TY - GEN
T1 - Measuring rating exigency
T2 - 22nd International Conference of the Catalan Association for Artificial Intelligence, CCIA 2019
AU - Nguyen, Jennifer
AU - Montserrat-Adell, Jordi
AU - Armisen, Albert
AU - Torrens, M.
AU - Agell, N.
N1 - Funding Information:
This research has been partially supported by the INVITE Research Project (TIN2016-80049-C2-1-R and TIN2016-80049-C2-2-R (AEI/FEDER, UE)), funded by the Spanish Ministry of Science and Information Technology.
Publisher Copyright:
© 2019 The authors and IOS Press. All rights reserved.
PY - 2019/9/6
Y1 - 2019/9/6
N2 - In a collaborative filtering recommender system context, users are matched with items liked by others who have similar interests. However, each person may evaluate items differently according to their own experiences and standards. Therefore, analyzing the degree of exigency of an individual with respect to others consuming the same items is relevant. We propose a fuzzy approach based on a measure of consensus among users considering ratings. The metric considers the distances between users and a central measure. A centroid of the ratings for each individual item is proposed as the benchmark from which people's exigency is measured as being more or less stringent than his peers. In addition, the method will allow us to identify people with different degrees of exigency towards a specific type of item which can facilitate a more relevant recommendation. The model is implemented in a real case dataset of restaurants from the Yelp platform.
AB - In a collaborative filtering recommender system context, users are matched with items liked by others who have similar interests. However, each person may evaluate items differently according to their own experiences and standards. Therefore, analyzing the degree of exigency of an individual with respect to others consuming the same items is relevant. We propose a fuzzy approach based on a measure of consensus among users considering ratings. The metric considers the distances between users and a central measure. A centroid of the ratings for each individual item is proposed as the benchmark from which people's exigency is measured as being more or less stringent than his peers. In addition, the method will allow us to identify people with different degrees of exigency towards a specific type of item which can facilitate a more relevant recommendation. The model is implemented in a real case dataset of restaurants from the Yelp platform.
KW - Application
KW - Consensus measurement
KW - Fuzzy linguistic terms
KW - Recommender Systems
UR - http://www.scopus.com/inward/record.url?scp=85084978929&partnerID=8YFLogxK
U2 - 10.3233/FAIA190132
DO - 10.3233/FAIA190132
M3 - Conference contribution
AN - SCOPUS:85084978929
T3 - Frontiers in Artificial Intelligence and Applications
SP - 256
EP - 265
BT - Artificial Intelligence Research and Development - Proceedings of the 22nd International Conference of the Catalan Association for Artificial Intelligence, CCIA 2019
A2 - Sabater-Mir, Jordi
A2 - Torra, Vicenc
A2 - Aguilo, Isabel
A2 - Gonzalez-Hidalgo, Manuel
PB - IOS Press
Y2 - 23 October 2019 through 25 October 2019
ER -