SentimentAnalysis

Python project: Analysis of sentiments from twiter (and perhaps other sources later)

Author: David Vincent

This project will attempt to compare happiness of terms in tweets.

What you need:

1. list of terms
2. list of sentiments with sentiment values

Demo walkthrough tutorial:

This tutorial is comparing windows, mac, linux

Steps: 1) create list, download sentiment list 2) download tweets 3) analize the tweets 4) review the data

1) To create the list start a new text file called termList.txt with follwoing lines:
    windows
    mac
    linux
   Download AFINN-111.txt from http://www2.imm.dtu.dk/pubdb/views/publication_details.php?id=6010
2) To download the tweets run the following command:

  python termDownload.py termList.txt > termData.txt
  
  The prevous command will save the tweet data in termData.txt, and output the progress to the screen.
  Wait unitl enough tweets have been downloaded then press ctl-c
  
3) to analize the tweets run the following command
  
  python happiestTerm.py AFINN-111.txt termList.txt termData.txt
  
  after the program finishes you should recieve the output in Decending order. 
  here is a sample output run on ~880 thousand tweets
  
  Processing Item [884989]...
  Terms   Score	  Count	  PercentCount
  **************************************
  windows	0.7714	115646	43.0
  mac	    0.6085	85083	  32.0
  linux	  0.398	  62666	  23.0
  
4) Review the data:

  Upon examining the data it looks like windows has the most happy results on average, however,
  there was about twice as many tweets in the data set about microsoft than there was about linux,
  so to make a better estimate, a paramater containing the max number of tweets to capture per term
  can be added when the program is executed. 
  
  For example to only allow 62500 tweets to be analized,
  run the following command
  
  python happiestTerm.py AFINN-111.txt termList.txt termData.txt 62500
  
  which on the dataset used in this example produced the following output:
  
  Processing Item [882725]...
  Terms    Score	Count	PercentCount
  **************************************
  windows	0.6712	62500	33.0
  mac	    0.5392	62500	33.0
  linux	  0.3927	62500	33.0
  
  Examining this data, it still shows that windows is regaurded as more happy than linux and mac, but the 
  score is a bit smaller than prevously
  
  So it looks like people that post to twitter are more happy about windows than they are about mac 
  and linux.
  
  Thanks for reading, and Happy Hacking!