Research On Entity Extraction For Animal Food Safety Hazards

In order for the government and food safety supervision departments to quickly and timely give targeted countermeasures when animal food safety incidents occur,it is necessary to quickly and accurately find valuable entity information in animal food safety incident reports expressed in natural language.Accurate identification and extraction of these entity knowledge information becomes the key.This article uses animal food safety incidents in the past two decades from 2001 to2020 as the research corpus,and uses crawling tools to crawl the animal food safety incident news on the Internet for the past two decades,Weibo,etc.,to define the animal food safety incident news materials involved The knowledge entity vocabulary in this field is labeled and the appropriate features and combination of features are selected,and the feature template that best reflects the performance of the recognition model is formulated.Through multiple animal food safety incident corpus experiments and ten cross-checks,the knowledge of animal food safety incidents is verified In entity recognition and extraction,the CRF++-based automatic recognition and extraction model constructed in this paper has the best performance.Finally,based on the conditional random field knowledge entity recognition model,a three-tier knowledge entity extraction tool based on the CRF++ model is constructed to realize the application of animal food safety event entity extraction.