Primjena rekurentnih neuronskih mreža za klasifikaciju tekstualnog sadržaja

master's thesis

• Title (english): Applying Recurrent Neural Networks for Classifying Textual Content
• Author: Lovro Filipović
• Mentor: Marin Vuković (mentor)
• Committee member: Marin Vuković (predsjednik povjerenstva)
• Committee member: Dragan Jevtić (član povjerenstva)
• Committee member: Boris Vrdoljak (član povjerenstva)
• Granter: University of Zagreb; Faculty of Electrical Engineering and Computing; Zagreb
• Defense date and country: 2018-07-05, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Computing

Abstract

U ovom radu opisane su umjetne neuronske mreže, s naglaskom na rekurentne neuronske mreže, sa njihovim karakteristikama i dijelovima. Obrađeni su pristupi i zadaci područja prirodne obrade jezika, posebno zadatak klasifikacije teksta te kako se rješava taj zadatak koristeći rekurentne neuronske mreže. Implementirana je rekurentna neuronska mreža u radnim okolinama TensorFlow i Keras te je primijenjena na klasifikaciju sentimenta skupa podataka Amazon Reviews i klasifikaciju tekstualnog sadržaja s obzirom na tip osobnosti autora. Na kraju su predočeni i analizirani dobiveni rezultati.

Abstract (english)

This article describes artificial neural networks, with emphasis on recurrent neural networks and their characteristics and parts. The approaches and tasks of natural language processing are elaborated, in particular, the task of texts classification and how to solve this task using recurrent neural networks. A recurrent neural network is built using TensorFlow and Keras frameworks and is applied to the sentiment classification of Amazon Reviews dataset and classification of the textual content with regard to the type of author's personality. Finally, the results obtained were presented and analyzed.

• Keywords: neuronske mreže; rekurentne mreže; obrada prirodnog jezika; klasifikacija teksta; tip osobnosti; sentiment
• Keywords (english): neural networks; recurrent networks; natural language processing; text classification; MBTI; sentiment
• Language: croatian
• URN:NBN: urn:nbn:hr:168:218708
• 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: 2019-03-13 20:13:53