AI-based system for analyzing students' reading behavior

Abstract

This invention relates to the field of reading recognition technology, in particular an artificial intelligence-based reading analysis system comprising a behavior collection module, a skimming detection module, a path construction module, a tag generation module, and a cognitive modeling module. In this invention, the completeness of the reading operation chain is improved and the granularity of the analysis of student reading behavior data is increased by frequently collecting scrolling and clicking events; the dimensions of skimming detection are extended and the accuracy of behavioral assessment is improved by using chapter structure thresholds and analyzing non-scrolling areas; and a comparison system with the first visit is established by collecting changes in dwell time on return paths, thereby improving the ability to detect students' deep reading behavior.By tracking the increasing trend in dwell time, a hierarchical learning tag system is constructed; by integrating behavior frequency and time fluctuation dual indicators and mapping them onto a knowledge graph based on clustering results, the dynamic comparison and difference analysis of chapter reading behavior is strengthened, thereby achieving a precise positioning and analysis of student reading behavior.

Images