Exploring the Effectiveness of Question for Neural Short Answer Scoring System
Gaoyan Lv, Wei Song, Miaomiao Cheng, Lizhen Liu · 2021
Automatic short answer scoring is a significant branch in the field of Natural Language Processing. It can release teachers from the complicated correcting work. It is an effective method for students to recognize their knowledge weakness as well. With the development of neural network and deep learning, researchers have built many neural models to resolve the problem. To the best of our knowledge, no one uses question as a feature in neural models. Starting from human experience and common sense, we explore the effectiveness of incorporating question into neural models for short answer scoring. The experimental results show that, on average, our model gains 1.1% improvement compared to the state-of-art model.