本文選題：網(wǎng)絡輿情 + 情感傾向　； 參考：《湘潭大學》2017年碩士論文

【摘要】：隨著移動互聯(lián)網(wǎng)的快速發(fā)展,社交網(wǎng)絡已經(jīng)成為用戶獲取信息、表達意見、交流看法的重要平臺。熱點事件一旦發(fā)生后,網(wǎng)絡用戶可以通過文本、圖片、小視頻等方式表達自己對某個社會事件的態(tài)度、認知、意見和情感等主觀性信息。信息經(jīng)過轉發(fā)、評論和點贊等方式進行傳播,同時若用戶在轉發(fā)與評論信息時加入個人主觀性情感,從而促進了事件的演化。近年來,網(wǎng)絡群體性事件數(shù)量急劇上升,在網(wǎng)民中引起了巨大的輿論反響,當突發(fā)事件爆發(fā)時若不對不良情感進行控制和引導,輿論則很容易極端化,甚至危及社會安全與穩(wěn)定。因此,有必要面向網(wǎng)絡輿情進行用戶情感傾向性分析研究,為政府有效掌握和監(jiān)控網(wǎng)絡輿情突發(fā)事件提供相應的理論支持和對策建議。本文以“羅一笑”網(wǎng)絡熱門話題事件為例,對輿情信息進行情感分析和輿情追蹤。主要的研究工作包括:第一,利用網(wǎng)絡爬蟲工具采集事件相關微博數(shù)據(jù),并進行整理分析。第二,以知網(wǎng)HowNet等詞典為基礎對情感詞進行擴展,構建一個比較全面的情感分類詞典,同時對各情感詞所表達的情感極性和強度進行識別和標記。第三,構建情感傾向分析模型,判斷網(wǎng)絡輿情的情感類型和統(tǒng)計情感詞頻,并對該事件中的用戶情感進行挖掘與可視化分析。第四,運用實證分析研究,對該事件的輿情演化階段進行劃分,分別對各階段用戶情感演化特征及規(guī)律進行分析。為后續(xù)網(wǎng)絡輿情情感引導對策的提出提供參考依據(jù)。實驗表明,網(wǎng)絡輿情從生成到最終消亡是一個完整的生命周期,通過對網(wǎng)絡輿情演化進行科學的階段劃分,可以發(fā)現(xiàn)各階段特征:(1)開始期微博發(fā)布數(shù)量少,網(wǎng)民對網(wǎng)絡輿情事件的態(tài)度紛繁復雜,但是通過對文本中用戶情感的挖掘、觀點的提取有利于進一步跟蹤事件的后續(xù)發(fā)展趨勢;(2)爆發(fā)期微博發(fā)布數(shù)最多,用戶參與度最高,影響范圍和影響效果極大,網(wǎng)民對事件的態(tài)度、觀點、情感等信息能夠為網(wǎng)絡輿情分析和監(jiān)控提供足量的數(shù)據(jù)基礎,同時,爆發(fā)期的情感傾向很大程度上定義了網(wǎng)絡輿情事件的總體情感演化趨勢,相關部門應對爆發(fā)期的網(wǎng)絡輿情情感演化多加關注,并引導輿情朝著正確的方向發(fā)展;(3)發(fā)酵期網(wǎng)民對網(wǎng)絡輿情事件的新資訊、新動態(tài)較為敏感,正面信息公開與輿情披露在此階段能夠起到良好的效果;(4)消解期和反思期用戶參與程度較低,但仍需要對網(wǎng)絡輿情事件進行跟蹤報道,規(guī)避謠言,肅清網(wǎng)絡環(huán)境,避免網(wǎng)絡輿情事件的二次發(fā)酵。
[Abstract]:With the rapid development of mobile Internet, social network has become an important platform for users to obtain information, express opinions and exchange views. Once a hot event occurs, Internet users can express their attitude, cognition, opinion and emotion on a social event by means of text, picture, small video and so on. The information is transmitted by way of forwarding, commenting and liking, and if the user adds personal subjective emotion to transmit and comment on the information, it promotes the evolution of the event. In recent years, the number of network mass incidents has risen sharply, causing a huge public opinion response among Internet users. When emergencies break out, if they do not control and guide bad emotions, public opinion is easy to become extreme. Even endanger social security and stability. Therefore, it is necessary to analyze and study the emotional tendency of users in order to provide corresponding theoretical support and countermeasures for the government to effectively grasp and monitor the sudden events of network public opinion. Taking Luo Yixiao as an example, this paper analyzes and tracks public opinion information. The main research work is as follows: firstly, the Weibo data are collected and analyzed by using web crawler tools. Secondly, based on the HowNet dictionary, we construct a comprehensive emotion classification dictionary, and identify and mark the emotion polarity and intensity expressed by each emotion word. Thirdly, we construct an emotional tendency analysis model to judge the emotional types and statistical affective word frequency of network public opinion, and mine and visualize the user emotion in this event. Fourthly, using the empirical analysis, the public opinion evolution stage of the event is divided, and the characteristics and rules of user emotion evolution in each stage are analyzed respectively. It provides the reference for the following network public opinion emotion guidance countermeasure. The experiment shows that the network public opinion is a complete life cycle from the generation to the final extinction. By dividing the evolution of the network public opinion into scientific stages, we can find that the number of Weibo releases at the beginning of each stage is small. Internet users' attitude to network public opinion events is complicated, but through the mining of users' feelings in the text, the point of view extraction is conducive to further tracking the future development trend of events.) in the outbreak period, the number of Weibo releases is the most, and the participation of users is the highest. The influence range and the influence effect are great, the netizens' attitude, viewpoint, emotion and so on information can provide the sufficient data foundation for the network public opinion analysis and the monitoring, at the same time, The emotional tendency of the outbreak period has largely defined the overall emotional evolution trend of the network public opinion event, and the relevant departments should pay more attention to the evolution of the network public opinion emotion in the outbreak period. And guide the public opinion to develop in the right direction.) during the fermentation period, the netizens are more sensitive to the new information about the network public opinion events. Positive information disclosure and public opinion disclosure can play a good effect in this stage. (4) the level of user participation in the period of resolution and reflection is relatively low, but it is still necessary to track and report online public opinion events, to avoid rumors, and to eliminate the network environment. Avoid the secondary fermentation of network public opinion events.
【學位授予單位】：湘潭大學
【學位級別】：碩士
【學位授予年份】：2017
【分類號】：C912.63

【參考文獻】

相關期刊論文 前10條

1 劉培玉;荀靜;費紹棟;朱振方;;基于隱馬爾可夫模型的主觀句識別[J];中文信息學報;2016年04期

2 周志杰;劉功申;;基于投票機制的文本主客觀分類系統(tǒng)研究[J];微型電腦應用;2015年02期

3 侯敏;滕永林;李雪燕;陳毓麒;鄭雙美;侯明午;周紅照;;話題型微博語言特點及其情感分析策略研究[J];語言文字應用;2013年02期

4 柳軍;蔡淑琴;;微內容的網(wǎng)絡輿情傳播特征分析[J];情報雜志;2013年01期

5 唐超;;網(wǎng)絡情緒演進的實證研究[J];情報雜志;2012年10期

6 肖強;朱慶華;;Web2.0環(huán)境下的“網(wǎng)絡推手”現(xiàn)象案例研究[J];情報雜志;2012年09期

7 羅旭;;網(wǎng)絡政治參與中的輿論領袖研究:演進與規(guī)制[J];探索;2012年04期

8 代大明;王中卿;李壽山;李培峰;朱巧明;;基于情緒詞的非監(jiān)督中文情感分類方法研究[J];中文信息學報;2012年04期

9 李本陽;關毅;董喜雙;李生;;基于單層標注級聯(lián)模型的篇章情感傾向分析[J];中文信息學報;2012年04期

10 朱恒民;李青;;面向話題衍生性的微博網(wǎng)絡輿情傳播模型研究[J];現(xiàn)代圖書情報技術;2012年05期

相關重要報紙文章 前1條

1 尹潘嶸;陶建華;;情感與認知 相互觸發(fā)[N];計算機世界;2005年

相關博士學位論文 前4條

1 黃勝;Web評論文本的細粒度意見挖掘技術研究[D];北京理工大學;2014年

2 紀雪梅;特定事件情境下中文微博用戶情感挖掘與傳播研究[D];南開大學;2014年

3 李澍淞;基于蜂擁策略的網(wǎng)絡輿論演化模型研究[D];復旦大學;2011年

4 張玉強;網(wǎng)絡輿情危機的政府適度反應研究[D];中央民族大學;2011年

相關碩士學位論文 前1條

1 趙城利;基于Web的信息智能感知技術及應用[D];國防科學技術大學;2004年

，

本文編號：1990163