This repository provides:
- Methods for experiments with the RuSERRC dataset: RAKE, BERT-based models, ERNIE.
- Word embeddings trained on Russian Wikidata and models for NER and RC trained on the RuSERRC dataset (available at https://github.com/iis-research-team/ruserrc-dataset)
- Dictionary with scientific terms (available at https://github.com/iis-research-team/ner-rc-russian/tree/master/dict_extractor)