Food Safety Information Collection And Processing Based On Machine Learning

Internet information is messy and the amount of data is huge,and a large amount of food safety-related news and quality inspection data are generated every day.However,manual sorting efficiency is low,and large-scale data lacks effective combing,and it is difficult to dig out an effective basis for food safety regulatory assistance decision-making.As the country pays more attention to food safety,research how to realize automatic collection and structured information extraction of text data in the field of food safety,and obtain valuable information,which can better assist in food safety early warning prediction and prevention in major events Control work.This article takes the historical food safety texts on the Internet media as the research object,and the work done includes the following aspects:1)Data collection.Based on the web crawler method that combines the Scrapy framework and the Beautiful Soup parser,it grabs relevant texts of food safety news and food safety cases from web pages.2)Information processing.Perform pre-processing on the collected raw data such as deduplication,Chinese word segmentation,and stop word removal.The text is classified based on convolutional neural networks to obtain a food safety data set.This paper proposes an improved named entity method,named entity recognition model based on BERT and adversarial training extracts information from food safety data sets and organizes them into structured information.3)Research on artificial intelligence-based food safety auxiliary decision-making system.The Apriori algorithm is used to mine and analyze the five important attributes of food to obtain association rules and improve the knowledge base of the expert system.This paper design a reasonable reasoning machine,and realize decision support for food risk levels.