A Deep Learning Model Capable of Identifying Low-Value Pages on Professional Theses written in the Turkish Language in PDF Format
Low-Value pages are defined as follows: Table of Contents, Table of Figures, Tables of Tables, References, and Appendices.
The model detects these pages and outputs a list that contains the page number of every page it has labeled as a Low-Value page.
- Can read from CSV, PDF, and URL's
- High Accuracy Values (~98%)
- Fast Algorithm
- Usable with any type of text in the Turkish Language.
Run Last_version.ipynb