Computer subject evaluation needs to summarize the overall information of the subject. It relies too much on expert experience, and the reuse degree of previous subject evaluation information is low. In response to this problem, this study proposes a method for constructing a knowledge graph of computer subject evaluation. The method constructs a knowledge graph model based on the CIR model, designs an unsupervised named-entity and relation extraction method based on dependency parsing for text data and a combination model of data flow for table data to extract knowledge, and visualizes the knowledge graph using Neo4j. This model can provide ideas and references for the construction of knowledge graphs of other subjects.