Natural Language Processing @ KAIST
CS492 Machine Learning for Natural Language Processing
The idea of using CNN for text-based multiple choice question answering is a novel idea back in 2017. And surprisingly the CNN model outperformed traditional LSTM models in this task. Our project for this course is to reimplement a CNN based question answering model from a paper and carry out modifications in order to investigate how we may further improve the model.
 CNN for Text-Based Multiple Choice Question Answering](/project/cnn_qna/nlp_hu24c1bdaa8f75dca39fdd9d782da41bad_33544_2321f5ab45b2fac7b430a255049876af.webp)
