twitter_sentiment_railway

Some useful links--- https://developer.twitter.com/en/docs https://apps.twitter.com/app/14307319/ http://www.geeksforgeeks.org/twitter-sentiment-analysis-using-python/

this is very simple version of our project... Everything is inside railway_feedback.zip file. rest is useless.

1. I've extracted around 20k+ tweets from using tweepy. They are saved in twitDB1.csv file in the data folder.
2. saved the text of those tweets in the tweetText.csv file
3. pre_processed.py then pre_processes them (removing all @,links etc). saved in pre_processed.csv
4. Then classify.py does 2 things -- a. remove all non english words from the tweets(yes there is still a lot of garbage in them) using nltk's wordnet module b. classifies them as tweets related to punctuality, cleanliness etc(just 2 sections till now...we could add more like food) 5.Finally sentiment_analysis.py simply predicts the tweet as positive or negative. So I've printed the % of tweets in each section (cleanliness/punctua..) as + or -

I've used this playlist by sentdex to learn nltk. https://www.youtube.com/watch?v=FLZvOKSCkxY&index=1&list=ac

please give ur suggestions and/or feedback