A Procedure for measuring the performance of a Thai Word Segmentation software is as follow:
1. Download our Test Data(Member only)
- Article
- Encyclopedia
- News
- Novel
Click Here to Download
|
2. Use your software to segment the data. Produce twelve output files , one for each genre, using UTF-8 encoding only. |
3. Test each output file with the BEST Evaluation Tool.
|
|