基于情感分析的新闻浏览平台设计与实现

The Design and Implementation of News Browsing Platform Based on Sentimental Analysis

随着信息技术的迅速发展,传统的以关键字为检索手段的信息获取技术已不能满足人们对信息的需求,如何高效准确地从海量的新闻和新闻评论中找到读者感兴趣的信息成为了信息检索研究的焦点。基于此,设计与开发了一个基于情感分析的新闻浏览平台,采用jsp、servlet、javabean的MVC模式实现视图层、业务逻辑层、持久层的分离,后台使用oracle数据库,采用贝叶斯分类器对新闻和新闻评论进行情感分析获取相应的情感极性,前台搜索引擎可以将查找内容和情感极性两个参数传入系统后台,系统后台通过这两个参数返回具有情感极性分类的新闻列表,从而达到从海量的新闻中查找读者感兴趣的内容,提高用户浏览体验的目的。

With the rapid development of information technology,the traditional information acquisition technique in which key words are used as retrieval means can no longer satisfy people＇s demand for information. How to efficiently and accurately find the information that readers are interested in from the mass of news and news commentary becomes the focus of information retrieval research.This paper designs a news browsing platform based on sentimental analysis by using MVC pattern of JSP,servlet,javabean implementation view layer,business logic layer and persistence layer separation. Its background is supported by oracle database,and Bayesian classifier is used to conduct emotional analysis on news and news comments to obtain corresponding sentiment polarity.Then the foreground search engine can put the content searched and sentimental polarity into the system background. Through these two parameters,the background of the system returns to the list of news with sentimental polarity classification,so as to find the content that readers are interested in from the mass of news,improving the browsing experience of users.