Open AccessActivity-based Friend Recommendation System (ARS) in Location-based Social Network
Author(s) -
Jae Hyuk Kim,
Yeunwoong Kyung
Publication year - 2022
Publication title -
webology
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.259
H-Index - 18
ISSN - 1735-188X
DOI - 10.14704/web/v19i1/web19295
Subject(s) - computer science , recommender system , location data , social network (sociolinguistics) , activity detection , location tracking , similarity (geometry) , tracking (education) , information retrieval , location aware , location based service , social activity , world wide web , data mining , social media , artificial intelligence , computer network , real time computing , psychology , pedagogy , social science , sociology , image (mathematics)
Common friend and place recommendation services in Location-based Social Network (LBSN) is based on user’s location tracking. However, since each user can do different activities even in the same place, location data is not enough to provide accurate recommendation for LSBN. To address this problem, Activity-based friend and place Recommendation System (ARS) is proposed. ARS considers two additional factors to improve recommendation accuracy: time and activity. ARS collects the time-related activity and location data from users through the developed scheduler application and then performs the recommendation for users based on the calculated similarity among them. Performance evaluation shows that ARS can provide accurate recommendation between users who have similar activity and location patterns according to time.