加權(quán)社交網(wǎng)絡(luò)中權(quán)重序列與最短路徑隱私保護(hù)算法研究.pdf

1、學(xué)位論文獨(dú)創(chuàng)性聲明本人聲明所呈交的論文是本人在導(dǎo)師指導(dǎo)下進(jìn)行的研究工作及取得的研究成果。據(jù)我所知，除了文中特別加以標(biāo)注和致謝的地方外，論文中不包含其他人已經(jīng)發(fā)表或撰寫過的研究成果， 也不包含已獲得江西理工大學(xué)或其他教育機(jī)構(gòu)的學(xué)位或證書而使用過的材料。 與我一同工作的同志對本研究所做的任何貢獻(xiàn)均已在論文中做了明確的說明并表示謝意。申請學(xué)位論文與資料若有不實(shí)之處，本人承擔(dān)一切相關(guān)責(zé)任。研究生簽名： 時(shí)間： 年 月 日學(xué)位論文版權(quán)使用授權(quán)書本

2、人完全了解江西理工大學(xué)關(guān)于收集、保存、使用學(xué)位論文的規(guī)定：即學(xué)校有權(quán)保存按要求提交的學(xué)位論文印刷本和電子版本， 學(xué)校有權(quán)將學(xué)位論文的全部或者部分內(nèi)容編入有關(guān)數(shù)據(jù)庫進(jìn)行檢索， 并采用影印、 縮印或掃描等復(fù)制手段保存、 匯編以供查閱和借閱； 學(xué)校有權(quán)按有關(guān)規(guī)定向國家有關(guān)部門或者機(jī)構(gòu)送交論文的復(fù)印件和電子版。 本人允許本學(xué)位論文被查閱和借閱， 同意學(xué)校向國家有關(guān)部門或機(jī)構(gòu)送交論文的復(fù)印件和電子版，并通過網(wǎng)絡(luò)向社會公眾提供信息服務(wù)。保密的學(xué)位論

3、文在解密后適用本授權(quán)書。學(xué)位論文作者簽名（手寫）： 導(dǎo)師簽名（手寫）：簽字日期： 年 月 日 簽字日期： 年 月 日AbstractIIAbstractWith the changes of people’s lifestyles and communication methods, the socialnetwork covers a large number of related information such as indivi

4、duals andenterprises on the basis of information technology. Due to the analysis and applicationof various fields in scientific research and data mining of data in social networks willlead to the leakage of a large numbe

5、r of users' privacy. Among them, due to theweight of the weighted network and the leakage of users' privacy caused by theshortest path, it is gradually becoming the focus and difficulty of the research onprivacy

6、protection. The attacker will use the edge weight sequence to identify thenode. At the same time, the existing methods for privacy protection of the shortestpath have great changes to the network structure and cannot gua

7、rantee the validity ofsocial network data. Based on the proposed problem, this paper proposes a moreeffective defense algorithm for the weighted sequence and the attribute attacks of theshortest path in weighted social n

8、etworks. The main contents of this paper are asfollows:1. Weighted privacy protection of existing weighted social networks is mainlybased on the use of Gaussian multiplication and histogram anonymity methods toconstruct

9、the network, lack of research on privacy leakage caused by the weightedsequence of edges, and insufficient consideration of data validity. Based on this, thispaper will use the vector set to combine the distance between

10、the inner diameter of thegroup and the relative distance of the nodes to achieve the privacy protection of theweight sequence. The algorithm can guarantee data validity of the data and resist theweight sequence attack.2.

11、 On the basis of k ? anonymity, privacy protection is established by addingedges or deleting edges to construct k ? shortest path. This method has a greatinfluence on the structure of social networks, and the shortest pa

12、th privacy levelbetween each node pair only is k . It can only be said. Based on this, this articleclassifies all edges in the network into three categories: unvisited edges ? ? NE ,partially accessed edges ? ? PE and mu

13、st-accessed edges ? ? ME , and adjusts the edgesby the weight ratio policy. If no k paths fulfills the requirement of at least k theshortest path, the algorithm returns the current result directly and uses the privacylev