The Convergence Of Causal Relationship In NLU And Its Application On Test And Measurement

The application of NLU technology in test and measurement is one of the most important and difficulty application of A.I. technology. This text tries to use the Cluster Model of Chinese natural language discourse understanding technology to comprehend the needs analysis of the test and measurement, focusing on chapters in the convergence of causal relationship and eventually through natural language to express the user needs to understand and analyze, realize the good exchange between the system and customers, provide the support for the following design.Firstly, methods of semantic analysis and expressing of knowledge are carefully introduced in paper. Combining with the characteristic of natural language understanding system which is based on knowledge, conceptual dependency theory and framing theory are chosen as knowledge representation method to express knowledge using the form of template. Secondly, the expression method of causal relationship is divided into hypotaxis and parataxis, and identifiers representing causality are discussed. Then, the classification of causation is discussed. Thirdly, the cluster model of causation expressed by parataxis using causal knowledge is established. Then, combining with time sequence, the cluster model of causation expressed by hypotaxis is introduced. At last, it is applied to the test and measurement domain, and its result turns to be satisfied.