ePrints.FRI - University of Ljubljana, Faculty of Computer and Information Science

Sentiment analysis on rtvslo.si user comments

Brina Škoda (2013) Sentiment analysis on rtvslo.si user comments. EngD thesis.

Download (1655Kb)


    The main purpose of this thesis was getting to know the field of opinion mining on texts in Slovene language in greater detail. We then used the acquired knowledge in connection with a practical example. For the data we used the news and comments from the rtvslo.si news portal, which is one of the two most popular portals in Slovenia. We tested various algorithms trying to classify comments as positive or negative with the greatest possible accuracy, then we assumed our results as correct and did some more opinion mining on them. At the beginning, we have set ourselves the basic goals which we then upgraded with acquired knowledge. There has not been much done in this field in Slovene language, consequently we came across a great deal of problems. Firstly, we studied the theory and presented it, so the further work is more understandable to the reader. We presented the used methods, findings and results.

    Item Type: Thesis (EngD thesis)
    Keywords: opinion mining, sentiment analysis, natural language processing, data mining, machine learning, SVM, maximum entropy, naive Bayes, classification, gold standard, features
    Number of Pages: 57
    Language of Content: Slovenian
    Mentor / Comentors:
    Name and SurnameIDFunction
    doc. dr. Dejan Lavbič302Mentor
    Link to COBISS: http://www.cobiss.si/scripts/cobiss?command=search&base=50070&select=(ID=10128724)
    Institution: University of Ljubljana
    Department: Faculty of Computer and Information Science
    Item ID: 2138
    Date Deposited: 13 Sep 2013 09:52
    Last Modified: 23 Sep 2013 12:21
    URI: http://eprints.fri.uni-lj.si/id/eprint/2138

    Actions (login required)

    View Item