Teaching Chinese Pattern Extraction and Its Knowledge Base Construction for Specific Domain Texts

Certain specialized component collocations and high-frequency expressions, known as Sentence Pattern Forms, can be used in textbook compilation and in-person teaching. The selection of Sentence Pattern Forms and the organization of knowledge points depend on the expert’s subjective opinion and lack the support of objective evidence. Manual extraction also limits the compilation’s effectiveness and thoroughness. This study creates a corpus of vocational textbooks on various majors. At the same time, we develop a knowledge base of extraction principles and provide corresponding algorithms for different Sentence Pattern Forms. Candidate expressions are automatically extracted and filtered by necessary manual screening. This method is more effective and scientific than pure manual extraction since it can cover more Sentence Pattern Forms that expert opinion cannot exhaust.