Klasifikacija tekstualnih dokumenata

master's thesis

• Title (english): Classification of text documents
• Author: Kristina Bukvić
• Mentor: Pavle Goldstein (mentor)
• Committee member: Pavle Goldstein (predsjednik povjerenstva)
• Committee member: Zoran Vondraček (član povjerenstva)
• Committee member: Nela Bosner (član povjerenstva)
• Committee member: Eduard Marušić-Paloka (član povjerenstva)
• Granter: University of Zagreb; Faculty of Science; (Department of Mathematics); Zagreb
• Defense date and country: 2021-09-22, Croatia
• Scientific / art field, discipline and subdiscipline: NATURAL SCIENCES; Mathematics

Abstract

Cilj rada je klasifikacija tekstualnih dokumenata pomoću algoritma stroja potpornih vektora i n-gramskog rječnika. Prvo smo definirali pojmove iz linearne algebre i optimizacije koji su potrebni za razumijevanje obrađenih tema u radu. Objašnjen je pojam klasifikacije teksta te pristupi: “Bag of words” i “Bag of n-grams”. Također je opisana matematička pozadina algoritma stroja potpornih vektora te njegova implementacija u programskom jeziku Python. Za analizu smo koristili dvije kolekcije tekstualnih dokumenata, pri čemu smo trening i test skup podijelili u omjeru 50 : 50. Nakon što smo napravili rječnik n-grama, proveli smo testiranje te smo usporedili rezultate klasifikacije za navedene pristupe.

Abstract (english)

The aim of this work is to classify text documents using the support vector machine algorithm and the n-gram dictionary. First, we present basic concepts from linear algebra and optimization theory required for the understanding of topics covered in the paper. Also, we explained the concept of text classification and approaches: “Bag of words” and “Bag of n-grams”. Next, the mathematical background for support vector machine algorithm is described with its implementation in Python programming language. We used two sets of text documents for analysis, dividing the training set and test set into 50 : 50 ratio. After creating the n-gram dictionary, we conducted the test and compared the classification results for the approaches listed above.

• Keywords: algoritam stroja potpornih vektora; n-gramski rječnik; programski jezik Python
• Keywords (english): support vector machine algorithm; n-gram dictionary; Python programming language
• Language: croatian
• URN:NBN: urn:nbn:hr:217:754302
• Study programme: Title: Mathematical Statistics; Study programme type: university; Study level: graduate; Academic / professional title: magistar/magistra matematike (magistar/magistra matematike)
• Type of resource: Text
• File origin: Born digital
• Access conditions: Open access
• Created on: 2021-10-25 12:19:19