本文選題：在線網(wǎng)絡(luò)社區(qū) + 數(shù)據(jù)采集��； 參考：《哈爾濱工程大學(xué)》2012年碩士論文

【摘要】：最近幾年，隨著互聯(lián)網(wǎng)的快速發(fā)展，網(wǎng)絡(luò)社區(qū)用戶數(shù)量急劇增加，大規(guī)模在線網(wǎng)絡(luò)社區(qū)逐漸形成。在線網(wǎng)絡(luò)社區(qū)用戶已經(jīng)成為互聯(lián)網(wǎng)的重要用戶群，在線網(wǎng)絡(luò)社區(qū)應(yīng)用的重要性已經(jīng)可以和搜索引擎、即時(shí)通訊工具等相提并論。SNS社區(qū)、微博等在線網(wǎng)絡(luò)社區(qū)作為新興的主流互聯(lián)網(wǎng)應(yīng)用，已經(jīng)聚集了大量的用戶。隨著智能手機(jī)等移動(dòng)終端的發(fā)展，網(wǎng)絡(luò)社區(qū)用戶數(shù)量更是快速增長(zhǎng)。 網(wǎng)絡(luò)社區(qū)的用戶關(guān)系是一種復(fù)雜網(wǎng)絡(luò)，復(fù)雜網(wǎng)絡(luò)的社區(qū)結(jié)構(gòu)發(fā)現(xiàn)一直都是研究熱點(diǎn)。通過(guò)對(duì)現(xiàn)有社區(qū)結(jié)構(gòu)發(fā)現(xiàn)算法的研究，發(fā)現(xiàn)現(xiàn)有的社區(qū)結(jié)構(gòu)發(fā)現(xiàn)算法關(guān)于社區(qū)結(jié)構(gòu)或者搜索目標(biāo)的定義存在過(guò)緊或者過(guò)松的問(wèn)題，本文提出一種新的社區(qū)結(jié)構(gòu)定義方法；在該方法的基礎(chǔ)上，本文還提出了一種基于邊介數(shù)的快速社區(qū)結(jié)構(gòu)發(fā)現(xiàn)算法。該算法具有新的社區(qū)結(jié)構(gòu)發(fā)現(xiàn)策略，并且針對(duì)網(wǎng)絡(luò)中不同的社區(qū)結(jié)構(gòu)都能夠快速收斂。同時(shí)，該算法還提出了關(guān)于重疊節(jié)點(diǎn)的發(fā)現(xiàn)策略，解決了現(xiàn)有很多社區(qū)結(jié)構(gòu)發(fā)現(xiàn)算法把社區(qū)結(jié)構(gòu)發(fā)現(xiàn)簡(jiǎn)單當(dāng)做圖形分割的問(wèn)題，較好地解決了某些節(jié)點(diǎn)同時(shí)屬于多個(gè)社區(qū)的情況。 網(wǎng)絡(luò)社區(qū)隨著時(shí)間的推移不斷變化，本身具有很強(qiáng)的動(dòng)態(tài)性。社區(qū)的節(jié)點(diǎn)數(shù)量會(huì)增加或者減少，節(jié)點(diǎn)間關(guān)系會(huì)變得更加緊密或者疏遠(yuǎn)；與此同時(shí)，網(wǎng)絡(luò)社區(qū)中滿足社區(qū)結(jié)構(gòu)的節(jié)點(diǎn)簇的規(guī)模以及它們是否還構(gòu)成社區(qū)都會(huì)發(fā)生變化。為了預(yù)測(cè)網(wǎng)絡(luò)社區(qū)在未來(lái)的演化狀況，需要根據(jù)網(wǎng)絡(luò)社區(qū)在過(guò)去表現(xiàn)出來(lái)的演化特征，建立能夠反映社區(qū)演化規(guī)律的模型。本文針對(duì)在線網(wǎng)絡(luò)社區(qū)演化不規(guī)則的情況，根據(jù)在過(guò)去演化過(guò)程中各特征的變化率來(lái)對(duì)未來(lái)進(jìn)行預(yù)測(cè)，彌補(bǔ)了現(xiàn)有模型對(duì)規(guī)則特征變化的依賴。模型除了關(guān)注網(wǎng)絡(luò)整體的演化外，還關(guān)注了網(wǎng)絡(luò)中社區(qū)的演化規(guī)則，確保模型的演化結(jié)果跟真實(shí)網(wǎng)絡(luò)社區(qū)具有相近的社區(qū)結(jié)構(gòu)。 在線網(wǎng)絡(luò)社區(qū)結(jié)構(gòu)發(fā)現(xiàn)與演化分析基礎(chǔ)平臺(tái)是一個(gè)綜合性基礎(chǔ)分析平臺(tái)。論文對(duì)在線網(wǎng)絡(luò)社區(qū)數(shù)據(jù)提取方法進(jìn)行了研究，，針對(duì)在線網(wǎng)絡(luò)社區(qū)數(shù)據(jù)普遍存在限制的情況，提出了一種在線網(wǎng)絡(luò)社區(qū)受限信息的提取方法；同時(shí)，在線網(wǎng)絡(luò)社區(qū)的部分?jǐn)?shù)據(jù)是動(dòng)態(tài)生成，并不是直接寫入靜態(tài)網(wǎng)頁(yè)，針對(duì)該情況，提出了動(dòng)態(tài)網(wǎng)頁(yè)數(shù)據(jù)的提取方案。最后，結(jié)合論文中提出的社區(qū)結(jié)構(gòu)發(fā)現(xiàn)方法、演化分析技術(shù)以及社區(qū)數(shù)據(jù)提取方法，實(shí)現(xiàn)了在線網(wǎng)絡(luò)社區(qū)結(jié)構(gòu)發(fā)現(xiàn)和演化分析基礎(chǔ)平臺(tái)，第五章對(duì)該平臺(tái)的設(shè)計(jì)和實(shí)現(xiàn)方法進(jìn)行了介紹，平臺(tái)能夠?yàn)樵诰�網(wǎng)絡(luò)社區(qū)的結(jié)構(gòu)發(fā)現(xiàn)和演化分析提供基礎(chǔ)性的支持。
[Abstract]:In recent years, with the rapid development of the Internet, the number of online community users has increased dramatically, and a large scale of online communities have gradually formed.Online network community users have become an important group of Internet users. The importance of online network community applications can be compared with search engines, instant messaging tools, and so on.Weibo and other online communities as a new mainstream Internet applications, has gathered a large number of users.With the development of mobile terminals such as smart phones, the number of network community users is growing rapidly.The user relationship of the network community is a kind of complex network, and the community structure discovery of the complex network is always the research hotspot.Through the research of the existing community structure discovery algorithm, it is found that the existing community structure discovery algorithm has the problem of being too tight or too loose in the definition of community structure or search target. A new community structure definition method is proposed in this paper.On the basis of this method, a fast community structure discovery algorithm based on edge mediums is proposed.The algorithm has a new community structure discovery strategy and can converge rapidly for different community structures in the network.At the same time, the algorithm also puts forward the discovery strategy of overlapping nodes, which solves the problem that many existing community structure discovery algorithms treat community structure discovery simply as graph segmentation.It solves the problem that some nodes belong to multiple communities at the same time.Network community changes with the passage of time, itself has a strong dynamic.The number of nodes in the community will increase or decrease, and the relationship between the nodes will become closer or estranged. At the same time, the size of the nodes that satisfy the community structure in the network community and whether they will also constitute the community will change.In order to predict the evolution of the network community in the future, it is necessary to establish a model that can reflect the evolution law of the network community according to the evolution characteristics of the network community in the past.Aiming at the irregular evolution of the online network community, this paper predicts the future according to the change rate of each feature in the past evolution process, which makes up for the dependence of the existing model on the rule feature change.The model not only pays attention to the evolution of the whole network, but also pays attention to the evolution rules of the community in the network, so as to ensure that the evolution result of the model is similar to that of the real network community.The online network community structure discovery and evolution analysis foundation platform is a comprehensive basic analysis platform.In this paper, the method of online community data extraction is studied, and a method of extracting the limited information of online network community is put forward in view of the limitation of online network community data, at the same time,Part of the data of the online network community is generated dynamically, but not written directly to the static web page. In view of this situation, a scheme of extracting the dynamic web page data is proposed.Finally, combined with the community structure discovery method, evolution analysis technology and community data extraction method proposed in the paper, the online network community structure discovery and evolution analysis platform is realized.The fifth chapter introduces the design and implementation of the platform. The platform can provide the basic support for the structure discovery and evolution analysis of the online network community.
【學(xué)位授予單位】：哈爾濱工程大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2012
【分類號(hào)】：TP393.09

【參考文獻(xiàn)】

相關(guān)期刊論文 前10條

1 汪小帆;劉亞冰;;復(fù)雜網(wǎng)絡(luò)中的社團(tuán)結(jié)構(gòu)算法綜述[J];電子科技大學(xué)學(xué)報(bào);2009年05期

2 李濤;裴文江;;針對(duì)重疊社團(tuán)結(jié)構(gòu)的復(fù)雜網(wǎng)絡(luò)多靶向攻擊策略[J];北京郵電大學(xué)學(xué)報(bào);2010年03期

3 熊中敏;黃冬梅;;可多邊并行移出的社團(tuán)發(fā)現(xiàn)方法[J];計(jì)算機(jī)工程;2009年12期

4 朱小虎;宋文軍;王崇駿;謝俊元;;用于社團(tuán)發(fā)現(xiàn)的Girvan-Newman改進(jìn)算法[J];計(jì)算機(jī)科學(xué)與探索;2010年12期

5 何東曉;周栩;王佐;周春光;王U

本文編號(hào)：1737288