Implementacija funkcionalnosti analize i usporedbe dokumenata

undergraduate thesis

• Title (english): Implementation of document analysis and comparison functionalities
• Author: Klara Zagajski
• Mentor: Marina Bagić Babac (mentor)
• Committee member: Marina Bagić Babac (predsjednik povjerenstva)
• Committee member: Mihael Kovač (član povjerenstva)
• Committee member: Kristijan Poje (član povjerenstva)
• Granter: University of Zagreb; Faculty of Electrical Engineering and Computing; Zagreb
• Defense date and country: 2024-09-06, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Computing

Abstract

Ovaj rad bavi se razvojom automatiziranog sustava za prepoznavanje teksta, ekstrakciju informacija, klasifikaciju i usporedbu dokumenata. Sustav obrađuje različite formate dokumenata i koristi tekstualne ugradnje (TF-IDF, Word2Vec, Doc2Vec, GloVe) i modele strojnog učenja (Linear SVM, K-Nearest Neighbors, Multinomial Naive Bayes) za poboljšanje točnosti klasifikacije. Evaluiran na skupu podataka sačinjenog od životopisa s Kaggle-a koji sadrži 22 klase, linearni SVM s TF-IDF-om postigao je najbolje rezultate. Ovaj rad nudi uvid u buduća unapređenja automatizirane analize dokumenata.

Abstract (english)

This thesis develops an automated system for text recognition, information extraction, document classification and document comparison. The system handles various document formats and uses text embeddings (TF-IDF, Word2Vec, Doc2Vec, GloVe) and machine learning models (Linear SVM, K-Nearest Neighbors, Multinomial Naive Bayes) to enhance classification accuracy. Evaluated on a Kaggle-sourced resume dataset with 22 classes, the Linear SVM with TF-IDF achieved the best performance. This work offers insights for future advancements in automated document analysis.

• Keywords: analiza dokumenata; prepoznavanje teksta; ekstrakcija informacija; klasifikacija dokumenata; tekstualne ugradnje; strojno učenje; TF-IDF; Word2Vec; Doc2Vec; GloVe; Linear SVM; K-Nearest Neighbors; Multinomial Naive Bayes; automatizirana obrada dokumenata
• Keywords (english): Document Analysis; Text Recognition; Information Extraction; Document Classification; Python; Text Embeddings; Machine Learning; TF-IDF; Word2Vec; Doc2Vec; GloVe; Linear SVM; K-Nearest Neighbors; Multinomial Naive Bayes; Automated Document Processing
• Language: english
• URN:NBN: urn:nbn:hr:168:302520
• Study programme: Title: Computing; Study programme type: university; Study level: undergraduate; Academic / professional title: sveučilišni/a prvostupnik/ prvostupnica (baccalaureus/baccalaurea) inženjer/inženjerka računarstva (sveučilišni/a prvostupnik/ prvostupnica (baccalaureus/baccalaurea) inženjer/inženjerka računarstva)
• Type of resource: Text
• File origin: Born digital
• Access conditions: Closed access
• Created on: 2024-10-12 22:31:24