Classification of Twitter Data Belonging to Sudanese Revolution Using Text Mining Techniques

Huda Jamal Abdelhameed, Susana Muñoz- Hernández

Twitter has attracted a great deal of attention recently. It is one of the most common social networking sites for chats, sharing ideas, and transfer of information and news through text. This paper focused on sentiment classification of twitter data belonging to Sudanese revolution which written using either Modern Standard Arabic or Sudanese dialectical Arabic. Twitter’s API was used to collect tweets related to Sudanese revolution. The dataset consists of 6482 tweets with a good balance of positive and negative sentiments. Three different classifiers were used on the dataset namely; Support Vector Machine (SVM), Naive Bayes (NB) and Decision Tree (DT) to classify the tweets based on its polarity into positive or negative. We evaluated our work by four different measures which are Precision, Recall, Accuracy and F-measure. We made a comparison between the three classifiers based on those measures. The results show that, SVM achieved the best Accuracy and F-measure and it equals 75.2%, 83.9% respectively. While NB achieved best Precision and it equals 75.2%. Also, DT achieved best Recall and it equals 99.9%. In addition, the percentages of positive and negative opinions toward the government was calculated. 9.4% represents the percentage of positive opinions related the government, while 90.6% represents the percentage of negative opinions related the same government.

https://doi.org/10.31871/IJNTR.5.7.15