Topic mining and semantic analysis of unsafe behavior based on miner big data

In order to prevent and control coal mine safety accidents, we should deeply mine the hidden information and knowledge of unsafe behavior data. Based on Python, LDA and NetDraw, 44069 pieces of unsafe behavior data of a large coal mine group in Shaanxi Province from 2017 to 2021 are selected for word segmentation and topic extraction, and the miner unsafe behavior semantic network diagram is drawn to analyze the centrality of miner unsafe behavior semantic network. 5 miners’ unsafe behavior high frequency topics and 3 miners’ unsafe behavior high incidence sites are obtained.