hide
Free keywords:
Patent Analysis, Relational Databases, Search Engines, Information Retrieval
Abstract:
Patent databases contain large amounts of (almost) unstructured references to non-patent literature (NPL). To identify these references is a general research request, as they are an important indicator for determining and quantifying various relationships between science and industry. In the present pilot study, we introduce a Patent reference matching method (PARMA) that is able to process a wide range of patent records by using a combination of full text search technology with filtering and matching routines in an RDBMS. Results show that the approach establishes a solid foundation for future analytic studies on the topic.
