yxpku / project-lesson-category

text categorization tool to categorize project lessons

Project Lessons Finder Category

The project used machine learning methods to classify the input project evaluation documents into 8 pre-defined lessons categories (46 sub-categories), based on the World Bank IEG project lessons text data.

Models Used

Naive Bayes, Random Forest and ANN

Note for Users

• The model used pretrained text corpus that are saved under the folder data, you can calibrate the text and categories
• To generate output lessons category for your evaluation notes, put the word documents under the folder input_doc
• Run the program main.py, and specify the methods names and the expected number of output categories (8 or 46)

Output

For each input document, the top 3 most probable categories are presented. The results are saved under the output folder

Notice

The sample data is used to preserve the privacy of the project documents