基于flex的大学英语作文自动评阅系统的设计与实现-计算机技术专业论文.docxVIP

摘 摘 要 I I 摘 要 随着社会的快速发展，英语的学习已变得越来越重要和普及。而在英语教学中， 英语作文的训练是非常重要的，人工评阅往往不够客观，且会给教师带来很大的工作 量，英语作文的自动评阅技术应运而生，并成为了一项非常有前景的研究课题。 英语作文评分模型是英语作文自动评阅系统的关键，本文在深入研究潜在语义技 术的基础上构建出英语作文评分模型，通过文本预处理、生成特征值—文本矩阵、奇 异值分解、选取最优 K 值以及文本相似性比较等步骤来实现英语作文文本内容质量上 的自动评分，并对模型进行了详细设计和实现。本文从英语作文自动评阅系统实现的 角度通过系统需求分析、系统模块设计、体系结构设计、数据库设计和系统分层设计 等几步完成了系统的实现，其中系统分层主要分为持久层、业务逻辑层和表示层。 整个系统分为学生端和教师端，本文主要介绍学生端的开发过程，学生端子系统 实现了登录注册模块、作文范文模块、写作训练模块、习作查询模块、批阅标准模块 和写作知识模块等六大功能模块，并结合自动评分模型，在写作训练模块中实现了作 文自动批阅，给出了系统得分和对作文质量的分析，最后的测试结果表明了本系统中 给出的机器评分与人工评分有较高的一致性。 关键词：Flex；作文自动评分；潜在语义分析 Ab Abstract II II Abstract With the rapid development,English language learning has become more and more important and popular. English composition training is very important in English teaching, But the artificial reviewers tend to be less objective and will give teachers a great deal of work, So the automatic writing of the English Review become a very promising research topic. The model of review for English composition is the key to the automatical review system for English composition，This paper have build out English composition marking model based on latent in-depth study of semantic technologies, Through the text pre-processing,generate feature value - the text matrix, singular value decomposition,select the optimal K value and text similarity comparison it have achieve the automatically marked of English writing in quality，and give the detailed design and implementation of model. This article is by the point of the implementation for the system, and through the system requirements analysis, system module design, architecture design, database design and system hierarchical design it have complete the implenmentation of system, which the system is divided into persistence layer, business logic and presentation layer. The whole system is divided into student and faculty end, This paper is mainly responsible for the development of students side, Student Terminal System have achieve the Home module, writing essay module, wri