Overcoming the problems with patents - how far have we come?

Moira Duncan looks at patenting, the problems associated with extracting critical data, and the tools available for data mining - Jan 2002

How the patent system started
Why are patents so important?
The sting in the tail
How information retrieval has developed
What about free patent searching on the internet?
A few problems solved

How the patent system started
When Henry VI founded Eton college in the fifteenth century he also introduced something that doesn't find its way into the history books quite so often: the patent. This particular patent was granted to John of Utynam in 1449 who was making stained glass for the windows using a method he'd brought with him from the continent. Henry's august family continued with this new -fangled idea, giving rise to the world's longest-running patent system. By the reign of Elizabeth I some fifty patents were granted each year.

The concept of ownership of ideas (and the subsequent privileges) that Henry inferred is not much changed today. Hundreds of thousands of patents are granted worldwide each year by some forty licensing agencies. Each country grants its own patents and has its own filing system which may be subtly - or not so subtly - different to that of its neighbours. Patents can also be issued internationally under a series of treaties and conventions.

Different filing systems mean a variety of classification systems and patents are filed, naturally, in the language of the country where the application is made.

Why are patents so important?
Patents grant their owners the right of intellectual property and the right of manufacture in a given country for a given period of time. They are of immense commercial value, but patent owners need not take the patent office up on its offer - they can sell the rights to other companies, known as licensing. In addition, patents contain detailed technical descriptions of their subjects and also show their relationship to similar inventions. In the strictest sense they are unique sources of information.

Thus we can begin to see the vital role patents play in research. Although superficially a patent search might be carried out to ascertain originality (a 'prior art' search), the role of patents in competitor intelligence cannot be underestimated. And now that we know that the patent trail doesn't go cold for several centuries, we can see how important it is historically in documenting the development of science and technology.

All patents contain links to other relevant patents and solutions to technical problems can often be found this way. And there's one final treat: research may uncover inventions that are now out of protection.

The sting in the tail
Of course, there's a catch to all this. Although absolutely in the public domain, patents can be alarmingly elusive. There's a good reason for this: the owner of the patent doesn't necessarily want you to find it. Abstracts, one of the main inroads, are often written in such a way as to obscure the real nature of the patent, and so the intentions of the patent owner. Front pages don't use scientific and technical terms, and foreign language patents produce their own difficulties - linguistic skills may be needed.

Also, technologies change at different speeds and the researcher really needs to know how far back to search to make real sense of the information found.

How information retrieval has developed
Many of the problems associated with patents searching can be overcome by using sophisticated document management software over the Internet. It was not always so. When Monty Hyams founded Derwent over fifty years ago information retrieval was in its infancy, and although by the mid-1960s customers could search the accumulated content of the Pharmaceutical Patents Journal using punched cards (otherwise known as 'peek-a-boo' cards), it was not until the mid-1970s that remote databases became accessible by computer (the size of refrigerators - PCs had yet to be invented). The Derwent Chemical Patents Index was one of the first online services available on what came to be known as 'hosts', in this case SDC-Orbit, and later Dialog. Online searching revolutionised the lives of researchers, just as the Internet and full-text searching was to do some twenty years later.

David Younghusband is Information Systems Specialist at Unilever, a multinational company specialising in food and homecare. 'Patents play an important part in the company's intellectual effort and are also useful to follow competitors,' says David. 'We try to predict future trends and opportunities.' He regularly uses Derwent World Patents Index and MicroPatent, and Inpadoc.

What about free patent searching on the internet?
David Younghusband thinks the Internet has been the most significant development for patents searchers in recent years, but urges caution. 'There is now a misconception that all information is freely available and is easily searchable at no cost. The reality is very different: you get what you pay for.

'The time spent floundering trying to find 'free' information is an invisible cost, and is usually money down the drain. Potential inventors have no incentive to find prior art, and can spend time and money preparing a patent for submission that will not pass the most basic patent examination.'

Edlyn Simmons is a section manager in the Business Information Services department of Procter and Gamble. The company has a policy of obtaining global patent coverage of all significant products and improvements on products and processes. She also uses all the major databases and cites the Internet as a revolutionary development. She doesn't necessarily agree with David Younghusband about free or low-cost searching. 'It allows simple searches to be done by everyone rather than being limited to the most prosperous corporations,' she says. However, she does identify another problem. 'Unfortunately, the Y2K problem is significant as it has generated a host of new document numbering formats, practically erasing the standardisation of patent data that was once provided by hosts.'

She finds terminology a problem in patents searching. 'The technical language of patents includes new terminology, non-standard terminology, and the creative use of generic terms by patent agents and attorneys who are attempting to broaden patent coverage beyond the scope of their examples.

'Patents are written in all the world's languages, and even equivalent patents published in English can contain badly translated text. The breadth of the patent literature introduces more than the usual amount of false retrieval based on the use of a single term in many technologies. And of course there are Markush structures in chemical patents.'

A few problems solved
The only way to counter these problems is by proper, new-fashioned indexing. 'Controlled index terms, classification codes and databases of limited technological scope help to avoid false retrieval that would be inevitable in a free-text search,' she comments. 'The only way to search for generic chemical structures is to use fragmentation and topological structure indexing.'

So what would Edlyn and David wish for from the patent fairy?
'I'd like standardisation of patent data formats across databases, across databanks, and across time,' says Edlyn. She'd also like clear-cut scope notes provided with every search of every database, 'so that people know when a search covers only a limited subset of the documents they expect.'

David isn't asking for much. 'I'd like a personal patent that will make me enough money to be financially comfortable and independent for the rest of my life!'

And why not.

Moira Duncan - is an experienced writer and editor for titles including Managing Information, Aslib, AccountingWeb, Library Association Record and FT.com. She began to work freelance in 1996, and has since worked for a variety of print and electronic publications, as well as marketing and promotions companies.