Analysis & Classification of Hoax in MAFINDO Dataset

Credits :

1. Yudistira Dwi Cahya
2. Wulan Akhsah
3. Kamal Muftie Yafi
4. Rachel Thyffani Margaretha S
5. Vesya Padmadewi

Set

• Dataset: Hoax dataset obtained from MAFINDO (Masyarakat Anti Fitnah Indonesia);
• Slang: Modified Kamus Alay based on Kamus Alay (Colloquial Indonesian Lexicon);
• Feature Extraction: Bag-of-Words, TF-IDF;
• Classifier: Naive Bayes, SVM, Logistic Regression, Decision Tree, kNN, ANN.
• Cross-Validation: GridSearch, RandomSearch

About The Data

This dataset contained two label values, namely "1" for hoax and "0" for not hoax. The total data in this dataset is 4,701. Each label has a varied amount of data distribution, including 3850 data for hoax and 851 data for not hoax.

Label	Hoax	Not Hoax
Total Data	3850	851

Algorithm included

• Text cleaning/preprocessing
• Non-standard word replacement
• Feature extraction: BoW, TF-IDF
• Classification: Naive Bayes, SVM, Logistic Regression, Decision Tree, kNN, ANN
• Cross-Validation: Grid Search, Random Search
• Post analysis: topicwizard, Voyant Tools, WordCloud