Context-Net

This repository houses model and code for a recurrent neural network (LSTM + attention) based moderator agent for critical discourse in discussion forums. Currently the models are limited to student discussion data from Massive Open Online Courses.

A pre-print version of this work is available on arXiv: https://arxiv.org/pdf/1905.10851.pdf Please cite as follows if you find it useful:

Chandrasekaran, M. K., & Kan, M. Y. (2019). When to reply? Context Sensitive Models to Predict Instructor Interventions in MOOC Forums. arXiv preprint arXiv:1905.10851.

@article{chandrasekaran2019reply,
  title={When to reply? Context Sensitive Models to Predict Instructor Interventions in MOOC Forums},
  author={Chandrasekaran, Muthu Kumar and Kan, Min-Yen},
  journal={arXiv preprint arXiv:1905.10851},
  year={2019}
}