Patinformatics: Tasks to tools

This article starts with an overview of the field of patinformatics––the science of analyzing patent information to discover relationships and trends. This is followed by a survey of many common analysis tasks in this field, and many of the software tools available to tackle these tasks. The survey is set out under the tasks of list cleanup and grouping of concepts; list generation; cooccurrency matrices and circle graphs; clustering of structured data; clustering of unstructured data; mapping document clusters; adding temporal component to cluster map; citation analysis; subject/action/object functions. The author concludes that patinformatics has developed very rapidly over the last few years, and provides continuing challenges and opportunities in making optimal use of the resources available to achieve reliable and meaningful results. Useful tables summarizing aspects of this survey are included. 2003 Elsevier Ltd. All rights reserved.