融合BERT和眼动技术的语文阅读能力训练系统

Chinese Reading Ability Training System Integrating BERT and Eye Movement Technology

本文介绍一种新型语文阅读能力训练系统。该系统应用眼动仪技术,记录眼球移动的轨迹,收集眼动轨迹和停留时间等数据,使用热点图呈现用户浏览和注视的情况。同时结合关键词抽取等自然语言处理算法,从阅读速度、阅读技巧、关键信息提取、记忆力、注意力五个角度科学自动地给出阅读能力提升的方案。通过对比实验,选取眼动仪最优的摆放角度和位置,并与基准模型比较,从推理时间和响应时间测试最优的关键词抽取模型。

This paper introduces a new type of Chinese reading ability training system.The system uses eyetracker technology to record the trajectory of eye movement,collect data such as eye trajectory and dwell time,and use focus information maps to present the user's browsing and gaze situation.Combined with natural language processing algorithms such as keyword extraction,it scientifically and automatically provides a plan for improving reading ability from five perspectives:reading speed,reading skills,key information extraction,memory,and attention.Through comparative experiments,selecting the optimal placement angle and position of the eye tracker,and compared with the benchmark model,choosing the optimal keyword extraction model by the reasoning time and response time.