基于RI-GRU的英语阅读兴趣分类预测  

Based on the RI-GRU Helped English Reading Interested in Classification Prediction

作　　者：叶彦汝 YE Yanru(Anhui Yangtse Vocational and Technical College,Wuhu 241000,Anhui,China)

机构地区：[1]安徽扬子职业技术学院,安徽芜湖241000

出　　处：《山西师范大学学报（自然科学版）》2024年第4期154-160,共7页Journal of Shanxi Normal University(Natural Science Edition)

摘　　要：随着各种应用场景下所产生的信息量飞速增长,英语阅读资料的存储量也在不断增加,如何充分挖掘潜在的数据信息成为了研究难点.英语阅读兴趣的预测旨在探究英语学习者的兴趣,以门控循环单元GRU为核心,搭建了英语阅读兴趣预测模型(RI-GRU),并以英语阅读资料的比重数据为例,通过捕捉各项关系属性特征,利用GRU计算不同序列数据下的各项误差,以最小化误差为目标寻找最优模型参数,实现较高精度的英语阅读兴趣预测.最终实验表明,比较其他方法模型,RI-GRU预测模型可有效表征英语阅读资料数据的特性,减小了各项分类误差,分类预测效果较好.With the rapid growth of information generated in various application scenarios,the storage of English reading materials is also increasing.How to fully tap the potential data information has become a research difficulty.The prediction of English reading interest aims to explore the interest of English learners.With the GRU structure of the gated cycle unit as the core,a prediction model of English reading interest(RI-GRU)was built.Taking the proportion data of English reading materials as an example,by capturing the characteristics of various relationship attributes,the RNN network realized by GRU was used to calculate various errors under different sequence data,and the optimal model parameters were found with the goal of minimizing errors,to achieve a high accuracy of English reading interest prediction.The final experiment shows that,compared with other models,the RI-GRU prediction model can effectively characterize the characteristics of English reading data,reduce the classification errors,and the classification prediction effect is better.

关 键 词：英语阅读 兴趣预测 GRU RNN网络 关系属性 

分 类 号：TP391.1[自动化与计算机技术—计算机应用技术]