Our Open Source NLP Libraries

In Starlang, we believe that the development of both science and technology heavily relies on the free circulation of knowledge. That is why we offer 6 free and open source libraries for developers. You can employ our libraries in your Turkish NLP processes.

All 6 of our open source libraries are available in 3 different programming languages to ensure the ease of use.


KeNet is a lexical database in WordNet format where you can find semantic relations between Turkish words including synonymy, antonymy, hyponymy and meronymy. Related lemmas are grouped into synsets.


HisNet is designed for sentence-based sentiment analysis in Turkish texts.


TRopBank is a cohesive PropBank developed for Turkish where you can find annotated verbal arguments and propositions.

Morphological Analyser

Morphological Analyser breaks words into their morphemes, so that their internal structure can be determined.


DEASCIIFIER converts Turkish texts written with ASCII-only characters into proper Turkish.

Morphological Disambiguator

Morphological Disambiguator analyses the internal structure of the words in order to solve ambiguities.

You can get access on our github profile.