Abstract: With the advent of big data era, knowledge graph and data visualization technology present the data in a structured, visual way and establish a keyword-oriented knowledge system and render the relationship in a fast and clearly way. In this paper, a historical figures entity relationship visualization system has been established by means of data visualization and knowledge graph. In the system,the complex data are divided into character data and event data by data preprocessing. In the parsing stage, a label traversing method and a method based on weight of links are applied to the divided data respectively. With the layered B/S structure design, the system is based on the Nodejs platform in which a historical figures database is founded. The users can obtain the knowledge graphs of relevant historical figures and events according to distinct needs. This system provides a variety of interactive with good scalability and maintainability and makes contributions to comprehension and exploration of the data and relationship quickly by presenting the data in visual forms. To some extent, it owes reference value to research staff as well.

Key words: knowledge graph, entity relationship, data parse, data visualization