Model should correctly classify the sentiment in the text.
-
Data : It has collected customer review of the product using the real e-commerce website like flipkart, amazon etc.
-
Data Preprocessing : Since data can only understand the numerical data, It is necessary to convert text data to numerical value. It has done using standard methods like lemmatization, stemming from nltk library alternative libraries like fast.ai are also available.
-
Modelling : Train the classifier multiple sentiments (positive, negative and neutral). Implement the naive bayes algorithm and using scikit learn library.
-
Visualization:
-
User Interface : using flask framework