Research on Text Mining of Material Science Based on Natural Language Processing

Facing large-scale and rapidly growing material science literature data, text mining has become a research hotspot of material science. In recent years, natural language processing technology and machine learning methods have become the main technical means of text mining in materials science. The main task of text mining is to transform unstructured text data into structured material data by information extraction methods such as Named Entity Recognition and entity relationship extraction. This research proposes a general solution framework for material information extraction tasks, and introduces the main concepts and processes of text processing, text annotation, entity relationship extraction, etc., and discusses the current research progress and possible future research directions.