This is the repository for the COLING 2022 paper: "QSTS : A Question-Sensitive Text Similarity Measure for Question Generation".
The model for question-class prediction was trained using T5-large (https://huggingface.co/docs/transformers/model_doc/t5) on the TREC dataset (https://cogcomp.seas.upenn.edu/page/resource_view/49)
We used Stanza (version 1.2) for obtaining dependency parses and NER information (https://stanfordnlp.github.io/stanza/)