Naslov
An emotion-based approach to developing a video recommendation system

Autori
Cvitanović, Ivana

Vrsta, podvrsta i kategorija rada
Ocjenski radovi, diplomski rad, preddiplomski

Fakultet
Fakultet elektrotehnike i računarstva

Mjesto
Zagreb

Datum
12.07

Godina
2019

Stranica
40

Mentor
Bagić Babac, Marina

Ključne riječi
Recommendation system, Natural language processing, Text classification, Collaborative filtering

Sažetak
Discovering new items, books, movies, music, etc is easier today than ever before and Recommender systems contribute to making this possible. Recommender systems make the online experience much more enjoyable for the user, by providing custom recommendations. Making these recommendations as precise and accustomed to the user is an ongoing challenge. In this thesis, the problem is approached from the emotional aspect. User’s feedback (comments) is observed to gain information on the user’s emotional stance regarding the item. Text classification is used to classify the comments based on the expressed emotion. Collaborative filtering ideas are adopted, and users are matched based on the similarity of emotion expressed on the same item.

Izvorni jezik
Engleski

Znanstvena područja
Računarstvo

POVEZANOST RADA