Disaggregating User Evaluations Using the Shapley Value

Matthias Feldotto, Claus-Jochen Haake, Alexander Skopalik, Nadja Stroh-Maraun

Research output: Chapter in Book/Report/Conference proceedingConference contributionAcademicpeer-review

Abstract

We consider a market where final products or services are compositions of a number of basic services. Users are asked to evaluate the quality of the composed product after purchase. The quality of the basic service influences the performance of the composed services but cannot be observed directly. The question we pose is whether it is possible to use user evaluations on composed services to assess the quality of basic services. We discuss how to combine aggregation of evaluations across users and disaggregation of information on composed services to derive valuations for the single components. As a solution we propose to use the (weighted) average as aggregation device in connection with the Shapley value as disaggregation method, since this combination fulfills natural requirements in our context. In addition, we address some occurring computational issues: We give an approximate solution concept using only a limited number of evaluations which guarantees nearly optimal results with reduced running time. Lastly, we show that a slightly modified Shapley value and the weighted average are still applicable if the evaluation profiles are incomplete.
Original languageEnglish
Title of host publicationProceedings of NetEcon '18 the 13th Workshop on the Economics of Networks, Systems and Computation (NetEcon 2018)
Subtitle of host publication the 13th Workshop on the Economics of Networks, Systems and Computation : June 18, 2018, Irvine, California, USA
Place of PublicationNew York, NY
PublisherAssociation for Computing Machinery (ACM)
Pages5:1-5:6
Number of pages6
ISBN (Print)978-1-4503-5916-0
DOIs
Publication statusPublished - 1 Jun 2018
Externally publishedYes
EventNetEcon 2018: the 13th Workshop on the Economics of Networks, Systems and Computation - Irvin, United States
Duration: 18 Jun 201818 Jun 2018
Conference number: 13
http://www.cse.cuhk.edu.hk/~cslui/NetEcon2018/

Conference

ConferenceNetEcon 2018
Abbreviated titleNetEcon 2018
CountryUnited States
CityIrvin
Period18/06/1818/06/18
Internet address

Fingerprint Dive into the research topics of 'Disaggregating User Evaluations Using the Shapley Value'. Together they form a unique fingerprint.

Cite this