Sustav za davanje kontekstualiziranih preporuka na temelju rudarenja teksta

master's thesis

• Author: Lena Kamenjaš
• Mentor: Luka Grubišić (mentor)
• Committee member: Goran Muić (predsjednik povjerenstva)
• Committee member: Josip Tambača (član povjerenstva)
• Committee member: Luka Grubišić (član povjerenstva)
• Committee member: Marcela Hanzer (član povjerenstva)
• Granter: University of Zagreb; Faculty of Science; (Department of Mathematics); Zagreb
• Defense date and country: 2018-11-28, Croatia
• Scientific / art field, discipline and subdiscipline: NATURAL SCIENCES; Mathematics

Abstract

U ovom je radu predstavljen jedan način kako poboljšati pretraživanje po dokumentima pisanim prirodnim jezikom - otkrivanjem ključnih riječi dokumenata. Ukratko se priča o obradi prirodnog jezika, važnoj disciplini kod analize dokumenta. Zatim se priča o procesu otkrivanja ključnih riječi i podjeli metoda. Detaljnije se obrađuju metode korištene pri izradi aplikacije: TextRank i $tf-idf$ algoritam. Prije opisa implementacije, navode se i ukratko opisuju korišteni alati i tehnologije za izradu... More aplikacije. Zatim se predstavlja postupak izrade aplikacije, koji se sastoji od pretprocesiranja, primjene algoritama (osnovna dva i njihove tri modifikacije) te postprocesiranja. Naposljetku se navodi usporedba rezultata te primjer dokumenta i nađenih ključnih riječi. Less

Abstract (english)

This thesis presents one way of improving the process of searching documents written in natural language - by discovering keywords. It starts with a brief description of natural language processing, a sub-field of computer science, information engineering, and artificial intelligence that is very important for text analysis. Next chapter presents keyword extraction and its classification of methods. Two methods used in application are discussed in detail: TextRank and $tf-idf$ algorithm.... More Before describing the implementation process, a list and a short description of used tools and technologies in application is given. Then follows the description of implementation process, which consists of pre-processing, application of algorithms (two basic and their three modifications) and post-processing. Finally, comparison between used methods and an example is given. Less

• Keywords: pretraživanje dokumenata; ključne riječi; obrada prirodnog jezika; rudarenje teksta
• Keywords (english): searching documents; keywords; natural language processing; text mining
• Language: croatian
• URN:NBN: urn:nbn:hr:217:555967
• Study programme: Title: Computer Science and Mathematics; Study programme type: university; Study level: graduate; Academic / professional title: magistar/magistra računarstva i matematike (magistar/magistra računarstva i matematike)
• Type of resource: Text
• File origin: Born digital
• Access conditions: Open access
• Created on: 2019-05-16 09:08:26