A carregar...
Publicação
Exploring patterns and trends of master dissertations : a text mining application
| Nome: | Descrição: | Tamanho: | Formato: | |
|---|---|---|---|---|
| 2.32 MB | Adobe PDF |
Autores
Orientador(es)
Resumo(s)
Data mining is a technique that is used to discover trends and patterns in a dataset (Feldman et al., 1998). However, the majority of datasets are unstructured and contain textual data (Sulova & Nacheva, 2017). Therefore, in the last decade, different text mining techniques are developed to gain competitive advantage by extracting information from these datasets. However, text mining techniques in the educational context is still scarce (Nie & Sun, 2017). Therefore, this study applied text mining techniques on master dissertations to explore trends and patterns of these studies over the years. Dissertations that were published at NOVA IMS from March 2004 until May 2018 were collected from its repository. Thereafter, different techniques were applied to pre-process the data. More specific, unevaluable characters and stop-words were removed from the dataset and stemming was applied to reduce noise and the amount of unique words. Thereafter, two clustering techniques were applied. First, Topic modelling was executed with an LDA algorithm which suggested four topics. After inferring the topics, the following definitions were indicated: Geodata Information, Behavioural Studies, Information and Decision Systems, and Implementing Systems. Thereafter, clustering k-means was executed which resulted in three clusters. The clusters indicated Geodata Analysis, Online Behaviour, and Business Analysis. Furthermore, trends and patterns were analysed for each study area (i.e. Advanced Analytics, Statistics and Information Management, Information Management, Geospatial Technologies and Geographic Information Systems and Science). The results of this study create opportunities for NOVA IMS among others to reallocate resources based on these trends (Sullivan, 2001) and to allocate financial resources based on current patterns (Simoudis, 1996) and to encourage students to examine new interesting trends (Feldman et al., 1998).
Descrição
Project Work as partial requirement for obtaining the Master’s degree in Information Management, with a specialization in Information Analysis and Management
Palavras-chave
Dissertations Text Mining Knowledge Trends Patterns Cluster Topic Modelling