Movie Recommender System Based On Collaborative Filtering

Autor: Fain, Maria
Přispěvatelé: Delač, Goran
Jazyk: chorvatština
Rok vydání: 2022
Předmět:
Popis: Sustavi preporuke stvoreni su u svrhu unaprijeđivanja pretraživanja i korištenja sadržaja korisnicima u virtualnom svijetu. Jedna vrsta sustava za preporuku su sustavi preporuke zasnovani na suradničkom filtriranju čiji način rada je opširnije opisan u ovome radu. Ponuđen je opis metrika koje se koriste u algoritmima preporuke i opis izdvojenih algoritama preporuke zasnovanih na suradničkom filtriranju. Napravljeno je jednostavno programsko ostvarenje sustava preporuke filmova zasnovanog na suradničkom filtriranju koristeći bazu podataka MovieLens. Nakon programskog ostvarenja, provedeno je vrednovanje uspješnosti i preciznosti korištenih algoritama preporuke. Recommendation systems were built with the purpose of upgrading the search and use of content for users in the virtual world. This paper explains the functionality of a certain kind of recommendation systems called recommendation systems based on collaborative filtering. A description of similarity metrics that are used in recommendation algorithms is given along with a description of chosen recommendation algorithms based on collaborative filtering. A simple film recommendation system based on collaborative filtering is implemented using the MovieLens database. Following the implementation, a calculation using accuracy metrics has been made for each of the used algorithms.
Databáze: OpenAIRE