Skip to content

Latest commit

 

History

History
35 lines (32 loc) · 1.13 KB

README.md

File metadata and controls

35 lines (32 loc) · 1.13 KB

TwitterAnalysis

CREATING SQL DATABASE

  1. Change line in sqlscript.py (line 60) db_con = MySQLdb.Connect(, , , 'tweets')
  2. Run sqlscript.py This will create a database tweets with different tables

CREATING STOP WORDS

  1. utility/create_stop_words.py is run stand alone to generate stop words using idf
  2. This script has the training dataset hard coded. Can be changed for a new dataset
  3. Produces file Lists/StopWordsIDF.txt (don't have to recreate if it exists)

RUNNING THE CODE

  1. Change collector.config. Query terms are also given in this file. Currently query has to be done manually by running Collector.py

  2. Once query is done, run sentinal.py (with updated SQL info)

TO DO

  1. Stemming
  2. Pivoting
  3. Bi-grams
  4. Confidence of prediction
  5. Neutral class
  6. Max entropy classifier
  7. MI of feature (and feature selection based on this)