Inteligentni sustav za uklanjanje glasova sa skrivenim namjerama u online anketama

undergraduate thesis

• Title (english): Intelligent System for Elimination of Hidden Agenda Votes in Online Polls
• Author: Hrvoje Kovačević
• Mentor: Goran Delač (mentor)
• Committee member: Goran Delač (predsjednik povjerenstva)
• Committee member: Marin Šilić (član povjerenstva)
• Committee member: Klemo Vladimir (član povjerenstva)
• Granter: University of Zagreb; Faculty of Electrical Engineering and Computing; Zagreb
• Defense date and country: 2017-07-10, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Computing

Abstract

Problem detektiranja manipulacije mišljenja problem je koji stranice sa glasovima ili recenzijama korisnika moraju riješiti. Ovaj rad se bavi detektiranjem numeričkih glasova koji imaju skrivene namjere. Problem pripada nadziranom učenju te se rješava strojnim učenjem klasifikacijskog modela. Kako bi se to postiglo potrebno je pripremiti podatke, izraditi i testirati modele te potom vrednovati rezultate. Priprema podataka svodi se na razumijevanje, vrednovanje i filtraciju podataka. Prilikom izrade modela potrebno je definirati mjeru uspješnosti modela te izabrati pravi algoritam za učenje modela. Ovaj rad opisuje proces poduzet kako bi se dobio model koji s 90% preciznošću klasificira glasove na ilegalne (sa skrivenom namjerom) i legalne.

Abstract (english)

The problem of opinion manipulation is a common problem that internet sites with vote and review systems need to handle. The goal of this paper is detection of numerically represented votes with hidden agenda. This problem is of supervised learning type and is resolved by training classification model with machine learning algorithms. To achieve that one should prepare data, train and test models and finally validate results. Preparation of data is composed of understanding, validating and filtrating data. In process of training a model one should define performance measure and chose the right machine learning algorithm. This paper describes the process taken that results in model that has a 90% precision in vote classification, where classes are illegal (hidden agenda votes) and legal votes.

• Keywords: manipulacija mišljenja; strojno učenje; proces strojnog učenja; priprema podataka; izrada modela strojnog učenja; unakrsno vrednovanje; knn; SVM; DT; stabla odlučivanja; RF; slučajna šuma; klasifikacija
• Keywords (english): opinion manipulation; machine learning; machine learning process; data preparation; construction of machine learning models; cross validation; knn; SVM; DT; decision tree; RF; random forest; classification
• Language: croatian
• URN:NBN: urn:nbn:hr:168:673488
• 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: 2019-03-13 21:49:41