This project attempts to assist educators in analysing the sentiment of Malay social media posts. The output from the sentiment can be used to enhance their teaching and learning activities. In this project, training and testing data was acquired from Husein in 2018, the Malay Stopwords List that used in data preprocessing stage was based on the research of Fatimah Ahmad (1995). All datasets need to be prepared using preprocessing, including tokenization, stop word removal, lower casing, removing numbers, and removing punctuations. Then the TF-IDF vectorization method was used. In this project, we implemented Support Vector Machine (SVM). The performance of trained models were evaluated using Confusion Matrix and Evaluation Matrix. From the experiment this project tends to produce 93% accuracy, 92% for prediction and 92% for recall...

218 hits