Patent Application Text Pre-processing for Patent Examination Procedure

This paper is devoted to the development of text pre-processing module applied in natural language for statistical and semantic analysis. Due to increasing the number of incoming requests, the term of applications examination is sometimes up to several years. Therefore, there is a need to develop a variety of decision support systems that would accelerate the process of patent examination. We divided patents claims on simple sentences and build on their basis of semantic networks, followed by simplification. Implementation of the system will reduce the timing of the examination on the merits. The paper includes an overview of modern systems and models, description of the proposed natural language text pre-processing model, a description of the process automation and a general description of the developed system. Relevance of the developed system consists in the necessity standardize the patent applications text for more convenient to search for similar patents and analysis.