Sustavi za preporučivanje zasnovani na suradničkom filtriranju

master's thesis

• Title (english): Recommendation Systems Based on Collaborative Filtering
• Author: Nikolina Jukić
• Mentor: Marin Šilić (mentor)
• Committee member: Marin Šilić (predsjednik povjerenstva)
• Committee member: Goran Delač (član povjerenstva)
• Committee member: Klemo Vladimir (član povjerenstva)
• Granter: University of Zagreb; Faculty of Electrical Engineering and Computing; Zagreb
• Defense date and country: 2023-07-04, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Computing

Abstract

Širenjem internetskih tehnologija povećava se broj korisnika što uzrokuje rast dostupne količine informacija, odnosno dolazi do preopterećenja podatcima. Jedno od rješenja je razvoj sustava za preporučivanje. Opisane su vrste, svojstva i algoritmi sustava za preporučivanje s fokusom na sustave zasnovane na suradničkom filtriranju. Programski je ostvaren težinski hibridni sustav za preporučivanje koji koristi metodu k-najbližih susjeda i matričnu faktorizaciju. Provedeno je vrednovanje sustava na tri skupa podataka različitih veličina, ljestvica ocjena i informacija. Uspoređeni su rezultati pojedinačnih pristupa s hibridnim pristupom te je navedena korištena literatura.

Abstract (english)

With the rapid growth of Internet technologies number of users increases resulting in growth of available data and information overload. One solution is implementation of recommender systems. This thesis covers some types of recommender systems, their properties and algorithms they use, with focus on collaborative filtering methods. A weighted hybrid recommender system using k-nearest neighbors algorithm and matrix factorization was implemented. The evaluation of proposed recommender system was done on three different datasets, varying in dataset size, rating scale and type of information. Results of individual approaches were compared against the weighted hybrid approach.

• Keywords: matrična faktorizacija; težinski hibridni sustav; metoda k-najbližih susjeda; sustavi zasnovani na memoriji; sustavi zasnovani na modelu
• Keywords (english): matrix factorization; weighted hybrid recommender system; k-nearest neighbors algorithm; memory-based recommender systems; model-based recommender systems
• Language: croatian
• URN:NBN: urn:nbn:hr:168:866606
• Study programme: Title: Computing; Study programme type: university; Study level: graduate; Academic / professional title: magistar/magistra inženjer/inženjerka računarstva (magistar/magistra inženjer/inženjerka računarstva)
• Type of resource: Text
• File origin: Born digital
• Access conditions: Closed access
• Created on: 2023-10-11 09:11:08