With the continuous development of the era of big data and the corresponding technical means, the technical demand for data mining and analysis research and judgment in maintaining food safety is also increasing, and how to combine the trend of The Times to strengthen the fight against food safety crimes and innovate the means of handling cases is particularly important. Therefore, in order to further improve the ability and efficiency of relevant departments in investigating food cases, this paper, based on actual case data, uses big data police analysis technology as a means to build a clue analysis model for illegal addition of toxic and harmful substances to food crimes, establishes a multi-channel data collection system, preprocesses case texts and defines standards, and innovatively uses knowledge ontology database construction technology. Then it is classified according to the business theme, and the overall framework of the analysis model of food crime clues of illegal addition of toxic and harmful substances is reused vertically, and the horizontal expansion is flexible, and the analysis, judgment and comprehensive treatment ability of hot issues of food crime of illegal addition of toxic and harmful substances in the disposal of food safety crimes the model is greatly improved.