Detekcija registracijskih formi za prikupljanje podataka o korisnicima

master's thesis

• Title (english): Detection of registration forms that collect user data
• Author: Krešimir Bačurin
• Mentor: Marin Vuković (mentor)
• Committee member: Marin Vuković (predsjednik povjerenstva)
• Committee member: Dragan Jevtić (član povjerenstva)
• Committee member: Krešimir Pripužić (član povjerenstva)
• Granter: University of Zagreb; Faculty of Electrical Engineering and Computing; Zagreb
• Defense date and country: 2023-07-11, Croatia

Abstract

Detekcija anomalija u registracijskim podacima
U ovom radu raspravljalo se o konceptu anomalija u analizi podataka, zašto su anomalije
važne te njihova primjena u stvarnom životu. Rad je usredotočen na detekciju anomalija u
registracijskim podacima koji se pojavljuju na web stranicama.
Kako bi algoritmi strojnog učenja za detekciju anomalija bili efikasni, opisani su i
primijenjeni principi pretprocesiranja podataka. Smanjen je „šum“ i poboljšana je kvaliteta
skupa podataka kako bi algoritmi mogli pružiti dobre rezultate.
Razvijena je web aplikacija koja korisnicima omogućava unos registracijske stranice na
temelju koje se korisnika obavještava traži li unesena stranica neuobičajene podatke ili pak
one sasvim regularne. Korisnik također ima uvid u učestalost pojave najčešćih podataka
koji se traže od korisnika pri registraciji.

Abstract (english)

Detection of anomalies in registration data
This paper discussed the concept of anomalies in data analysis, why anomalies are
important and their use in real life. The work is focused on the detection of anomalies in
registration data that appear on web pages.
In order for machine learning algorithms for anomaly detection to be more effective, the
principles of data preprocessing are described and applied. The "noise" was reduced and
the quality of the data set was improved so that the algorithms could provide good results.
A web application was developed that allows users to enter URL of registration page,
based on which the user is informed whether the entered page is asking for unusual data or
completely regular data. The user also has an insight into the frequency of occurrence of
the most common data requested from the user upon registration.

• Keywords: Web pauk; registracijska forma; detekcija anomalija; strojno učenje
• Keywords (english): Web spider; registration form; anomaly detection; machine learning
• Language: croatian
• URN:NBN: urn:nbn:hr:168:053088
• 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: 2023-10-11 08:19:18