Towards Results-level Proportionality for Multi-objective Recommender Systems

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F00216208%3A11320%2F22%3A10447483" target="_blank" >RIV/00216208:11320/22:10447483 - isvavai.cz</a>

Výsledek na webu

<a href="https://doi.org/10.1145/3477495.3531787" target="_blank" >https://doi.org/10.1145/3477495.3531787</a>

DOI - Digital Object Identifier

<a href="http://dx.doi.org/10.1145/3477495.3531787" target="_blank" >10.1145/3477495.3531787</a>

Jazyk výsledku

angličtina

Název v původním jazyce

Towards Results-level Proportionality for Multi-objective Recommender Systems

Popis výsledku v původním jazyce

The main focus of our work is the problem of multiple objectives optimization (MOO) while providing a final list of recommendations to the user. Currently, system designers can tune MOO by setting importance of individual objectives, usually in some kind of weighted average setting. However, this does not have to translate into the presence of such objectives in the final results. In contrast, in our work we would like to allow system designers or end-users to directly quantify the required relative ratios of individual objectives in the resulting recommendations, e.g., the final results should have 60% relevance, 30% diversity and 10% novelty. If individual objectives are transformed to represent quality on the same scale, these result conditioning expressions may greatly contribute towards recommendations tuneability and explainability as well as user&apos;s control over recommendations. To achieve this task, we propose an iterative algorithm inspired by the mandates allocation problem in public elections. The algorithm is applicable as long as per-item marginal gains of individual objectives can be calculated. Effectiveness of the algorithm is evaluated on several settings of relevance-novelty-diversity optimization problem. Furthermore, we also outline several options to scale individual objectives to represent similar value for the user.

Název v anglickém jazyce

Towards Results-level Proportionality for Multi-objective Recommender Systems

Popis výsledku anglicky

The main focus of our work is the problem of multiple objectives optimization (MOO) while providing a final list of recommendations to the user. Currently, system designers can tune MOO by setting importance of individual objectives, usually in some kind of weighted average setting. However, this does not have to translate into the presence of such objectives in the final results. In contrast, in our work we would like to allow system designers or end-users to directly quantify the required relative ratios of individual objectives in the resulting recommendations, e.g., the final results should have 60% relevance, 30% diversity and 10% novelty. If individual objectives are transformed to represent quality on the same scale, these result conditioning expressions may greatly contribute towards recommendations tuneability and explainability as well as user&apos;s control over recommendations. To achieve this task, we propose an iterative algorithm inspired by the mandates allocation problem in public elections. The algorithm is applicable as long as per-item marginal gains of individual objectives can be calculated. Effectiveness of the algorithm is evaluated on several settings of relevance-novelty-diversity optimization problem. Furthermore, we also outline several options to scale individual objectives to represent similar value for the user.

Druh

D - Stať ve sborníku

CEP obor

—

OECD FORD obor

10201 - Computer sciences, information science, bioinformathics (hardware development to be 2.2, social aspect to be 5.8)

Projekt

<a href="/cs/project/GA22-21696S" target="_blank" >GA22-21696S: Hluboké vizuální reprezentace nestrukturovaných dat</a><br>

Návaznosti

P - Projekt vyzkumu a vyvoje financovany z verejnych zdroju (s odkazem do CEP)

Rok uplatnění

2022

Kód důvěrnosti údajů

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Název statě ve sborníku

SIGIR 2022 - Proceedings of the 45th International ACM SIGIR Conference on Research and Development in Information Retrieval

ISBN

978-1-4503-8732-3

ISSN

—

e-ISSN

—

Počet stran výsledku

6

Strana od-do

1963-1968

Název nakladatele

ACM

Místo vydání

New York, NY, USA

Místo konání akce

Madrid, Spain

Datum konání akce

11. 7. 2022

Typ akce podle státní příslušnosti

WRD - Celosvětová akce

Kód UT WoS článku

—