Designing an Educational Chatbot with Joint Intent Classification and Slot Filling
Yujia Wang, Shuqi Liu, Linqi Song · 2022
In the post-pandemic era, with the continuous shifting trend from offline to online teaching mode and the development of technologies in virtual communication, the necessity of a virtual assistant for education-related application scenarios is gradually being recognized. However, simple and educational domain-specific chatbot is not well studied. In this work, we develop an educational chatbot based on a Natural Language Understanding (NLU) model that can understand task-oriented natural language texts to provide education-related services. The NLU model utilizes a large pre-trained language model of BERT variants as backbones to jointly optimize intent classification and slot filling modules, enabling the proposed educational opendomain goal-oriented intrapersonal chatbot to capture critical information from user queries. Our system achieves good performance. The system is demonstrated to be able to assist in reducing the communication burden between students and teachers in the online teaching environment, ease the shift of students from physical to virtual classrooms, and alleviate the pressure on the teaching staff.