Access restricted to higher education institution's students and staff
master's thesis
Analysis of automotive sales in Croatia by using Google Trends

Dragan Kafadar
Metadata
TitleANALIZA POTRAŽNJE ZA AUTOMOBILIMA U REPUBLICI HRVATSKOJ KORISTEĆI GOOGLE TRENDS
AuthorDragan Kafadar
Mentor(s)Ana Perišić (thesis advisor)
AbstractPrediktivna analitika obuhvaća razne statističke tehnike od prediktivnog modeliranja, strojnog učenja do rudarenja podacima kojima analizira činjenice iz prošlosti i sadašnjosti kako bi napravila predviĎanja budućih odnosno nepoznatih događaja, te je primjenjiva u gotovo svim oblicima poslovanja. Cilj ovoga rada je ispitati mogućnosti korištenja podataka dobivenih putem pretraživačkih upita u svrhu predikcije potražnje za automobilima. Korišteni su podaci dobivenim putem internetske tražilice Google, to jest koristeći alat Google Trends. Google Trends je besplatni online alat pomoću kojega korisnik na jednostavan način može vidjeti što ljudi pretražuju na internetu u određenom vremenskom razdoblju. Analizirano je razdoblje od 2005. do 2015. godine u Republici Hrvatskoj, gdje su korišteni mjesečni podaci o novoregistriranim automobilima i mjesečni podaci o pretraživačkim upitima dobivenih putem internetske tražilice Google. Pokazano je kako pretraživački upiti i potražnja za automobilima imaju sličnu dinamiku.
Keywordsprediction Google Trends automotive sales demand
Parallel title (English)Analysis of automotive sales in Croatia by using Google Trends
Committee MembersDivna Goleš (committee chairperson)
Ivana Beljo (committee member)
Ana Perišić (committee member)
GranterVeleučilište u Šibeniku
PlaceŠibenik
StateCroatia
Scientific field, discipline, subdisciplineSOCIAL SCIENCES
Economics
Organization and Management
Study programme typeprofessional
Study levelspecialist graduate
Study programmeManagement
Academic title abbreviationSruč.spec.oec.
Genremaster's thesis
Language Croatian
Defense date2016-09-06
Parallel abstract (English)Predictive analytics encompasses various statistical techniques from predictive modeling, machine learning, and data mining that analyze both current and historical facts in order to predict future events. Predictive analytics can be applied to almost any type of business. The aim of this paper is to examine the possibility of using data obtained through search queries in order to predict automotive sales. The data were obtained through internet search engine Google, using Google Trends tool. Google Trends is a free online tool which shows what people search on the internet in a specific time period. The analysis was performed for the period from 2005 to 2015. Monthly data of newly registered cars in Croatia and monthly data of search queries through internet search engine Google were obtained. The results confirmed similar dynamics of search queries and automotive sales.
Parallel keywords (Croatian)predikcija Google Trends automobili potražnja
Resource typetext
Access conditionAccess restricted to higher education institution's students and staff
Terms of useAll rights reserved
URN:NBNhttps://urn.nsk.hr/urn:nbn:hr:143:359338
CommitterNives Paškov Milošević