Tweet-Streamer-Analyser

Web app to show tweets, show WordCloud of tweets, Sentiment analysis of tweets for a given user or #tag

The web app contains both streamer dashboard and analysis dashboad in the same page

Done using flask as backend,

packages used :-

• Tweepy to access twitter API
• tweet-preprocessor for cleaning the tweets
• pycorenlp for sentiment analysis.

To run the web app the server for stanfordcorenlp has to be started

1. Download and unzip StanfordCoreNLP(done in ubuntu):

wget https://nlp.stanford.edu/software/stanford-corenlp-full-2018-10-05.zip https://nlp.stanford.edu/software/stanford-english-corenlp-2018-10-05-models.jar

unzip stanford-corenlp-full-2018-10-05.zip

2. Setup packages for english:

mv stanford-english-corenlp-2018-10-05-models.jar stanford-corenlp-full-2018-10-05

3. Start the server:

cd stanford-corenlp-full-2018-10-05

java -mx6g -cp "*" edu.stanford.nlp.pipeline.StanfordCoreNLPServer -timeout 5000

(change timeout for more data transfer, change -mx6g for amount of memory used by server( 2g, 4g ... )

Start the web app by running web_app.py

The streamer uses jinja input and script to display the tweets

whereas the analyser creates svg to display worcloud and pie chart

Future works:

• Caching the searches and data for quicke display as Twitter API and StanfordCoreNLP take significant time
• Adding more Charts of sentiment analysis such as most used words in negative, positive sentiments
• Adding location in script so as to display the trending #tags locally for the user of webapp
• Adding search output filters such as time limit and number of tweets limit
• Having Different pages for Stremer and Analyser
• Including more images and links in the webpage to better the user experience