Mn031 / Panian-Word-Generator-NLP-

Panian-Word-Generator-NLP-

Why We came up with the project

Pāṇini achieved with his grammar what is called “non-lossy compression” in computer science, in the sense that he was able to create the original words and forms from a very compressed set of rules. Using a simple and intuitive set of rules, we can arrive at very complex words signifying an array of meanings. These rules also make the generation of new words and their verification very easy. For example, given a root word with the “abhipray” or the intention of speaking in mind. Then run it through the Aṣṭādhyāyī to find a suffix that is most relevant to the “abhipray”. We have the intended word. We can employ the same mechanism while verification of new words to see if the word we have matches the one we constructed using the above mechanism. Since the entire process has an algorithmic structure and efficiency to it, we decided to test for ourselves whether we could code the sutras in a computer in a union of two distant but somehow related disciplines. Coding the Aṣṭādhyāyī with all its interdependencies and _ is a very arduous task. Yet to accomplish what we started out to do and do justice to the magnificence of the text we coded the IT Markers and the affix “aniyar” which means (to be) “worthy of”.

Our Methodology

Since there exists a plethora of root words and their appropriate suffixes and rules, we decided to start with a small subset of the various sutras that can perform and convey one particular “abhipraay” (of being worthy).

How the Aṣṭādhyāyī does this is as follows:

1. Take a Dhatu from the Dhatupaath and remove the IT markers
2. Add the suffix aniyar to it after removing its IT markers
3. Combine the two using sandhi rules
4. Accommodate for certain changes after the Sandhi.

The various nuances to this are as follows

The various nuances to this are as follows

1. The Dhatus in the Dhatupaath are not listed according to their Dhatu but rather with IT markers for various uses. Hence we take this IT marked Dhatu and first remove the ITs
2. We take the suffix aniyar and remove its IT markers
3. The sandhi between the two words is interesting:

• If the dhatu ends with a Hal, the penultimate swar does Guna.
• If the dhatu ends with an AJ, the ultimate swar does a Guna Next, replace ultimate vowels with certain sets of letters

4. Under certain conditions, the n in aniyar changes to ण ्and this accommodated for