The identification of patent information has gradually become the focus of global attention. Patent information plays a very important role in the implementation of intellectual property strategy, the construction of an innovative country and the development of strategic emerging industries. In this paper, a zero-expansion patent data processing method based on genetic algorithm is adopted, which can effectively overcome the “zero distance” phenomenon that may appear in patent data and analyze patent data more accurately, thus effectively solving the error problem caused by too many zeros. The test of zero expansion patent data analysis system shows that the prediction of the number of patent applications for zero expansion model based on genetic algorithm shows that in the 19th month, the predicted data is 257 and the historical data is 248, the difference between them is 9, and the prediction difference is very low.