A Novel Approach for Patent Document Summarization

Abstract

Patent papers are valuable intellectual assets intended to safeguard individual interests and creativity. In a recent study field, patent retrieval helps to assist patent analysts in the collection, compilation, and review of patent records. Patent applications are extensive and full of scientific and legal terms, so it requires much time to translate, understand so review a single patent application, particularly for domain experts. It is also important to streamline this process and to assist the analysts in evaluating the relationship between the form and the patents obtained. Typical patent retrieval activities also involve analyzing whether various patent records are similar/different in several respects. This approach analyzes and summarizes a patent document in terms of similarities and differences. Patent documents are lengthy and consumes time for analysis; this approach automates the process and gives a comparative summary of multiple patent reports. The algorithm takes two patent documents as input and provides a comparative overview of the two patents. The final output is a summary of two patents that cover both discriminative and similar aspects of the documents.