Query Algorithm Based Social Network Friends Relationship Diagram Research And Applications

In recent years, social network data has become a novel and increasingly important data type. Compared with traditional data types, social network data is more complicated and the semantics of expression is more abundant. They are reflected in:Firstly, the physical size of social network is often millions, sometimes tens of millions or billions, and in the worst the relationship between the entities is square to the size of the entities, increasing the scale of the data; Secondly, the relationships in social network are often uncertain, such as the strength, direction, type of the relationships, etc. So it is more difficult to query on social network data. For such a complex data type, it is necessary to find effective data organization and efficient query processing method to improve quality and efficiency analysis of social network data.This paper studies the social network friendship and builds a model for it. On this basis, graph query optimization technique is designed and used to conduct the shortest path query on friendships. Firstly, the existing shortest path query methods are introduced. Then a shortest path query method based on hyperbolic space is designed. This method embeds social network graphs into hyperbolic coordinate system, the query for the shortest path length between nodes in social graphs is then transformed to the computation of distance between nodes in coordinate system, so as to efficiently reduce the computational cost O(n+m) of using breadth-first search(BFS) to O(1). Experiments are done to verify the shortest path query method based on hyperbolic space having a high efficiency and accuracy. Finally, this method is applied to the computation of graph separation metrics, graph centrality and ranked social search, further validating the usefulness and accuracy of this method.