Analysis of E-Learning Logs to Estimate Students' Phonemic Perception Confusion in English Word Recognition

摘要

In the paper, we analyzed students' learning log data obtained in a word listening learning system to estimate phonemic perception confusion of students. We had two student groups take a word dictation test and a word choice test. English words in the word dictation test were resolved into phonemes and the abilities of both groups on phonemic perception were inferred by Bayes' rule. By comparing the results estimated by Bayes' rule with those obtained in the word choice test, we suggested that the estimation is valid and helpful to build adaptive word recognition training systems.