seniment-analysis-using-sentimnt-words-imdb

sentiment analysis on imdb data set with a deferent aproach the imdb data set is of most of the hollywood movies and the reviews are aslo by most of the USA citizens. thats why this model become biesed towards wastern culture and their accent of speaking and writing.

ini this model i have used some sentiment lexicons by indian commumnity that desides the sentence to be positive or negative

sentiment lexicons

sentiment lexicons are some bag of words that has both positive and negative labled words you can see that in positive words.txt and negative words.txt files

using these lexicons and then making the model

HOW TO USE THIS REPO?

step 1:

set up internet connection and your environmentshaving following dependencies

1. tensorflow
2. numpy
3. h5py
4. maths
5. keras please do install others if you need

step 2:

Run Sentiment_Analysis_training_new.py file first

step 3:

now Run Sentiment_Analysis_new.py file to get result

Result

the result will be the array of two elements

first element of the array will be showing the pobability of being the sentence or review negative

the second element of array is the probability of being the sentence or review positive

the Sentiment_Analysis_new.ipynb file is the python noteboomk developement file for this project