/SentiMiner

Analyzing Cashtag tweets for Sentiment and make predictions about the stock market using multi class logistic regression.

Primary LanguagePython

#SentiMiner

To extract stock information, run StockMining.py from the terminal

To execute, run mining.py from the terminal

By default, the program will listen indefinitely for incoming tweets with the keyword(s) specified in mining.py

The output will be a csv file. It will be named .csv CSV file information is returned in the following order: ID, text, isVerified, #retweets, #favorites, #followers, sentiment, #capitals, # of !, # of ?, #emoticons

  1. How to set a tweet limit before disconnecting:

    • In mining.py, go to the line which instantiates MineListener
      • listener = MineListener()
    • Add an argument tweet_limit
      • ex: listener = MineListener(tweet_limit = 10)
    • Now after 10 tweets are gathered and sent to the csv file, the stream will disconnect
    • Similarly to the above, one can override the path of the file to which the data will be sent
      • ex: listener = MineListener(csv_path = 'myFile.csv')

  2. How to specify what keywords to search for in mining.py as well

    • In mining.py, go to the line where the stream starts running
      • stream.filter(track = ["party"])
    • Add items to the list which track is set to
      • ex: stream.filter(track = ["party","feelings","fun"])
    • Now all the specified keywords will be included in the search

  3. How to extract stock information:

    • In StockMining.py, change stockTicker to desired stock symbol/ticker
    • Change intervalInSeconds to desired time interval in seconds
    • Change periodInDays to number of days that you wish to extract the stock info of
    • To write the extracted data to a CSV file, uncomment the final two lines