1. Data pre-processing - Remove Blank rows in Data, Change all the text to lower case, Word Tokenization, Remove Stop words, Remove Non-alpha text, Word Lemmatization (here have to use all the grams including bygram)
2. cross-validation using traing and testing data set
3. word vectorization with TFIDf
4. attribute reduction with fuzzy-rough quick reduct algorithm
5. Fuzzy rough nearest neighbour algorithm(FRNN) for the classification of tweets to three classes of positive - 1, negative - 0 and neutral - 2.
5. Need the confusion matrix, precision, recall, f1-score, support and accuracy
6. Classify another data set using the developed classifier as whether the tweets are positive, negative or neutral (This is expecting to change the label value in the csv file under the label column)
7. Need all the reports and comments where necessary in the code
8. Do the project in Python using jupyter notebook
I attached the labeled data set to build and test the classifier and data set to be classified separetely
About the recuiterMember since Nov 11, 2022 Haroun Saleh
from Eastern, Uganda