BOUN-Pars creates dependency parse trees of Turkish sentences in CoNLL-U format .

It is based on Stanford's graph-based neural dependency parser and uses linguistically oriented rules and benefits from morphological information of words.

The pre-processing steps of parsing from raw text: the segmentation, morphological tagging, and lemmatization tasks are performed by a pre-trained model by TurkuNLP pipeline.

BOUN-Pars is written on Python language and the source codes are available here .

BOUN-Pars is developed by Şaziye Betül Özateş, Arzucan Özgür, Tunga Güngör from the Department of Computer Engineering, and Balkız Öztürk from the Department of Linguistics, at Bogazici University.

This research is supported by the Scientific and Technological Research Council of Turkey (TUBITAK) under grant number 117E971.

Please cite the following paper if you make use of this resource:

@article{ozates2020hybrid,

        author ={ Şaziye Betül Özateş, Arzucan Özgür, Tunga Güngör and Balkız Öztürk},

        title ={A Hybrid Approach to Dependency Parsing: Combining Rules and Morphology with Deep Learning},

        journal ={Under Review}

}



Home    Demo    Publications    Resources    TabiLAB