diplomsko delo visokošolskega študijskega programa Informacijska varnost
Benjamin Steiner (Author), Simon Vrhovec (Mentor)

Abstract

V Sloveniji je letno ukradenih med 900 in 1100 vozil, ampak policiji uspe izslediti le 25% teh vozil. Trenutno ne obstaja storitev, preko katere lahko poljubno vozilo primerjamo s seznamom ukradenih vozil in izluščimo le najbolj podobna vozila, ali posledično najdemo ukradeno vozilo. Za naslovitev problema je bil ustvarjen model, ki s pomočjo uporabe algoritma za iskanje najbližjih sosedov išče ukradena vozila, ki so najbolj podobna poljubno izbranemu vozilu. Da bi testirali razviti model, smo izbrali 200 vozil za vsako izmed 11 najpogostejših znamk avtomobilov na spletni strani Avto.net. Razdalje med avtomobili smo razvrstili v 3 opisne stopnje glede na podobnost: zelo podobna vozila, podobna vozila in malo podobna vozila. Opisne stopnje podobnosti so ustrezno merilo za ocenjevanje podobnosti vozil, saj so mejne razdalje med vozili jasno določene in neodvisne od znamke avtomobilov.

Keywords

podatki;javno dostopni podatki;rudarjenje podatkov;ukradena vozila;diplomske naloge;

Data

Language: Slovenian
Year of publishing:
Typology: 2.11 - Undergraduate Thesis
Organization: UM FVV - Faculty of Criminal Justice
Publisher: [B. Steiner]
UDC: 004.8(043.2)
COBISS: 85463299 Link will open in a new window
Views: 241
Downloads: 48
Average score: 0 (0 votes)
Metadata: JSON JSON-RDF JSON-LD TURTLE N-TRIPLES XML RDFA MICRODATA DC-XML DC-RDF RDF

Other data

Secondary language: English
Secondary title: Use of publicly available data to assess the similarity of vehicles with stolen ones
Secondary abstract: In Slovenia, between 900 and 1100 vehicles are stolen every year, but the police only manage to recover about 25% of the stolen vehicles. Currently, there is no service that enables a comparison between the selected and the stolen vehicle and extracts only the most similar vehicles, or consequently finds the stolen vehicle. To solve this problem, a model was created. In the model, we used Nearest Neighbours algorithm to search for stolen vehicles that are most similar to the selected vehicle. To test the developed model, we selected 200 vehicles for each of the 11 most popular car brands on the Avto.net website. The distances between vehicles were classified into 3 descriptive levels depending on their similarity: very similar vehicles, similar vehicles, and slightly similar vehicles. Descriptive similarity levels are a suitable criterion for assessing vehicle similarity because the distances between vehicles are clearly defined and independent of car brand.
Secondary keywords: Python;Orange;stolen vehicles;finding nearest neighbours;
Type (COBISS): Bachelor thesis/paper
Thesis comment: Univ. v Mariboru, Fak. za varnostne vede, Ljubljana
Pages: VI f., 27 str.
ID: 13729658