- For model prediction
Naive Bayes Classifier
is used. - A dataset is used which contains random files of spam and ham,that uses feature vectorization for processing the data before sending it to model.
- Implementation has been made in
Flask
. - Accuracy -
96%
SPAM | HAM | TOTAL |
---|---|---|
1556 | 1556 | 3112 |
- No biased data is encountered.
- 80% training data 20% testing data.
- Flask Installation Guide
- Once Flask Installation and Virtual Environment setup has been done , execute the following command to start the server locally.
python main.py
- A text file containing the message to be tested has to be uploaded and the result would be outputted.