Y-Classifier

Y-Classifier is an application for automatic text document classification. It allows the user finding to what class (genre, thematic category) a text document belongs. The application employs an original algorithm developed by V.A. Yatsko that is based on deviations of stop-words frequencies from Zipf's score, see https://www.researchgate.net/publication/354541048_A_New_Method_of_Automatic_Text_Document_Classification.

The application works in user-focused and predefined genres modes.

By default, user-focused mode is active.

To perform classification in this mode the user needs to 1) load reference texts; 2) load input texts; 3) choose output directory; 4) indicate class or genre you are interested in; 5) start classification using "threshold 1" or "threshold 2" option.

Genre-predefined mode

When in this mode, the user doesn't need to load reference texts, he/she should choose one of the predefined genres.

See UserGuide file for detailed instructions. The file is available in the application's directory.

Y-classifier can process Russian and English texts in .txt format, UTF-8 encoding.

The application works on Windows machines.

Go to Downloads section to get Y-classifier application.