magistrsko delo
Jan Vidmar (Author), Mirjana Kljajić Borštnar (Mentor)

Abstract

Raziskovalna področja se med seboj razlikujejo tudi po uporabi raziskovalnih strategij, kar se prav tako odraža v študijskih programih. Čedalje več je v uporabi tudi mešanih raziskovalnih strategij, kar ne preseneča, saj se tudi področja med seboj vse bolj prepletajo. Zato smo na primeru povzetkov diplomskih nalog s Fakultete za organizacijske vede želeli ugotoviti, ali je mogoče s pomočjo metod rudarjenja besedil ugotoviti raziskovalno strategijo in jo povezati s študijskim programom. Zbrali smo po 100 povzetkov diplomskih del iz treh osnovnih študijskih programov na fakulteti, torej skupno 300 povzetkov diplomskih nalog. Želeli smo tudi ugotoviti, ali je mogoče iz povzetka diplomske naloge prepoznati raziskovalno strategijo in napovedati, iz katerega programa prihaja diplomska naloga. V ta namen smo uporabili raziskovalno strategijo načrtovanja in razvoja, kot osnovno metodo razvojnega cikla pa smo izbrali CRISP-DM. Izdelali smo Python skripto, ki je omogočila ekstrakcijo povzetkov in jih uredila v urejen korpus. S področja rudarjenja besedil smo uporabili tako nenadzorovane kot nadzorovane metode: metode gručenja, besednega oblaka in napovedovanja razreda. Za uporabo metod rudarjenja besedil smo uporabili orodje Orange Data Mining tool. Ugotovili smo, da lahko iz povzetkov z visoko natančnostjo napovemo študijski program, v katerega sodi posamezna diploma. Rezultati kažejo, da je najpogostejša raziskovalna strategija na področju kadrovskih in izobraževalnih sistemov vzorčna raziskava, medtem ko sta na področju informacijskih sistemov to načrtovanje in razvoj.

Keywords

CRISP-DM;

Data

Language: Slovenian
Year of publishing:
Typology: 2.09 - Master's Thesis
Organization: UM FOV - Faculty of Organizational Sciences
Publisher: [J. Vidmar]
UDC: 004.6
COBISS: 95963651 Link will open in a new window
Views: 99
Downloads: 20
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: Analysis of research strategies in bachelor thesis using text mining methods
Secondary abstract: Research areas differ from each other in the usage of research strategies and the differences reflect in the study programmes as well. Nowadays, the usage of diverse research strategies is increasing as research areas intertwine. The aim of this master thesis was to use the text mining methods in order to identify research strategies in various bachelor thesis from the Faculty of Organizational Sciences and further connect them with the study programme. There were 100 bachelor thesis abstracts collected from three different study programmes, 300 in total. In order to achieve our aim, we used the research strategy of planning and development. CRISP-DM was chosen as the main method of the developmental cycle. We created a Python script, which enabled the extraction of abstracts and ordering them in a structured corpus. Orange Data Mining tool was used for text mining and the following non-controlled and controlled text mining methods: hierarchical clustering, word cloud and class predictions. Our research shows that the study programme can be predicted from the abstract of the bachelor thesis with a high degree of accuracy. Based on the results, the most common research strategy in the study programme of Organization and Management of Human Resources and Educational Systems is survey, whereas in the study programme of Organization and Management of Information System, the most common strategy is planning and development.
Secondary keywords: Podatkovno rudarjenje;Univerzitetna in visokošolska dela;
Type (COBISS): Master's thesis/paper
Thesis comment: Univ. v Mariboru, Fak. za organizacijske vede
Pages: VII, 71 f.
ID: 14134463