【摘要】：網(wǎng)絡輿情形成迅速，對社會影響巨大。隨著因特網(wǎng)在全球范圍內(nèi)的飛速發(fā)展，網(wǎng)絡媒體已被公認為是繼報紙、廣播、電視之后的“第四媒體”，網(wǎng)絡成為反映社會民生的主要載體之一。已日益成為輿情產(chǎn)生和傳播的重要領域，網(wǎng)絡輿情在社會生活中扮演著越來越重要的角色。其直接性，突出性，偏差性使網(wǎng)絡輿情雙刃劍角色的作用越發(fā)明顯。為了加強網(wǎng)絡管理和監(jiān)控，，如何更好的開展網(wǎng)絡輿情信息的監(jiān)測、追蹤與分析，已經(jīng)成為目前政府、企業(yè)、學校等機構(gòu)面臨的首要現(xiàn)實問題。網(wǎng)絡輿情監(jiān)測追蹤系統(tǒng)可以實現(xiàn)針對網(wǎng)絡中海量輿情信息自動實時的采集與分析，有效地解決以傳統(tǒng)的人工方式對輿情監(jiān)測、追蹤和分析的難題。 傳統(tǒng)的基于文本聚類的網(wǎng)絡輿情熱點追蹤算法，在處理海量網(wǎng)頁時，文本聚類速度過低，聚合結(jié)果較差。本文提出了一種基于關(guān)鍵詞提取的網(wǎng)絡輿情熱點追蹤方案，并根據(jù)微博、新聞、論壇和博客的不同特點分別設計了熱點分析模型。通過在筆者開發(fā)的啄木鳥網(wǎng)絡輿情監(jiān)測追蹤系統(tǒng)上的實驗表明，該方案行之有效，熱點分析模型識別熱點準確率高。 網(wǎng)絡輿情監(jiān)測與傳統(tǒng)搜索引擎最大的區(qū)別就在于其時效性強，滯后的信息毫無價值。及時性包括兩個方面，采集性的及時性與信息分析的高效性。本文提出了首先在海量的網(wǎng)站中根據(jù)影響力的排名，選取有限的主流權(quán)威站點作為首要信息采集源，將有限的資源充分利用，最終提升監(jiān)測效率。 最后，根據(jù)網(wǎng)絡的發(fā)展現(xiàn)狀，本文給出了一個具體的應用系統(tǒng)啄木鳥輿情監(jiān)測追蹤系統(tǒng)。引入近幾年來針對網(wǎng)絡輿情監(jiān)控方面進行系統(tǒng)建設的過程中出現(xiàn)的問題以及解決具體問題的思路，全面分析了設計建設的目標和原則，同時深入闡述了各項功能的詳細作用及最終實現(xiàn)，為及時準確預警網(wǎng)絡突發(fā)事件和全面應對分析做出一些自己的貢獻。
[Abstract]:The network public opinion forms rapidly, the influence to the society is enormous. With the rapid development of the Internet in the world, the network media has been recognized as the "fourth media" after newspapers, radio and television. The network has become one of the main carriers reflecting the people's livelihood. Network public opinion plays a more and more important role in social life. Its directness, prominence, deviation make the role of network public opinion double-edged sword more and more obvious. In order to strengthen the network management and monitoring, how to better carry out the monitoring, tracking and analysis of network public opinion information has become the most important practical problem that the government, enterprises, schools and other institutions are facing. The network public opinion monitoring and tracking system can realize the automatic real-time collection and analysis of mass public opinion information in the network, and effectively solve the difficult problem of monitoring, tracking and analyzing public opinion in the traditional manual way. The traditional hot spot tracking algorithm of network public opinion based on text clustering, when dealing with massive web pages, the speed of text clustering is too low, and the aggregation result is poor. This paper proposes a hot spot tracking scheme based on keyword extraction, and designs a hot spot analysis model according to the different characteristics of Weibo, news, forum and blog. The experiment on the woodpecker network public opinion monitoring and tracking system developed by the author shows that the scheme is effective and the hot spot analysis model has high accuracy rate of hot spot recognition. The biggest difference between online public opinion monitoring and traditional search engine lies in its timeliness, lag information is of no value. Timeliness includes two aspects: the timeliness of collection and the high efficiency of information analysis. In this paper, firstly, according to the rank of influence, we select the limited mainstream authoritative sites as the primary information collection source, make full use of the limited resources, and finally improve the efficiency of monitoring. Finally, according to the development of the network, this paper presents a specific application system, woodpecker public opinion monitoring and tracking system. This paper introduces the problems in the process of system construction in the area of network public opinion monitoring in recent years and the train of thought to solve the specific problems, and analyzes the objectives and principles of the design and construction in an all-round way. At the same time, the detailed function of each function and its final realization are expounded in detail, and some contributions are made for timely and accurate warning of network emergencies and comprehensive response analysis.
【學位授予單位】：河北大學
【學位級別】：碩士
【學位授予年份】：2013
【分類號】：TP391.1;G206

【參考文獻】

相關(guān)期刊論文 前10條

1 廉捷;劉云;;網(wǎng)絡輿情中的信息預處理與自動摘要算法[J];北京交通大學學報;2010年05期

2 羅坤瑾;;網(wǎng)絡輿情監(jiān)控與構(gòu)建公共信息空間[J];當代傳播;2012年06期

3 施聰鶯;徐朝軍;楊曉江;;TFIDF算法研究綜述[J];計算機應用;2009年S1期

4 劉毅;;略論網(wǎng)絡輿情的概念、特點、表達與傳播[J];理論界;2007年01期

5 姜勝洪;;網(wǎng)絡輿情熱點的形成與發(fā)展、現(xiàn)狀及輿論引導[J];理論月刊;2008年04期

6 曾潤喜;;網(wǎng)絡輿情信息資源共享研究[J];情報雜志;2009年08期

7 錢愛兵;;基于主題的網(wǎng)絡輿情分析模型及其實現(xiàn)[J];現(xiàn)代圖書情報技術(shù);2008年04期

8 殷衛(wèi)東;朱曉華;趙俊凱;;網(wǎng)絡輿情技術(shù)淺析[J];信息網(wǎng)絡安全;2012年03期

9 張壽華;劉振鵬;;網(wǎng)絡輿情熱點話題聚類方法研究[J];小型微型計算機系統(tǒng);2013年03期

10 范學貴;;網(wǎng)絡社情民意 推進社會主義民主政治建設探析[J];協(xié)商論壇;2012年02期

本文編號：2137571