A team of researchers at the University of Michigan's School of Information are working with IBM scientists to develop new tools for searching and analyzing patent data. The NSF-funded project will use natural language search technology developed at the IBM Almaden Research Center, where the Free Patent Server was launched in 1997. IBM renamed the service the Intellectual Property Network (IPN) and then Delphion IPN, before selling it to Thomson.
While the press release describes the project as benefitting "individuals and businesses of all sizes," the project description emphasizes the development of new tools to help businesses identify "patent thickets... dense webs of overlapping patent rights that an organization must hack its way through in order to commercialize new technology." Certainly, this will be of most interest to companies in patent-intensive industries such as biotech, pharma/chemical and IT. It's unclear what impact it will have on independent inventors and small and medium-sized entreprises.
I'm the librarian for research services in the Engineering and Science Library at Queen's University in Kingston, Ontario. I've been working with patent information since 1991, including seven years at the U.S. Patent and Trademark Office. I believe that the dissemination of patent information is a public good and should be promoted, especially in the education of science and engineering students.