基于侵权社区挖掘的P2P网络版权内容传播研究
发布时间：2019-06-04 01:19
【摘要】：P2P (Peer-to-Peer)对等网络因其非中心化,自组织,动态可扩展等特点,在资源共享和内容传播方面得到广泛应用的同时,由于缺少一种严格的内容授权和第三方监管机制,造成了不良资源和盗版侵权内容的泛滥传播,正面临着一场严重的信任危机。为了深入挖掘用户节点之间版权内容共享关系,深刻反映P2P网络中版权内容的传播与分布,最终有效地进行版权内容的监管与侵权内容传播的反制,本文引入社区发现理论,提出基于行为者网络模型的侵权社区挖掘方法及监管系统。以网络中社区结构为监管对象,通过对社区内版权内容分布的研究,对社区的威胁度进行量化分析,并探索侵权社区内容传播的反制策略。 通过对对等网络中侵权行为研究发现,侵权者之间相互共享和传递大量相似版权内容,这种基于内容共享和消息传递的节点组成社区的结构。社区发现技术是数据挖掘和探测性分析技术在复杂网络中极为重要的研究,在社会网络分析领域、数据挖掘和数据库技术领域、统计学和机器学习等领域有广泛的应用。 本文基于行为者网络(Actor-network model)模型构建侵权社区的研究方法包含：步骤1用于网络数据的预处理；步骤2构建行为者网络模型；步骤3完成侵权社区挖掘。其中第二步包含内容相似度图的构建、节点关系联接和消息筛选及过滤等步骤。在内容相似度图构建模块,本文首先对内容元数据进行本体描述,然后进行相似度的计算,确保版权内容描述的准确性和挖掘的全面性；节点关系联接模块,记录了消息时间戳、消息频率、消息类型在内的性质完成消息权重的量化；消息筛选及过滤过程大大简化了P2P网络数据,为实施社区挖掘算法做到提前优化。第三步是在研究社区挖掘经典算法Girvan-Newman (GN)算法的基础上提出有效的侵权社区挖掘的ANMGN算法,引入社团结构增益缩短算法时间。 目前,版权内容的监管一种是采取事后处理的方式,即发现盗版侵权内容后,通过采取阻止下载、文件清除等技术阻断侵权内容的传播；另一种是内容格式加密技术,符合版权方许可的用户才能下载,而这种技术并不是开源的,只适用于特定软件,没有普遍性。本文提出的基于侵权社区挖掘方法实现了对版权内容的传播流向和分布检测与监管,达到区域预警及时对侵权社区予以打击。为了解决P2P网络中的盗版侵权内容传播问题,本文在侵权社区发现的基础上,分析了社区的威胁度,对侵权社区内版权内容传播提出两种反制机制：断点反制策略和断边反制策略。版权内容传播时出现经过次数最多的节点,我们称之为桥接点,断点反制策略就是通过找到社区桥接点,在必要时删除这些桥接点,以达到阻碍版权内容向外传播；社区之间的连接边,我们称之为关键边,断边反制策略通过找到网络中的关键边,在必要时产出这些边,破坏版权内容的传播。
[Abstract]:P2P (Peer-to-Peer) peer-to-peer network has been widely used in resource sharing and content dissemination because of its decentralized, self-organization, dynamic scalability and other characteristics, at the same time, due to the lack of a strict content authorization and third-party supervision mechanism, As a result of the spread of bad resources and pirated infringing content, it is facing a serious crisis of trust. In order to deeply excavate the relationship of copyright content sharing among user nodes, reflect the dissemination and distribution of copyright content in P2P network, and finally effectively carry out the supervision of copyright content and the counter-system of infringing content dissemination, this paper introduces the theory of community discovery. The method and supervision system of tort community mining based on actor network model are proposed. Taking the community structure in the network as the regulatory object, through the study of the distribution of copyright content in the community, this paper makes a quantitative analysis of the threat of the community, and explores the countermeasures for the dissemination of content in the infringing community. Through the study of infringement in peer-to-peer network, it is found that infringers share and transfer a large number of similar copyright content to each other, which is a community structure based on content sharing and message transmission. Community discovery technology is a very important research on data mining and exploratory analysis technology in complex networks. It has been widely used in the field of social network analysis, data mining and database technology, statistics and machine learning. In this paper, the research methods of building tort community based on actor network (Actor-network model) model include: step 1 for network data preprocessing; step 2 to build actor network model; step 3 to complete infringement community mining. The second step includes the construction of content similarity graph, node relational join, message filtering and so on. In the construction module of content similarity diagram, this paper first describes the ontology of content metadata, and then calculates the similarity to ensure the accuracy of copyright content description and the comprehensiveness of mining. The node relational connection module records the nature of message timestamp, message frequency and message type to complete the quantification of message weight. The process of message filtering and filtering greatly simplifies P2P network data and optimizes the implementation of community mining algorithm in advance. The third step is to propose an effective ANMGN algorithm for tort community mining on the basis of studying the classical community mining algorithm Girvan-Newman (GN) algorithm, and introduce the community structure gain to shorten the algorithm time. At present, one of the supervision of copyright content is to deal with it after the event, that is, after the discovery of pirated infringing content, the dissemination of infringing content is blocked by technology such as preventing download, file removal and so on. The other is content format encryption technology, which can only be downloaded by users licensed by copyright. This technology is not open source, it is only suitable for specific software, and it is not universal. The method based on infringing community mining proposed in this paper realizes the detection and supervision of the dissemination direction and distribution of copyright content, and achieves the regional early warning to crack down on the infringing community in time. In order to solve the problem of the dissemination of pirated infringing content in P2P network, this paper analyzes the threat degree of the community on the basis of the discovery of the infringing community. This paper puts forward two kinds of countermeasures for the dissemination of copyright content in infringing community: the strategy of breaking point and the strategy of breaking edge. When the copyright content propagates, the node that passes through the most times, we call it the bridge point, the breakpoint reaction strategy is to find the community bridge point, delete these bridge points when necessary, in order to prevent the copyright content from spreading outward. The connecting edge between communities, which we call key edge, breaks the edge strategy by finding the key edge in the network and producing these edges if necessary, which destroys the dissemination of copyright content.
【学位授予单位】：山东大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TP393.02
本文编号：2492370
[Abstract]:P2P (Peer-to-Peer) peer-to-peer network has been widely used in resource sharing and content dissemination because of its decentralized, self-organization, dynamic scalability and other characteristics, at the same time, due to the lack of a strict content authorization and third-party supervision mechanism, As a result of the spread of bad resources and pirated infringing content, it is facing a serious crisis of trust. In order to deeply excavate the relationship of copyright content sharing among user nodes, reflect the dissemination and distribution of copyright content in P2P network, and finally effectively carry out the supervision of copyright content and the counter-system of infringing content dissemination, this paper introduces the theory of community discovery. The method and supervision system of tort community mining based on actor network model are proposed. Taking the community structure in the network as the regulatory object, through the study of the distribution of copyright content in the community, this paper makes a quantitative analysis of the threat of the community, and explores the countermeasures for the dissemination of content in the infringing community. Through the study of infringement in peer-to-peer network, it is found that infringers share and transfer a large number of similar copyright content to each other, which is a community structure based on content sharing and message transmission. Community discovery technology is a very important research on data mining and exploratory analysis technology in complex networks. It has been widely used in the field of social network analysis, data mining and database technology, statistics and machine learning. In this paper, the research methods of building tort community based on actor network (Actor-network model) model include: step 1 for network data preprocessing; step 2 to build actor network model; step 3 to complete infringement community mining. The second step includes the construction of content similarity graph, node relational join, message filtering and so on. In the construction module of content similarity diagram, this paper first describes the ontology of content metadata, and then calculates the similarity to ensure the accuracy of copyright content description and the comprehensiveness of mining. The node relational connection module records the nature of message timestamp, message frequency and message type to complete the quantification of message weight. The process of message filtering and filtering greatly simplifies P2P network data and optimizes the implementation of community mining algorithm in advance. The third step is to propose an effective ANMGN algorithm for tort community mining on the basis of studying the classical community mining algorithm Girvan-Newman (GN) algorithm, and introduce the community structure gain to shorten the algorithm time. At present, one of the supervision of copyright content is to deal with it after the event, that is, after the discovery of pirated infringing content, the dissemination of infringing content is blocked by technology such as preventing download, file removal and so on. The other is content format encryption technology, which can only be downloaded by users licensed by copyright. This technology is not open source, it is only suitable for specific software, and it is not universal. The method based on infringing community mining proposed in this paper realizes the detection and supervision of the dissemination direction and distribution of copyright content, and achieves the regional early warning to crack down on the infringing community in time. In order to solve the problem of the dissemination of pirated infringing content in P2P network, this paper analyzes the threat degree of the community on the basis of the discovery of the infringing community. This paper puts forward two kinds of countermeasures for the dissemination of copyright content in infringing community: the strategy of breaking point and the strategy of breaking edge. When the copyright content propagates, the node that passes through the most times, we call it the bridge point, the breakpoint reaction strategy is to find the community bridge point, delete these bridge points when necessary, in order to prevent the copyright content from spreading outward. The connecting edge between communities, which we call key edge, breaks the edge strategy by finding the key edge in the network and producing these edges if necessary, which destroys the dissemination of copyright content.
【学位授予单位】：山东大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TP393.02
【参考文献】
相关期刊论文 前5条
1 高琦,陈华钧;互联网Ontology语言和推理的比较和分析[J];计算机应用与软件;2004年10期
2 窦文,王怀民,贾焰,邹鹏;构造基于推荐的Peer-to-Peer环境下的Trust模型[J];软件学报;2004年04期
3 朱晓姝;;一种语义对等网模型[J];网络安全技术与应用;2006年03期
4 李秀莲;P2P软件使用用户版权侵权问题的解决对策[J];现代情报;2005年08期
5 徐德智;汪智勇;王斌;;当前主要本体推理工具的比较分析与研究[J];现代图书情报技术;2006年12期
，本文编号：2492370
本文链接：https://www.wllwen.com/guanlilunwen/ydhl/2492370.html
