- Import Sql database using sqlite3 library python.
- Seperate data based on Score.
- assign label to data positive negative depending on the score.
- Do text cleaning removing punctuation, removing html tags, removing stopwords, stemming, removing alphanumeric words.
- parse the cleaned data to numeric value such as array.
- apply cleaned data to the naive bayes classifier.
- make count vectorizer or bag of words to store the keywords.
- split data into train and test of 70 and 30 %.
- train the data using the NB classifier.
- Save the classified model using pickle which is again a python library.
- Predict the data from the new derived final.sqlite(cleaned data after preprocessing) file which was generated after training.
- Integrate with flask api and check the output with postman.