Darja Eržen (2012) A prototype of recommendation system for higher education course selection. EngD thesis.
The primary goal of this was to develop a prototype of a course recommendation system to provide decision support for the students and help them in their choice of curricula. In this, we imitated a standard on-line recommender system that serves customers by recommending them what product to buy. Two different kinds of recommendation systems were conceived. The first kind was based on mining of the textual description of the courses. Here, the students are advised to take courses whose descriptions are similar. As an alternative, we have studied recommendation systems based on collaborative filtering by proposing a selection of courses considered by students with similar course profiles. Different recommendation systems that were prototyped in our work were thoroughly evaluated on the data of course selected by students from the Faculty of Computer and Information Science. The study shows that system based on collaborative filtering gives more accurate advises than text mining system. This is likely due to the fact that students like to visit courses from various fields.
|Item Type: ||Thesis (EngD thesis)|
|Keywords: ||recommender system, text mining, collaborative filtering, TF-IDF weights, Jaccard similarity coefficient, Dice's coefficient, cosine similarity, lemmatisation|
|Number of Pages: ||59|
|Language of Content: ||Slovenian|
|Mentor / Comentors: |
|Name and Surname||ID||Function|
|prof. dr. Blaž Zupan||106||Mentor|
|Link to COBISS: ||http://www.cobiss.si/scripts/cobiss?command=search&base=50070&select=(ID=00009258580)|
|Institution: ||University of Ljubljana|
|Department: ||Faculty of Computer and Information Science|
|Item ID: ||1716|
|Date Deposited: ||30 May 2012 22:53|
|Last Modified: ||04 Jul 2012 09:22|
Actions (login required)