ANALISIS SENTIMEN DAN KLASIFIKASI KATEGORI TERHADAP TOKOH PUBLIK PADA DATA TWITTER MENGGUNAKAN NAIVE BAYES CLASSIFIER
Twitter has been used extensively by various segments of society. Hence, it can be a media that represent trending issues in public. The people habbits of posting tweet can be a refference to find out the public sentiment toward public figures. The needs of sentiment analysis toward public figures a...
| Main Authors: | , |
|---|---|
| Format: | Thesis |
| Published: |
[Yogyakarta] : Universitas Gadjah Mada
2014
|
| Subjects: |
| Summary: | Twitter has been used extensively by various segments of society. Hence, it
can be a media that represent trending issues in public. The people habbits of
posting tweet can be a refference to find out the public sentiment toward public
figures. The needs of sentiment analysis toward public figures are usually
necessary when people want to know the public sentiment and response toward
the public figures especially before elections.
This research analyzes and classifies tweets in Indonesian language using
Naive Bayes Classifier combined with negation detection and term weighting
feature using term frequency and TF-IDF. The classification in this research is
classified based on sentiment about public figure's feature.
Based on the evaluation of the application and RapidMiner tools, the
accuration result shows that the term frequency feature gives better result than that
of TF-IDF. The Support Vector Machine method gives better result than that of
Naive Bayes in classifying tweet. However, both of these methods have the same
good performance to classify tweet. |
|---|