Han-Coref: Thai Coreference resolution by PyThaiNLP
Wannaphong Phatthiyaphaibun, Peerat Limkonchotiwat · Zenodo (CERN European Organization for Nuclear Research) · 2023
This technical paper presents the development of a Thai coreference resolution dataset and model using an open-source approach. Due to the lack of datasets and models for coreference resolution in Thai, we aim to provide resources for future researchers in this field. We created a dataset from news articles and Wikipedia pages and had the data expertly annotated. In addition, we trained the model on the fastcoref library, which enables fast, precise, and user-friendly training. With their clear techniques and easily applicable methods, this paper allows other experts to contribute to growing NLP for the Thai Language.