Posted by: Jason White
Public Sector Information Holder: UK Intellectual Property Office
Information Asset: Patents - Bibliographic, full text, images
The UK Intellectual Property Office makes available individual patents through their web site and search engines. They have recently implemented the capability to download some data in batches in a very badly structured XML format.
The data should be freely available for all to download and make use of as they see fit. This requires a copy of the complete UK patent database, in a well-accepted XML format, such as ST36.
My ideal solution
Separate datasets for bibliographic-only and full-text, freely downloadable in XML ST.36 format. This format is well-understood by the patent community, is maintained by WIPO, and used by EPO. Another freely available dataset with all patent PDFs.
The downloadable database should include the fullback file to at least 1990, preferably earlier. Frontfile data should be published weekly for all three datasets.
Ancillary datasets, such as legal events, reclassifications, etc. should also be made available and downloadable in bulk.
What I would do
I would include UK patents (which are separate from EP patents) in an existing commercial application. This system already includes data from US, EP, and WO. The US now makes all this data freely available, and downloadable via Google.