TítolCollaborative Judgement
Publication TypeConference Proceedings
Year of Conference2015
AuthorsAndrejczuk E, Rodríguez-Aguilar JA, Sierra C
Conference NameProceedings of 18th International Conference on Principles and Practice of Multi-Agent Systems (PRIMA 2015)
Volume9387
EditorSpringer International Publishing
Paginació631-639
Conference LocationBertinoro, Italy
Date Published24/10/2015
ISBN Number978-3-319-25524-8
Paraules claurecommender systems, reputation, Trust
Resum

In this paper we introduce a new ranking algorithm, called Collaborative Judgement (CJ), that takes into account peer opinions of agents and/or humans on objects (e.g. products, exams, papers) as well as peer judgements over those opinions. The combination of these two types of information has not been studied in previous work in order to produce object rankings. We apply CJ to the use case of scientific paper assessment and we validate it over simulated data. The results show that the rankings produced by our algorithm improve current scientific paper ranking practice based on averages of opinions weighted by their reviewers’ self-assessments.