Design And System Implementation Of Link Prediction Algorithm In Social Networks

With the rapid development of information and the continuous improvement of people’s material living standards,especially the widespread use of the Internet,network socialization has become daily,it promotes communication and communication between people,and with the continuous evolution of social networks,a person’s image on the Internet will also become more complete through the communication between people and the tags attached to them.A social network is a group of socially related points connected by one or more relationships.Points or network members are units connected by relationships.The complex network is used as an analysis tool to study the link relationship containing a huge amount of information.Link prediction is a key part of the field of data mining.It connects complex networks with information science.It can not only deal with the basic problems in information science,that is,the restoration and prediction of missing information but also mine the potential structural information in the network.Therefore,we have a deeper understanding of the evolution of the network structure.The Apriori algorithm in data mining is also a powerful tool for predicting the development trend of things based on frequent itemsets and association rules.The thesis first briefly summarizes the similarity indicators based on node attributes and three types of similarity indicators based on structural information in link prediction and conducts an in-depth exploration of theoretical knowledge based on their advantages and disadvantages.Then,the ideas and steps of the Apriori algorithm in data mining are explained in detail,and combined with case analysis,so that the theoretical knowledge is no longer boring,but vivid and easy to understand.However,because the classic Apriori algorithm is too time-consuming and labor-intensive,therefore,it is found through analysis of the improved Apriori algorithm that not only the number of database scans is reduced,but also no connection operation is required.And by optimizing the database,while improving the efficiency and quality of data mining,it also improves the performance of the system.The campus service and friend recommendation system platform built using the SpringMVC framework solves the "difficult and incurable diseases" in campus services,such as the problem of finding things and retrieving lost things.For problems existing in campus construction,you can anonymously make suggestions or express opinions.And finally,select users with similar attributes through node similarity and combine the Apriori algorithm to mine strong association rules to achieve their friend recommendation,which truly reflects the convenience of teachers and students.