prikaz prve stranice dokumenta Raspodijeljeni sustav za preporučivanje na platformi Apache Spark
No public access
master's thesis
Raspodijeljeni sustav za preporučivanje na platformi Apache Spark
Zagreb: University of Zagreb, Faculty of Electrical Engineering and Computing, 2017. urn:nbn:hr:168:601260
Popić, Filip
University of Zagreb
Faculty of Electrical Engineering and Computing

Institutional repository: FER Repository

Cite this document

Popić, F. (2017). Raspodijeljeni sustav za preporučivanje na platformi Apache Spark (Master's thesis). Zagreb: University of Zagreb, Faculty of Electrical Engineering and Computing. Retrieved from https://urn.nsk.hr/urn:nbn:hr:168:601260

Popić, Filip. "Raspodijeljeni sustav za preporučivanje na platformi Apache Spark." Master's thesis, University of Zagreb, Faculty of Electrical Engineering and Computing, 2017. https://urn.nsk.hr/urn:nbn:hr:168:601260

Popić, Filip. "Raspodijeljeni sustav za preporučivanje na platformi Apache Spark." Master's thesis, University of Zagreb, Faculty of Electrical Engineering and Computing, 2017. https://urn.nsk.hr/urn:nbn:hr:168:601260

Popić, F. (2017). 'Raspodijeljeni sustav za preporučivanje na platformi Apache Spark', Master's thesis, University of Zagreb, Faculty of Electrical Engineering and Computing, accessed 14 November 2024, https://urn.nsk.hr/urn:nbn:hr:168:601260

Popić F. Raspodijeljeni sustav za preporučivanje na platformi Apache Spark [Master's thesis]. Zagreb: University of Zagreb, Faculty of Electrical Engineering and Computing; 2017 [cited 2024 November 14] Available at: https://urn.nsk.hr/urn:nbn:hr:168:601260

F. Popić, "Raspodijeljeni sustav za preporučivanje na platformi Apache Spark", Master's thesis, University of Zagreb, Faculty of Electrical Engineering and Computing, Zagreb, 2017. Available at: https://urn.nsk.hr/urn:nbn:hr:168:601260

Metadata
TitleRaspodijeljeni sustav za preporučivanje na platformi Apache Spark
Title (english)A Distributed Recommender System on Apache Spark
AuthorFilip Popić
MentorKrešimir Pripužić (mentor)
Committee memberKrešimir Pripužić (predsjednik povjerenstva)
Committee memberDragan Jevtić (član povjerenstva)
Committee memberMarin Vuković (član povjerenstva)
GranterUniversity of Zagreb
Faculty of Electrical Engineering and Computing
Zagreb
Defense date and country2017-07-11, Croatia
Scientific / art field,
discipline and subdiscipline		TECHNICAL SCIENCES
Computing
Abstract
U ovom radu izrađen je raspodijeljeni sustav za preporučivanje. Algoritam preporučivanja temeljen je na suradnji korisnika i korištenju matrice sličnosti objekata preporučivanja. Za izračun matrice sličnosti, podržano je nekoliko različitih mjera sličnosti, koje na temelju ulaznih statistika o korisničkim konzumacijama, računaju sličnost za sve parove objekata. Pri izračunu matrice sličnosti, omogućen je postupak normiranja redaka matrice sličnosti. Za izračun korisničkih preporuka, podržane su tri različite tehnike matričnog množenja u raspodijeljenoj okolini. Postupak predobrade podataka ostvaren je u programskom jeziku C++, dok su sve funkcionalnosti preporučitelja ostvarene u programskom jeziku Scala, uz korištenje programskog okvira Apache Spark koji omogućava raspodijeljeno izvođenje. Testiranje i evaluacija performanci implementiranog algoritma preporučivanja obavljeni su na računalnom grozdu, korištenjem stvarnih podataka iz odabranog studijskog slučaja.
Abstract (english)
The result of this thesis is a distributed recommender system based on the item-item collaborative filtering. The recommendation algorithm builds an item-item similarity matrix based on the collaboratively collected data on user-item interactions, for all users in the system. The recommendation algorithm supports several similarity measures including a vector normalisation of rows in the matrix. Moreover, the recommendation algorithm supports three different distributed matrix multiplication algorithms. The entire recommender system source code is written in Scala programming language based on Apache Spark. However, the data pre-processing scripts are written in C++ programming language executed in a single-node environment. The tests and performance evaluation of the implemented algorithm were executed on a Cloudera cluster using real dataset obtained from the particular case study.
Keywords
Keywords (english)
Languagecroatian
URN:NBNurn:nbn:hr:168:601260
Study programmeTitle: Computing
Study programme type: university
Study level: graduate
Academic / professional title: magistar/magistra inženjer/inženjerka računarstva (magistar/magistra inženjer/inženjerka računarstva)
Type of resourceText
File originBorn digital
Access conditionsClosed access
Terms of useLicence In copyright icon
Public note
Created on2019-03-14 00:07:09