Research on Construction and Evolution of Relational Graph of the Science and Technology Data

Abstract

In recent years, the amount of scientific and technological data has been increasing, and the requirements for data analysis have been continuously improved. Among of them, the mining of data association relations and the construction and evolution of relational models are the research hotspots in recent years. How to obtain entities and their relationships from massive data, and then use the relational model to construct the relational graph and support the evolution of the graph is the key problem to be solved. Aiming at the above problems, this paper proposes a method for constructing and evolving relational graphs based on structured data, including source data preprocessing, entity and attribute identification, information extraction, relationship matching, and relational graph construction and evolution. The evolution and update algorithm of the graph based on Simhash and Hamming distance is proposed, and the update strategy of the entity similarity in the graph is analyzed. Combined with experimental verification, this method can automatically realize the extraction of structured data, the construction of relational graph and the evolution and update process of the graph, and the accuracy of the results was about 97%.