目录
摘要III
关键词III
Abstract IV
引言
引言1
1 概述1
1.1 研究背景和意义1
1.2 国内外研究现状1
1.2.1 文本情感分析1
1.2.2 用户观点挖掘2
1.2.3 文本聚类2
1.3 研究目标及内容2
1.3.1 研究目标2
1.3.2 研究内容2
2 需求分析3
2.1 功能需求3
2.2 系统用例图3
2.3 系统类图4
2.4 时序图5
3 系统设计8
3.1 系统功能模块设计8
3.2 数据库设计9
3.2.1 ER 图9
3.2.2 数据库表设计 13
4 系统实现 15
4.1 系统框架 15
4.2 系统实验环境 15
4.3 系统功能实现 15
4.3.1 核心功能模块 15
4.3.1.1 用户评论情感分析 15
4.3.1.2 用户评论观点挖掘 22
4.3.1.3 资料多维度特点展示 24
4.3.2 基本功能模块 25
4.3.2.1 注册登录 25
4.3.2.2 用户信息管理 26
4.3.2.3 关键字查询 27
4.3.2.4 发表经验贴 27
4.3.2.5 资料信息展示 28
5 系统测试 30
5.1 测试目的 30
5.2 测试内容 30
5.3 测试方法 30
5.4 测试过程 30
5.5 测试结果分析 32
6 总结与展望 32
6.1 总结 32
6.2 展望 33
致谢 33
参考文献 33
考研资料评论网站及自动评价方法的设计与实现
摘 要
近年来， *51今日免费论文网|www.51jrft.com +Q: ￥351916072￥ 
考研已经逐渐成为大学生毕业的主要选择之一，伴随考研人数的暴增，各种考研资料也在市场大量涌现。而随着互联网购物模式的兴起，考研学生普遍选择在网购平台上购买考研资料，并会在网购后会对资料做出评论，超七成的学生也会在网购资料时浏览其他消费者的评论。但是学生在浏览评论时会遇到评论数据量大、无价值评论多、评论分散在不同的网购平台上，不便于综合对比等问题，难以获取有价值的信息。为了解决这一系列问题，本系统采用Django框架的MVT设计模式，搭建了一个专门的考研资料评论网站，使得资料评论数据集中化，同时利用朴素贝叶斯算法对用户提交的资料评论进行情感分析，获取学生对资料的总体评价；基于依存句法分析技术设计观点抽取规则来挖掘用户观点，并采用Kmeans聚类算法对挖掘出的用户观点进行处理，生成资料标签，使考研学生能从多维度了解资料的特点。
DESIGN AND IMPLEMENTATION OF POSTGRADUATE DATA REVIEW WEBSITE AND AUTOMATIC EVALUATION METHOD
ABSTRACT
In recent years, postgraduate entrance examination has gradually become one of the main options for college graduates. With the explosion of the number of postgraduate entrance examinations, various postgraduate entrance examination materials are also emerging in the market. With the rise of the Internet shopping model, postgraduate students generally choose to purchase postgraduate materials on the online shopping platform, and will comment on the materials after online shopping. Over 70% of the students will also browse other consumer reviews during online shopping. However, when browsing the reviews, students will encounter problems such as large amount of review data, many nonvaluable reviews, and scattered reviews on different online shopping platforms. It is difficult to compare them comprehensively, and it is difficult to obtain valuable information. In order to solve this series of problems, this system uses the MVT design model of the Django framework to build a dedicated postgraduate data review website to centralize the data of the data review, and at the same time uses the Naive Bayes algorithm to perform sentiment analysis on the data reviews submitted by users to obtain students overall evaluation of the materials. Based on the dependency parsing technology, design the viewpoint extraction rules to mine user opinions, and use the Kmeans clustering algorithm to process the mined user opinions to generate data labels, so that graduate students can understand the characteristics of the data from multiple dimensions.

原文链接：http://www.jxszl.com/jsj/wljs/606923.html