/fds

Primary LanguagePython

How to

  • Unzip merged-cleansed.7z in the root folder as merged-cleansed.json.
  • Run the script topic_analysis.py.
  • Once completed it will ouput LDA model and will open HTML view on http://localhost:8888.

Files

  • merged-cleansed.7z - Formatted largely cleansed twitter dataset. When unpacked it would be around ~400mb.
  • /analysis/ - Files relevant to analysis.
  • /analysis/models/ - Results from previous analysis.
    • meta - Containers metadata / params used to run the model.
  • /analysis/custom-stopwords.txt - list of custom stop words.