The goal was to develop a software module capable of recognizing and classifying chemical information, and then converting the resulting data into a single SMILES format for subsequent comparison. The work was aimed at creating a software product capable of automating the processes of analyzing and comparing formulas from patent documents, which can be very useful in the field of scientific research and intellectual property. The study describes the stages of developing a software module for analyzing various types of representations of chemical formulas extracted from patent documents, creating a common database of patents, comparing chemical compounds of different representations with each other to identify their correspondence or inconsistency with each other. The document reveals the approaches and technologies used to solve the problems of classifying the formats of representations of chemical compounds, their analysis, as well as comparisons with each other. The significance of the work lies in the developed algorithms for parsing a patent document, comparing chemical formulas, comparing patents based on their chemical formulas, and developed software module, and we analyze chemical formulas in patent documents.