联系我们

论文查询结果	返回搜索

论文编号：	10987
作者编号：	2120163283
上传时间：	2019/6/16 11:58:40
中文题目：	基于新闻评论文本的网络舆情分析研究 ——以红黄蓝幼儿园事件为例
英文题目：	Research on Internet Public Opinion Analysis Based on News Comment Text ——Taking the event of RYB Education as an example
指导老师：	王芳
中文关键字：	网络舆情；情感分析；红黄蓝；情感词典
英文关键字：	Internet public opinion；sentiment analysis；RYB Education；Sentiment dictionary
中文摘要：	我国的互联网正在发生着剧烈的变化、不断地发展，从小众普及到大众，互联网正在从各个方面改变人们的生活方式。由于互联网的准入门槛较低，大家通过互联网可以接触和了解到不同的信息，但是与此同时各类信息鱼龙混杂。越来越多网民习惯从互联网媒体中获取信息，从而发表对事件的看法、态度以及情感倾向。尤其是作为主要信息载体的新闻，已成为互联网资讯的集散地和放大器。互联网也成为一把双刃剑，既能提升信息的分享和消除消息鸿沟，但是也被一些别有用心的人利用，网络舆情事件影响也在随之扩大，动辄产生现象级的舆情话题，造成社会舆论，影响社会稳定。网民在互联网中最直接反应情绪的就是文字，并且大部分反映观点的文字较短，如微博评论不得多于140字，新闻类网站一般的评论字数在10到30字之间。与此同时，每一个评论用户还包括用户的个人信息，如性别、发文地点、年龄、工作等，因此，评论文本中蕴含着巨大的研究价值。如何能够从文本中获取信息并有效进行舆情分析是当前研究的关键，因此本文旨在对红黄蓝事件的评论文本进行深入的舆情分析，从大量的无序文本中快速提取关键的信息，并对其进行情感分析。 本文以舆情爆发的典型“红黄蓝”虐童虐童事件为例，详细的梳理该事件发生期间，网络舆情的变化过程，并对变化过程进行分析研究。主要研究工作如下：一、利用情感词典技术分析文本中出现的情感词，构建情感分析模型对其情感倾向进行判断，以此挖掘出文本中隐藏的情感倾向，分析出网络舆情的发展趋势，更好地把握网络舆情。二、计算出该事件的评论文本情感倾向性，并进行分析。三、通过计算文本情感值，对舆情进行阶段划分，研究每个阶段的情感趋势，并提出相关的建议对策。希望相关部门可以对于舆情早发现，早预防，早处理，并进行舆情的正确引导与监管，保证网络环境健康发展。
英文摘要：	China's Internet is undergoing dramatic changes, continuous development and popularization, and the Internet is changing people's lifestyles in all aspects. Due to the low barriers to entry on the Internet, everyone can access and learn about different information through the Internet, but at the same time all kinds of information are mixed. More and more netizens are accustomed to getting information from the Internet media to express their views, attitudes and sentiments towards events. In particular, news is the main carrier to become the distribution center and amplifier of Internet information. The Internet has also become a double-edged sword. It can not only improve the sharing of information and eliminate the message gap, but it is also used by people with ulterior motives. The influence of online public opinion events is also expanding, and the phenomenon of lyricism is generated, resulting in public opinion, affecting social stability. The most direct response of Internet users to the Internet is the text, and most of the responses are short-formed. For example, Weibo comments should not exceed 140 words, and news websites generally have between 10 and 30 words. At the same time, each comment user also includes the user's personal information, such as gender, place of posting, age, work, etc., so the comment text contains great research value, how to get information from the text and effectively analyze the public opinion is the key to current research. Therefore, this thesis aims to carry out in-depth analysis of the commentary texts of the red, yellow and blue events, quickly extract key information from a large number of disordered texts, and analyze the emotions. This paper takes the typical "red, yellow and blue" child abuse incidents that broke out in lyrics as an example, and combs the changes of the network public opinion during the incident, and analyzes the change process. The main research work is as follows: First, using the reptile technology, collecting news comment text data related to the red, yellow and blue events, and performing de-drying and word segmentation processing. Secondly, using emotional dictionary technology to analyze the emotional words appearing in the text, constructing the sentiment analysis model to judge its emotional tendency, so as to mine the hidden emotional tendency in the text, analyze the development trend of online public opinion, and better grasp the network public opinion. Third, by calculating the emotional value of the text, the lyrics are divided into stages, the emotional trends at each stage are studied, and relevant suggestions are proposed. It is hoped that the relevant departments can find out early, prevent, deal with, and conduct correct guidance and supervision of sensation to ensure healthy development.
查看全文：	预览  下载(下载需要进行登录)