Research On Auxiliary Decision-Making Method Of E-Government Project Review Based On Knowledge Graph

In recent years,with the development of government informatization,the demand for e-government project construction has increased greatly.This situation has brought great pressure to each evaluation link in the construction and management of e-government projects.Affected by the limitations of experts’ own knowledge,the current method of reviewing only by manpower can no longer meet the needs of project construction management.As a basic method to use domain knowledge to scientifically assist decision-making,knowledge management is also a scientific method for the construction and management of e-government projects.Expert review of e-government project construction scheme is actually a process of using expert knowledge to assist decision-making for management rules.Therefore,the use of knowledge management technology to realize the auxiliary decision-making of e-government project management has theoretical research and practical application value.As a semantic network,knowledge graph can scientifically express domain knowledge and support machine reasoning based on knowledge association.It is an effective tool to assist decision-making in the construction and management of e-government projects.Using the knowledge graph,fragmented project construction plan data can be instantiated into visual project knowledge,and relevant reasoning can be completed based on project management rules to realize tacit knowledge discovery.Based on the above logic,this paper proposes an auxiliary decision-making method for e-government project review based on knowledge graph,and completes the following main research work:(1)In order to scientifically express the management knowledge of e-government project construction,this paper constructs the knowledge ontology of e-government project approval.In ontology implementation,based on knowledge engineering technology,entity concept knowledge extraction and verification is carried out by using the fusion attention mechanism and BERT-Bi LSTM-CRF named entity recognition model.Further,after preprocessing the information in the e-government project scheme,the construction of the e-government project approval knowledge graph is completed by instantiating the ontology.(2)In order to integrate the deep-level information of review comments,this paper proposes a fine-grained mining method for expert review comments.This method combines short syntactic structure and dependency syntactic analysis to construct multiple rules to extract evaluation objects and evaluation words.With the help of sentiment dictionary and semantic similarity,the sentiment polarity of evaluation elements is judged,and the entity expansion of the knowledge graph of e-government project approval is realized.(3)Based on the knowledge graph of e-government project approval,this paper proposes a visual analysis method of knowledge reasoning oriented to practical application scenarios.The method uses the intuitive visual chart to solve the problems of system construction target review,repeated construction problems,and difficulty in mining historical declarations,and achieves objective,scientific and efficient auxiliary decision-making requirements.