Bijankhan is a large tagged corpus in Persian language. Unfortunately this corpus is not in utf-8 format and also has a lot of misspelled words. I tried to normalized Bijankhan corpus by some simple replacements.
موارد اصلاح شده:
- تبدیل حروف عربی به معادل فارسی
- در برخی از کلمات فاصله و نیم فاصله در کنار هم قرار گرفتهاند
- به دلیل نامشخصی برخی کلمات دارای همزه به اشتباه وارد دادگان شدهاند. برای مثال مطمان، ارااه، مساال، تااتر و بیشتر
Simply run the bijankhan.py to normalize the corpus. for help type 'python bijankhan.py -h'
If you like this project, please donate or consider becoming a patron:
Using or modifying this project has no limitation due to its license (GNU v3.0). But if you like to use Bijankhan corpus, you need to contact to the owner.