magistrsko delo
Tina Avbelj (Author), Tomaž Curk (Mentor)

Abstract

Matrična faktorizacija je metoda za zlivanje podatkov, ki jo lahko uporabimo za priporočilne sisteme. V magistrski nalogi smo se ukvarjali s priporočilnim sistemom iz domene turizma, ki uporabnikom priporoča turistična doživetja. Turistični ponudniki za predstavitev svoje ponudbe izberejo slike in opise. Naš cilj je bil izbrati take slike in opise, ki najbolje odražajo oceno ponudnika. Uporabljali smo sintetične podatke o uporabnikih in njihovih ocenah ponudnikov ter implementirali generator slik in opisov. S klasifikacijskimi algoritmi smo iz množice slik in besedil odstranili šum in izbrali slike in besedila, ki najbolje odražajo ponudnikovo povprečno oceno. Ustreznost izbranih primerov smo vrednotili tako, da smo jih uporabili za napovedovanje ocen z matrično faktorizacijo. Uspešnost napovedi smo primerjali z napovedmi matrične faktorizacije z uporabo naključno izbranih slik in besedil ter matrično faktorizacijo brez stranskih virov. Izkazalo se je, da uporaba izbranih slik in besedil ne izboljša napovedi v primerjavi z naključno izbranimi slikami ali besedili in matrično faktorizacijo brez dodatnih virov. Uporabili smo tudi zbirko realnih podatkov s spletnimi slikami in opisi nastanitev. Iz slik in opisov smo z algoritmoma k-najbližjih sosedov in naključni gozdovi napovedovali ocene nastanitev. Izkazalo se je, da iz podatkov nismo izluščili dovolj vzorcev, da bi z uporabljenimi pristopi lahko napovedali ocene na podlagi slik in opisov turističnih ponudnikov.

Keywords

priporočilni sistemi;matrična faktorizacija;zlivanje podatkov;računalništvo;računalništvo in informatika;magisteriji;

Data

Language: Slovenian
Year of publishing:
Typology: 2.09 - Master's Thesis
Organization: UL FRI - Faculty of Computer and Information Science
Publisher: [T. Avbelj]
UDC: 004:338.48(043.2)
COBISS: 41053955 Link will open in a new window
Views: 915
Downloads: 128
Average score: 0 (0 votes)
Metadata: JSON JSON-RDF JSON-LD TURTLE N-TRIPLES XML RDFA MICRODATA DC-XML DC-RDF RDF

Other data

Secondary language: English
Secondary title: Recommending photos and descriptions to tourism providers
Secondary abstract: Matrix factorization is a method for data fusion that can be used for recommender systems. In the thesis, we implemented a recommender system for recommending experiences to tourists. Tourism providers promote their experiences with images and descriptions. Our goal was to choose images and descriptions that best reflect a provider's rating. We used synthetic data for users and ratings and implemented a data generator for images and descriptions. We removed noise from the data using k-nearest neighbors algorithm and for each provider selected one image or one description that represents the provider's rating. The selected images and descriptions were used as a source for matrix factorization for predicting ratings. We compared our process to matrix factorization with no additional sources and to matrix factorization where we chose random images and descriptions and used them as a source. Our process with selected images and descriptions did not improve the AUC score compared to selecting random images and descriptions and using no additional sources. We also tested on online images and descriptions of real accommodations. We used k-nearest neighbors and random forests algorithms to predict average ratings for accommodations from images and descriptions. We concluded that with the methods we used, we were not able to extract patterns from data, which would enable the prediction of ratings based on images and descriptions.
Secondary keywords: recommendation systems;matrix factorization;data fusion;computer science;computer and information science;master's degree;
Type (COBISS): Master's thesis/paper
Study programme: 1000471
Embargo end date (OpenAIRE): 1970-01-01
Thesis comment: Univ. v Ljubljani, Fak. za računalništvo in informatiko
Pages: 74 str.
ID: 12195568