viewing recommendation lists and/or searching/fltering), action (performing actions on recommended items, e.g., clicking, consuming) and inaction (neglecting or skipping recommended items). Modern recommenders build machine learning models from recordings of such user interaction with the system, and in doing so they commonly make certain assumptions (e.g., pairwise preference orders, independent or competitive probabilistic choices, etc.). In this paper, we set out to study the efects of these assumptions along three dimensions in eight diferent single models and three associated hybrid models on a user browsing data set collected from a real-world recommender system application. We further design a novel model based on recurrent neural networks and multi-task learning, inspired by Decision Field Theory, a model of human decision making. We report on precision, recall, and MAP, fnding that this new model outperforms the others.

KW - Decision feld theory

KW - Decision making

KW - Recurrent neural networks

UR - http://www.scopus.com/inward/record.url?scp=85073341933&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85073341933&partnerID=8YFLogxK

U2 - 10.1145/3298689.3347065

DO - 10.1145/3298689.3347065

M3 - Conference contribution

AN - SCOPUS:85073341933

T3 - RecSys 2019 - 13th ACM Conference on Recommender Systems

SP - 29

EP - 33

BT - RecSys 2019 - 13th ACM Conference on Recommender Systems

PB - Association for Computing Machinery, Inc

Y2 - 16 September 2019 through 20 September 2019

ER -

From preference into decision making: Modeling user interactions in recommender systems

Abstract

Publication series

Conference

Bibliographical note

Keywords

Access

OpenUrl availability

Other files and links

Fingerprint

Cite this