« The Freedom to Be Who I Am | Main | The Emergence of Interactive Supercomputing Applications »

February 16, 2006

Comments

Steve Flinn

Irving,
Having just discovered your blogs, I really appreciate having access to your thoughtful perspectives.

With regard to promoting a more effective IP environment, it seems to me what is often lost in the ongoing IP debates is that the fundamental objective (which your re-emphasize in your blog) in encouraging invention disclosure within patents is that it is meant to educate the public and advance innovation (in exchange for the 20 years of exclusive property rights).

There seems to be two inter-related issues regarding the ongoing patent debates: the first issue is finding the optimal overall balance between the benefits of disclosure in exchange for inventor property rights, and the second issue is effectively promoting maximum learning from, and application of, patent applications and issued patents.

The first issue is basically a policy issue, and the key policy variable is the scope of what can be patented. In the USA that scope continues to increase (e.g., Lundgren). It can be debated whether that is positive or negative, but the “right” answer depends significantly on the second issue related to the ability to effectively learn from the disclosures of patents . . .

The second issue is a technical and process-related issue – it is fundamentally about applying technologies and processes to promote effective learning in the face of increased complexity (greater scope of patentability, greater volumes of patents, greater volumes of prior art, etc.). Although determining the right balance in patentability scope on an absolute basis is a thorny problem, on a relative basis it is certainly the case that the better the ability to assimilate patent knowledge and learn from it, the happier we should be about broadening patentability scope.

And here it seems to me that we need to think about effective learning both within the patent authorities (e.g., USPTO) to promote the highest quality patents, and for the public as whole. For example, the patent quality index approach sounds like a good step forward in that it can promote a simple metric that can gauge patent quality. This can be useful for the public as a whole, and also for patent authority examiners.

However, in the face of massive complexity and steep learning curves, it seems to me that we need to think about some additional methods of promoting effective learning in the face of increasing complexity, both within the patent authorities and for the public as a whole.

Specifically, a whole host of technical solutions for managing large volumes of complex information of varying quality has evolved over the past few years that increasingly leverage social-based information (leveraging the “Wisdom of the Crowd”). An example is Wikis technology, as applied in Wikipedia – a wonderful community resource that we all cite in our on-line writings. Certainly contemporary search engines also come to mind as instruments to manage increasing information volumes and complexity.

Extending beyond these specific technologies, it seems to me that organizing patents as a socially adaptive knowledge network with some of the following features could promote effective learning both within patent authorities (promoting higher quality issued patents) and for the public in general:

 Patents more formally represented as “objects”, with the patent objects tightly bound with meta-information, history etc.

 Socially-driven (explicit or inferred) interconnections between patents objects (“knowledge network”)

 Ability to create private, but optionally sharable, knowledge networks of patent objects with personal interconnections among the objects

 Ability to easily integrate other information (prior art, lab notes, etc.) into a patent object knowledge network for personal or shared use

 Ability for individuals or sub-communities (e.g., experts) to apply ratings, provide feedback, etc., to patent objects that can be used by others

 In addition to search capabilities, automatically generate personalized recommendations based on personal and social use patterns (and such a recommendations function could incorporate a patent quality index in its algorithm).

Such a system is feasible to implement right now. At the risk of a commercial plug, the best working example I know of is our business thought leadership site (www.manyworlds.com), which is an adaptive learning system for a complex knowledge base spanning a number of inter-related business-related domains. As an aside, reflecting your culture of learning, I must mention that at last count, IBM has more executives subscribing to wwww.manyworlds.com than any other company in the world!

Marc Ehrlich

Hi Steve -

You make some very compelling suggestions in your reply.

I agree it is vital to enhance the usability of the patent record to promote the intended public teaching both within the office (better exam) and in the public at large.

As you may well know, some aspects of the knowledge network you call for do already exist. There are a wealth of commercial tools and scholarly articles that cover networks of patents linked by the patent citations within the patent documents. This so-called patent citation analysis field is explored in depth in the book entitled Patents, Citations and Innovations: A Window on the Knowledge Economy, by Adam Jaffe.

Of course these citation-based patent networks are of limited utility as a learning mechanism since they are created by applicants and examiners in their necessarily limited recitation of relevant prior art during patent exam. The limitations in this process are well known as they omit (for example)citations to art seen as redundant from a patentability perspective, but potentially valuable from a learning perspective.

The only issue I would seek to clarify is your initial discussion about patentable subject matter.

I read your posting to say that we may be more comfortable with an expansion in patentable subject matter if patents were better at teaching the public. While I do agree that this ability to learn from patents bears on the utility of the patent system, I would point out that the harms of expanding patentable subject matter beyond their constitutional mandate (the 'useful arts') presents potential harms that cannot be remedied even by the most well crafted and fully knowledge enhanced patent documents.

Steve Flinn

Hi Marc,

Thanks for your reply and suggestions. Yes – agree the current citation management approaches contain the kernel of some of the capabilities needed for effective patent and relevant prior art – but I think we both agree there is much more that could and should be done to get to the next level of truly useful learning systems that can serve a variety of learning needs.

On the question of patentable subject matter – no, I’m not arguing for expanding patentable subject matter beyond the constitutional mandate. But, as we both know, even within the confines of that mandate, the interpretations of what is patentable subject matter has certainly expanded in the past couple of decades. And whether that is good or bad is clearly contentious. I personally believe it is a good thing, but if and only if patent examinations are of high quality and obviousness criteria are applied appropriately.

But although I believe this broader interpretation of patentable subject matter is a good thing, I can’t prove it a priori, and I have seen no fact-based analysis that can prove or disprove this hypothesis (if someone has I would be interested in taking a look at it).

As a recovering economist, the debate on the appropriate scope of patentability reminds me a bit of the debate in this country regarding socialism versus capitalism – until very recently, some the brightest minds in this country thought a predominantly socialistic economy would be more effective in the long run than a predominantly capitalistic economy. And frankly there was no theoretical a priori argument that was convincing one way or the other – the convincing came by looking at real-world performance over a number of decades. I think the same will be true of patent scope – we’ll have to evaluate the real-world consequences over a significant period of time to settle the debate.

One can visualize a value curve that stretches along a horizontal axis that measures the patentability scope chosen by a society. At one extreme is a policy of no patents and/or enforcement in any domain – there are many in the world that argue that that is the point at which the value curve is at its peak. At the other extreme is the broadest possible interpretation of patentable subject matter within the confines of the constitutional mandate. I believe the value curve peaks closer to this point – given the caveats about patent quality. Others will believe the value peak is somewhere in between.

But again – all the advocates of the different shapes of this value curve can do at this point is to generate plausibility arguments. On the other hand, what we all can agree on, is that higher patent quality and patent-based learning systems will serve to shift up our theoretical curve regardless of its actual shape.

Jonas Maebe

I do not agree that patent quality is somehow able to magically solve all problems faced by the patent system due to expanding subject matter. However, supposing it did, wouldn't the proper course of action be to first solve the quality problems and only then expand subject matter, instead of the other way round?

After all, the patent quality problems indeed aren't confined to fields like software and business methods, see e.g. this speech by David Martin from M-Cam: http://wiki.ffii.org/Martin041109En

His story (real world data!) about how low quality patents are hurting US companies operating in China and the amusing anecdote about the 683 patents that are currently in force covering the ethernet plug are quite interesting.

shrikant sharma

dr irving sir ihave gone thru yr nice expertise on patents. usa in perticuler and mankind ingeneral have been benifitted with the invention of the wordPATENT and the economics of patents are becoming complex more and more with the advancement of time.for aresearcher like me who has developeda new energy tapping system non fuel and sustainableSO MUCH THAT THATOIL DEPENDENCY OF ANY COUNTRY LIKE USA WILL BECOME ZERO IN THREE YEARS. how to proceed has become a problem. if i would have born in the era of Edison or so iwould have got the patents and not only usa but me also would have been benefitted. pl let me know what to be done maintaining the secrecy andpossibility of thefts minimum.

philippines property

If you talking about intellectual property (IP) Wikipedia’s right definition being the product of the mind or the intellect that has commercial value and may be legally protected in the same way as other forms of property.You have a very informative blog thank;s for sharing.

by:stephen

The comments to this entry are closed.

My Photo

November 2019

Sun Mon Tue Wed Thu Fri Sat
          1 2
3 4 5 6 7 8 9
10 11 12 13 14 15 16
17 18 19 20 21 22 23
24 25 26 27 28 29 30
Blog powered by Typepad