Ontology-based Research And Implementation Of Early Warning Platform From Food Complaint Texts

Food safety is related to the national health and social stability, and many other problems.In recent years, many food safety incidents occur frequently such as Mengniu corporates’poisoning, clenbuterol, beef extract and so on. All of these draw much attention to improvingfood safety. The occurrence of food safety problems not only makes our country suffer severeeconomic losses, but also affect consumer trust in government, and even cause threat to socialstability and long-term development of the country. So how to establish a long-term andeffective mechanism of food safety supervision system to ensure the long-term stable foodsafety regulation is an urgent problem which needs to explore and solve the problem.Along with the rapid development of Internet and gradually perfect food safetyregulatory system, consumer can be involved in the regulation of food safety by complainingvia the Internet. Due to the increase in the number of complaints documents accompanying,the question about how to classify effectively from the vast amounts of complaints,desultorily document to meet the needs of different users for different information comes up.The traditional document processing system is difficult to realize the semantic expansion, andit is hard for the users to find the information they need from large document accurately.This paper proposes a food complaint text early warning method based on the guidanceof ontology, and establishes a scientific and reasonable system of early warning, builds andimproves the food security early warning platform. All of those make this paper play asupplementary role in the research content of food safety regulators. Based on traditionalearly warning system, this paper constructs food safety complaints warning platform model,and builds the food domain ontology, and expands food complaint document semantics tohighlight the implicit semantics and improve the document’s semantic accuracy. Through thecalculation of similarity of theme characteristic vector and text categorization constructingclassifier, make the automatic classification of food complaint documents based on the themecome true for those which are not correctly classified documents for unsupervised clustering.which can be the purpose of food safety alarm. Then, it is possible to use complaints aboutfood safety for rapid and accurate text data processing, make the food safety regulatorsunderstand the food safety hidden trouble in time to protect consumers’ rights and interests.