基于LSTM模型的学生反馈文本学业情绪识别方法

Recognition Method for Academic Emotions of Students' Feedback Texts Based on LSTM Model

分析学生学习过程产生的反馈文本,是发现其学业情绪的重要方式。传统的学业情绪测量方法主要包括使用学业情绪测量问卷和访谈分析,但这两种方法难以大规模地应用于在线教育环境。本研究旨在通过构建学业情绪自动预测模型,对大量学生反馈文本进行快速有效的学业情绪分类。研究首先利用词向量训练工具,将文本转化为多维向量;然后基于深度学习网络LSTM构建学业情绪预测模型,以文本的多维向量作为模型输入;最后经过多轮训练,优化模型参数。实验显示,上述模型可快速有效识别学生反馈文本中所包含的学业情绪,该模型在测试数据集上的学业情绪识别准确率可达89%。

Analyzing emotional texts produced by students in the learning process is an important way to discover students’academic emotions.The traditional method of measuring academic emotion is to use academic emotion questionnaire or interview approach,but this method is difficult to be widely used in an online learning environment.Therefore,this study aims to quickly and effectively discovering the implicit categories of academic emotions in a large number of student feedback texts by constructing an automatic predictive model of academic emotions.This paper first uses the word vector training tool to transform the text into a multi-dimensional vector.Then,based on the deep learning network LSTM,the academic sentiment prediction model is constructed.The model consists of two layers of LSTM,with the multidimensional vector of the text as input.Finally,after several rounds of training,optimizing Model parameters,Experiments show that the above model can quickly and effectively identify the academic sentiment contained in the student’s feedback text.The accuracy of the school’s academic sentiment in the test data set can reach 89%.