本文選題：網(wǎng)絡(luò)輿情 + 話題提取��； 參考：《南京郵電大學(xué)》2015年碩士論文

【摘要】：互聯(lián)網(wǎng)信息的開放式、交互式傳播使其傳播范圍更廣、傳播速度更快,一些熱點(diǎn)問題、焦點(diǎn)事件或敏感話題一經(jīng)網(wǎng)絡(luò)傳播,其影響力會(huì)成倍放大,甚至?xí)谒查g將政府、企業(yè)、組織或者個(gè)人推向風(fēng)口浪尖,網(wǎng)絡(luò)輿情的監(jiān)控和管理一直受到多方面學(xué)者的關(guān)注和重視,目前也取得一些成果,本文從話題的角度來探究網(wǎng)絡(luò)輿情演化規(guī)律,將網(wǎng)絡(luò)中混雜的信息進(jìn)行合理的整理和組織,旨在從更細(xì)的粒度上來研究網(wǎng)絡(luò)輿情的話題演化以及傳播規(guī)律,為網(wǎng)絡(luò)輿情的監(jiān)控和管理提供一定的理論基礎(chǔ)和方法指導(dǎo),本文的研究?jī)?nèi)容主要分為以下幾個(gè)方面:(1)網(wǎng)絡(luò)輿情話題的提取�；赑LSA堅(jiān)實(shí)的統(tǒng)計(jì)學(xué)基礎(chǔ),提出了網(wǎng)絡(luò)輿情話題提取方法,通過實(shí)際案例分析,成功抽取到網(wǎng)絡(luò)輿情話題,利用其統(tǒng)計(jì)學(xué)特性準(zhǔn)確描述了話題、文本以及詞之間的概率語(yǔ)義關(guān)系,并通過特征詞及特征詞對(duì)話題的貢獻(xiàn)率定量定義了話題。(2)網(wǎng)絡(luò)輿情話題的演化。首先,構(gòu)建了網(wǎng)絡(luò)輿情話題的靜態(tài)演化模型,通過計(jì)算定義話題的特征詞之間的關(guān)聯(lián)度構(gòu)造了關(guān)聯(lián)矩陣,并利用社會(huì)網(wǎng)絡(luò)分析工具繪制了網(wǎng)絡(luò)輿情話題靜態(tài)演化路徑,研究發(fā)現(xiàn),字面上沒有關(guān)聯(lián)的話題之間也會(huì)存在某種聯(lián)系,而這種聯(lián)系與特征詞的共現(xiàn)概率有關(guān);然后,構(gòu)建了網(wǎng)絡(luò)輿情話題的動(dòng)態(tài)演化模型,將網(wǎng)絡(luò)輿情話題系統(tǒng)看作一個(gè)動(dòng)態(tài)網(wǎng)絡(luò),分析了動(dòng)態(tài)網(wǎng)絡(luò)的微觀結(jié)構(gòu),并通過實(shí)證分析研究了網(wǎng)絡(luò)輿情話題的動(dòng)態(tài)演化特性:話題演化路徑、熱點(diǎn)話題演化路徑以及焦點(diǎn)話題演化路徑,從而更深入得揭示了網(wǎng)絡(luò)輿情話題的演化機(jī)理。(3)網(wǎng)絡(luò)輿情話題的傳播。首先,根據(jù)經(jīng)典傳播模型SIR分析了話題的傳播機(jī)制,并通過仿真實(shí)驗(yàn),分析了話題傳播的影響因素(用戶行為和話題內(nèi)容)對(duì)話題傳播和擴(kuò)散的作用機(jī)制,在此基礎(chǔ)上,引入了話題關(guān)聯(lián)度,通過研究多話題環(huán)境下,話題關(guān)聯(lián)度對(duì)話題轉(zhuǎn)播概率影響的基礎(chǔ)上,提出了多話題傳播模型,并研究了話題關(guān)聯(lián)度、多話題發(fā)生時(shí)間間隔以及多話題初始傳播節(jié)點(diǎn)對(duì)各個(gè)話題以及整個(gè)輿情傳播的影響,仿真實(shí)驗(yàn)分析結(jié)果表明,本文所提的話題傳播模型能夠反映真實(shí)的話題演化規(guī)律,具有一定的實(shí)際意義。
[Abstract]:The open and interactive communication of Internet information makes it spread wider and faster. Once some hot issues, focal events or sensitive topics are spread on the Internet, their influence will be magnified, and even the government and enterprises will be immediately affected. Organizations or individuals push to the forefront, monitoring and management of network public opinion has been concerned and valued by many scholars, and some achievements have been made. This paper explores the evolution law of network public opinion from the point of view of topic. In order to study the topic evolution and dissemination law of network public opinion from a finer granularity, it can provide some theoretical basis and method guidance for monitoring and management of network public opinion. The research content of this paper is divided into the following several aspects: 1) extracting the topic of network public opinion. Based on the solid statistical foundation of PLSA, this paper puts forward the method of extracting the topic of network public opinion. Through the actual case analysis, the topic of network public opinion is extracted successfully, and the topic is accurately described by its statistical characteristics. The probabilistic semantic relationship between the text and the words, and the evolution of the topic of network public opinion is quantitatively defined by the contribution rate of the feature words and the feature words to the topic. First of all, the static evolution model of network public opinion topic is constructed, the correlation matrix is constructed by calculating the correlation degree between the characteristic words that define the topic, and the static evolution path of network public opinion topic is drawn by using social network analysis tool. The study found that there is also some connection between topics that are not literally related, and this connection is related to the co-occurrence probability of feature words. Then, the dynamic evolution model of the topic of network public opinion is constructed. This paper regards the topic system of network public opinion as a dynamic network, analyzes the micro structure of the dynamic network, and studies the dynamic evolution characteristics of the topic of network public opinion through empirical analysis: the path of topic evolution, The hot topic evolves path and the focal topic evolvement path, thus has revealed the network public opinion topic evolution mechanism. 3) the network public opinion topic dissemination. First of all, according to the classical communication model SIR, this paper analyzes the mechanism of topic propagation, and through simulation experiments, analyzes the influence factors (user behavior and topic content) on the mechanism of topic propagation and diffusion. Based on the study of the influence of topic correlation on the probability of topic transmission in multi-topic environment, this paper puts forward a multi-topic communication model, and studies the degree of topic correlation. The influence of the time interval of multi-topic occurrence and the initial transmission node of multi-topic on the spread of each topic and the whole public opinion. The simulation results show that the topic propagation model proposed in this paper can reflect the real topic evolution law. Has certain practical significance.
【學(xué)位授予單位】：南京郵電大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2015
【分類號(hào)】：D035;G206

【參考文獻(xiàn)】

相關(guān)期刊論文 前10條

1 蘭月新;董希琳;曾潤(rùn)喜;李昊青;張琦;;信息異化視角下網(wǎng)絡(luò)輿情衍生效應(yīng)模型研究[J];情報(bào)雜志;2015年01期

2 陳卓群;;基于共詞網(wǎng)絡(luò)的社交媒體話題演化分析[J];情報(bào)科學(xué);2015年01期

3 曹麗娜;唐錫晉;;基于主題模型的BBS話題演化趨勢(shì)分析[J];管理科學(xué)學(xué)報(bào);2014年11期

4 林萍;黃衛(wèi)東;;基于LDA模型的網(wǎng)絡(luò)突發(fā)事件話題演化路徑研究[J];情報(bào)科學(xué);2014年10期

5 任立肖;張亮;杜子平;李楊;;復(fù)雜網(wǎng)絡(luò)上的網(wǎng)絡(luò)輿情演化模型研究述評(píng)[J];情報(bào)科學(xué);2014年08期

6 曾潤(rùn)喜;王晨曦;陳強(qiáng);;網(wǎng)絡(luò)輿情傳播階段與模型比較研究[J];情報(bào)雜志;2014年05期

7 趙旭劍;楊春明;李波;張暉;金培權(quán);岳麗華;戴文鍇;;一種基于特征演變的新聞話題演化挖掘方法[J];計(jì)算機(jī)學(xué)報(bào);2014年04期

8 陳福集;陳婷;;基于SEIRS傳播模型的網(wǎng)絡(luò)輿情衍生效應(yīng)研究[J];情報(bào)雜志;2014年02期

9 黃衛(wèi)東;陳凌云;吳美蓉;;網(wǎng)絡(luò)輿情話題情感演化研究[J];情報(bào)雜志;2014年01期

10 于洪;楊顯;;微博中信息傳播特性分析及模型仿真研究[J];系統(tǒng)仿真學(xué)報(bào);2013年12期

相關(guān)博士學(xué)位論文 前1條

1 周杰;復(fù)雜系統(tǒng)中的信息傳播研究[D];華東師范大學(xué);2008年

，

本文編號(hào)：1849099