- Pipeline is now a `Python` package; - Ordinal classification available; - Massively updated Readme; - Function/class documentation; - Much better structure with folders for depositing data, accessing results etc; - Additional predictors (sentiment scores);
0.1.60000
This release tests RSConnect publishing
0.1.1
0.1.0
First release with the following features:
1. Pre-process data. 2. Build & tune text classification pipeline. 3. Analyse learner performances on test set.
Python code is divided into .py script modules. Modules should be run in the order in `text_mining_run_entire_process.py`.
This version uses a CSV file of the feedback text comments for Nottinghamshire Healthcare NHS Foundation Trust. A future release will handle and pull this and all other data (feedback text from partner trusts) directly from SQL.