Font Size: a A A

Research On Key Technologies Of Anchor Link Identification In Heterogeneous Social Networks

Posted on:2020-01-03Degree:MasterType:Thesis
Country:ChinaCandidate:F YangFull Text:PDF
GTID:2428330575461924Subject:Computer Science and Technology
Abstract/Summary:PDF Full Text Request
At present,people mainly share their lives through social networks and build more friends through social networks.With the continuous development of social networks,a single social network has gradually evolved into a heterogeneous social network,which in turn creates a huge social network structure in the network environment.Entities in heterogeneous social networks mainly include users,locations,topics,and the like.Entity anchor link identification in heterogeneous social networks is an important solution for cross-network user recommendation and network information diffusion.Heterogeneous social network user entity anchor link identification key technology and location entity anchor link identification key technology is the main research direction of this paper.Firstly,for the problem of heterogeneous social network user entity anchor link identification,this paper proposes a user anchor link recognition framework based on similarity judgment.The existing methods mainly use the user's natural attributes and behavior attribute data to establish a supervised or semi-supervised recognition model,in order to obtain higher recognition accuracy.However,the premise of the implementation of such methods is that a large number of user anchors are required to link prior knowledge,which is often difficult to obtain accurately in actual social networks.Aiming at the above problems,this paper proposes a user entity anchor link recognition algorithm based on similarity judgment based on unsupervised learning mode.The algorithm establishes anchor link recognition model by combining user natural attributes and improved user relationship analysis.The experimental results show that the user entity anchor link recognition algorithm based on similarity judgment proposed in this paper has certain advantages in recognition accuracy.Secondly,aiming at the identification problem of heterogeneous social network location entity anchor link,this paper proposes a location anchor link recognition algorithm based on anchor link user.At present,there are few researches on location anchor links.The main research and development method is to realize the identification of location entity anchor links through the identification of multiple entity anchor links in social networks.These methods do not take into account the user-location relationship,and analyze all users and locations independently,resulting in low location anchor link recognition.In view of the above problems,this paper is based on the heterogeneous social network user anchor link,and comprehensively considers the location attribute and the relationship between the user and the location,strengthens the relationship between the user and the location,and establishes the recognition model of the location entity anchor link.The experimental results show that the location entity anchor link recognition algorithm proposed in this paper has a certain degree of improvement in recognition accuracy compared with the existing methods.
Keywords/Search Tags:heterogeneous social network, user anchor link, location anchor link, similarity judgment, unsupervised learning
PDF Full Text Request
Related items