Automatic recognition of Ezāfe is one of the goals of Persian Language Processing, which is addressed in this tool. It provides one of the best results in this field through using the Conditional Random Fields (CRF) algorithm. Our results show an accuracy of 98% at a speed of over 100,000 words per second, which is unique in its kind.
A variety of applications can be imagined for this tool, including its use in text-to-speech tools, machine keyword extraction systems, machine translation, named entities recognition, and many other applications.