Research On Conceptual Graph Indexing Of Construction Standards

In China, behavior in engineering construction is regulated through adoption of reasonable standards and proper implementation. Construction standards play a crucial role in ensuring construction safety and quality. Proper adoption of engineering construction standard can promote the progress of science and technology, enhance the construction quality and boost social and economic development. However, the efficiency of inspection suffers when the inspectors have to search relevant standard documents literally on the spot, which gives rise to waste of time and potential mistakes. The solution to the problem of construction standard retrieval, and the strengthening of the computer support would definitely promote review efficiency, ensure the implementation level, and improve the construction process.As the industry codes, engineering construction standards have normalized format, various classification, rigorous logic, clear structure, concise items and specific regulations. A demand analysis of construction standards retrieval in China is proposed in this paper based on above characteristics and current retrieval methods. Conceptual graph is a semantic retrieval method, more efficient and accurate than keywords retrieval method. Accordingly, this paper puts forward the construction of conceptual graph to achieve the purpose of engineering construction standard retrieval, and focus on the process of automatic construction of the conceptual graph of standards.During the construction of conceptual graph, this paper adopts the maximum entropy model, and machine learning based on context information to achieve the goal of identifying the basic noun phrases in the standard clauses. By distinguishing the different rules of mandatory standards and recommended standards, different methods of recognizing the predicate center word are established. The sentence where the predicate center word locates is called the backbone template. This paper constructs a conceptual graph of the backbone template first. Then construct the conceptual graph of non-backbone template. Finally, merge them to form a complete conceptual graph. After the construction of standard conceptual graph, this paper analyzes the influence factors of the construction process through factor analysis method. We summarized and divided the indexes into three common factors, conceptual graph breadth, conceptual graph relationship and information amount. Then combined the three factors into analysis of automatic construction of conceptual graph, and put forward optimization path of automatic conceptual graph which laid the foundation for research of further application of conceptual graph in the field of engineering construction standards.