Patentssearcher

Abstract

There is an abundance of systems today to search for relevant patents, ranging from free ones like Google Patents (google.com/patents) to subscription ones like Delphion (delphion.com). After studying many existing systems, we found that they all apply general-purpose Information Retrieval (IR) techniques to rank patents. We argue that the quality of search can be significantly improved by exploiting the domain semantics: e.g., patents are organized into classes and subclasses, and have links to external publication and to other patents. Also patents' text is organized into various sections and uses specific legal wording.We present the PatentsSearcher system, available at PatentsSearcher.com, whose key contribution is to leverage the domain semantics to improve the quality of discovery and ranking. PatentsSearcher also offers other novel functionalities to help users locate and navigate relevant and important patents or applications.