Developing an open-domain question answering system for a low-resource language, Turkish, with machine translation.
We create a search engine to identify bichemically related molecules in CORD-19 abstracts.
We present the first large labeled Turkish NLI dataset (NLI-TR) and models using machine translation.
We present the first UI that enables online traversal of Turkish WordNet: KeNet.
We develop a web application to use a Turkish NLP library (nlptoolkit) online.
A dataset to measure sentence and paragraph level similarity in Turkish texts.