In assessing students’ language learning progress, it will be most helpful if a tool can automatically score students’ writing tasks. It can help reduce teachers’ workload and shorten the time to provide feedback to students. In this talk, we will introduce an efficient AI-driven automated scoring tool for a picture writing task which asks students to write a sentence to describe a given picture. In particular, we will first describe how we design a picture writing test, create a data set for training and testing, and then develop an AI scoring model by considering the content from both the picture and its textual description. Finally, we will report the testing performance of our AI model and give some concluding remarks. This is a joint work with Ruibin Zhao (MIT), Yipeng Zhuang (MIT), Di Zou (ELE) and Qin Xie (LML). Copyright © 2022 EDTECH Conference.
|Original language||Chinese (Traditional)|
|Publication status||Published - Jun 2022|
Citation楊良河 (2022年6月) ：人工智能驅動的看圖造句自動評分，論文發表於「EDTECH教育科技研討會2022：特殊教育科技的創新和發展」，香港，中國。
- Alt. title: AI-driven auto-grading of picture writing tasks