This repository contains a corpus of 660 manually annotated pairs <claim, sentence> from economic domain.
Each pair is labeled by four annotators as premise (1) or non-premise (0) to the given claim.
"Original model" – ruGPT-3Large model, "fine-tuned model" – fine-tuned ruGPT-3Large model.
To cite: Fishcheva I.N., Osadchiy D., Bochenina K.O. Kotelnikov E.V. Argumentative Text Generation in Economic Domain // Computational Linguistics and Intellectual Technologies: Proceedings of the International Conference "Dialog-2022".