Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[TaiwanIP]

If we limit the discussion to just the method claims (because let's face it, the invention will be in the method for e-commerce applications), what are some things to watch out for when drafting these claims?

For example, should great care be taken in making sure that we recite "'computer-implemented' method," such as in this Amazon patent:
http://www.google.com/patents/about?id=kzeuAAAAEBAJ&dq=7,433,835

Should we be sure to state that everything is being done and received by a server, such as in this application:
http://appft1.uspto.gov/netacgi/nph-Parser?Sect1=PTO2&Sect2=HITOFF&u=%2Fnetahtml%2FPTO%2Fsearch-adv.html&r=1&p=1&f=G&l=50&d=PG01&S1=e-commerce.TTL.&OS=ttl/e-commerce&RS=TTL/e-commerce

Does Bilski change things for how e-commerce patent claims should be drafted?

[MYK]

See the RIMM v NTP case for what happens when you only have method claims.  You *need* that valid system claim, otherwise if any single step is performed outside the USA, there is no infringement.

You might take a look at this:
http://www.kramerlevin.com/files/Publication/7b9a8811-e314-4dfb-b08e-0b19486a31f5/Presentation/PublicationAttachment/0256540a-4b87-4403-9fd4-0c3c4447fd08/NTPvRIMIPStrategist.pdf

[khazzah]

If we limit the discussion to just the method claims (because let's face it, the invention will be in the method for e-commerce applications)

Ah, but you should not limit the discussion. There are important differences between method claims and system/apparatus claims. Some of which were already discussed by MYK, and related ones which I address below.

Should we be sure to state that everything is being done and received by a server, such as in this application:

What's your thinking here? ie, explain what you think the consequences are for including the term "server", and for not including the term.

Does Bilski change things for how e-commerce patent claims should be drafted?

Ah, that's the million dollar question. Seriously. No one knows yet. The Fed Cir has not yet decided a post-Bilski case involving what I would call an "e-commerce" patent. RCT v. Microsoft looked at a method claim for processing images, and found it not abstract. That's the only Fed Cir guidance that we have so far.

See the RIMM v NTP case for what happens when you only have method claims.  You *need* that valid system claim, otherwise if any single step is performed outside the USA, there is no infringement.

Centillion v. Qwest points out another advantage of system claims over method. Namely, if the system claim includes client elements and server elements, the entity operating the client may be found to *use* the system as a whole. You can then go after the entity that operates the server for indirect infringement. Without the Centillion holding, there can be no single direct infringer of such a claim, and thus no indirect infringement. That leaves only vicarious liability, which simply doesn't exist in many fact situations (see, e.g., Golden Hour Data System v. emsCharts and Akamai v. Limelight).

The better approach is to avoid claims that look like those in Centillion, Akamai and GoldenHour. That is, don't write a claim which includes the actions of, or components operated by, two different entities. But sometimes that's easier said than done, so it's nice to have Centillion to fall back on.

[JimIvey]

The Fed Cir has not yet decided a post-Bilski case involving what I would call an "e-commerce" patent.

I'll just toss out my usual comment:

It's better to think of them as network-based service/server patents rather than "e-commerce" patents.  The only difference is that some data somewhere represents money in the latter category.  I have yet to see any hint that the Fed Cir will use that feature as a distinction in patent law.

Having said that, I agree with the comments so far.  I always try to include claims that can be infringed by the server or the client alone, and I always include Beauregard and computer system claims.

Regards.

[TaiwanIP]

There have been some excellent comments so far. I just started on the specification yesterday. At this beginning stage, I'm thinking in terms of the method because I'm very involved with the idea and the inventors. When and if the inventors decide to go through with their idea, various programmers will be involved in setting up the system and they will have to fully understood the processes involved. They probably won't be interested in looking at my apparatus claims or my drawing with the server shown with a processor, various databases, memories, etc. What I have to say about the method in the application, however, might be something they're interested in looking at, though. I was hoping to limit the discussion to the method simply because I'm only thinking in terms of the method at this preliminary stage. Of course I plan to eventually draft system claims and most likely claims directed to a computer-readable medium.

At this point, it would be of immense help if someone could point me to a published patent application which in their view would help in using as a guide to follow as I prepare the application.

[MLM]

I have been wrestling with this too. As a software engineer I thought that I would have a good handle on writing these kinds of applications, but for me they are by far the most difficult. The inventors in this space have too many, constantly evolving abstract concepts. Perhaps I am overthinking things, but I do trust my instincts, and my instincts tell me this is hard stuff. Oh, and always 103 rejections to deal with. Yay.

FWIW right now I am looking at another Amazon patent for inspiration, 7,720,723. It has a mix of flow diagrams and screen shots.

[khazzah]

FWIW right now I am looking at another Amazon patent for inspiration, 7,720,723. It has a mix of flow diagrams and screen shots.

Conventional wisdom says to use Flow charts to describe software. Kinda funny since most developers I talk to don't use flow charts, at least not as their main mode of expression. But maybe the thinking is that a) that's what everyone else does and b) examiners are familiar with flow charts.

Flow charts make the most sense to me when describing control flow (if-then, iteration, etc.). A flow chart that does nothing but illustrate a straight sequence of actions doesn't do much for me. Also, a flow chart doesn't convey what component is performing the step -- and that may be important.

Personally, I tend to use a mix of flow charts, block diagrams (one way of showing component interactions) and messaging diagrams (another way of showing component interactions).

[JimIvey]

I do pretty much the same as Karen describes.  Software development currently uses a different type of charting, and I don't remember the name of it.  For a while, I thought I should learn to use it since it's what developers would more likely understand.

However, I'll defend flowcharts.  Software people tend to anthropomorphize their software, describing it as trying to do this, wanting that, aspiring to be this, etc.  Flowcharts help us reduce the software to clear, concrete steps that are more amenable to patent protection.

Given the context of this particular topic (server/client applications), I'd say there's an important type of drawing missing in the cited Amazon patent.  I use transaction flow diagrams: side-by-side flowcharts showing the participation of multiple players (computers or perhaps processes in the same computer) in a transaction.  As a rule, I never draw a flowchart that includes steps performed by different players.

That helps one think in terms of one-sided activity that could be claimed, i.e., helps in drafting claims that can be infringed by the server or client alone.  When you have a single flowchart in which some of the steps are performed by the client and others by the server, it's difficult to conceptually separate them and those charts tend to lead to claims that won't ever be directly infringed by any client or server.

Regards.

[TaiwanIP]

I have been wrestling with this too. As a software engineer I thought that I would have a good handle on writing these kinds of applications, but for me they are by far the most difficult. The inventors in this space have too many, constantly evolving abstract concepts. Perhaps I am overthinking things, but I do trust my instincts, and my instincts tell me this is hard stuff. Oh, and always 103 rejections to deal with. Yay.

FWIW right now I am looking at another Amazon patent for inspiration, 7,720,723. It has a mix of flow diagrams and screen shots.

I have to come back to this specification at a later time now because another issue came up. I'll definitely look over this patent and try to provide more comments at a later time. Thank you for letting me know of this one. I briefly scanned over the claims and I see many similarities to the Amazon patent I provided a link to above. It’s obvious we're not the only ones that use past publications and patents for inspiration!

[TaiwanIP]

Finally back to this specification. I found "BPAI Finds Method for Interpreting an MRI Image Statutory Even Though Not Tied to a Machine Implementation" to be an interesting read related to this issue:
http://www.patents4software.com/?p=415

I realize that BPAI decisions carry little weight in the grand scheme of things but some of the points made had me nodding my head in agreement. Also, check out the first reply (the only reply as of this writing) which reads as follows:

Interesting observation. It makes me think back to Chief Justice Roberts at the oral argument of Bilski.
CHIEF JUSTICE ROBERTS: Mr. Stewart, I thought I understood your argument up until the very last footnote in your brief. And you say the method isn’t patentable because it doesn’t involve a machine. But then you say but it might be if you use a computer to identify the parties that you are setting a price between and if you used a microprocessor to calculate the price. That’s like saying if you use a typewriter to type out the process, then it is patentable. That takes away everything that you spent 53 pages establishing.

[khazzah]

I use transaction flow diagrams: side-by-side flowcharts showing the participation of multiple players (computers or perhaps processes in the same computer) in a transaction.  As a rule, I never draw a flowchart that includes steps performed by different players.

That helps one think in terms of one-sided activity that could be claimed, i.e., helps in drafting claims that can be infringed by the server or client alone. 

[Somehow I missed Jim's comment from a few weeks ago.]

Jim, your "transaction flow diagrams" are my "messaging diagrams". So I agree that applications involving client-server interactions are best served by this type of diagram rather than flow chart.

[TaiwanIP]

Should we be sure to state that everything is being done and received by a server, such as in this application:

What's your thinking here? ie, explain what you think the consequences are for including the term "server", and for not including the term.

It was a style I saw in an application I was using as a reference. I'm basically done with the claims (I think) and "server" was never mentioned!

The better approach is to avoid claims that look like those in Centillion, Akamai and GoldenHour. That is, don't write a claim which includes the actions of, or components operated by, two different entities. But sometimes that's easier said than done, so it's nice to have Centillion to fall back on.

This is good advice and something I was sure to do in my claims.

[JimIvey]

Finally back to this specification. I found "BPAI Finds Method for Interpreting an MRI Image Statutory Even Though Not Tied to a Machine Implementation" to be an interesting read related to this issue:
http://www.patents4software.com/?p=415

Well, from the linked article, here's the claim (emphasis supplied):

1. A method of classifying tissue in a magnetic resonance image, the method comprising:

(a) acquiring a magnetic resonance image of a region of interest;

(b) constructing a pixel intensity histogram of the magnetic resonance image; and

(c) applying a statistical regression analysis to the histogram to determine a pixel intensity threshold value for segmenting the histogram into at least two regions, wherein at least one of the regions is representative of a tissue of interest.

The author observes that the claim doesn't require a computer:  "What is interesting about this case and the RCT decision is that the method claims in question are not in any way tied to a computer implementation."  However, I don't see how you can "acquir[e] a magnetic resonance image" without an MRI machine -- I don't believe the law is that the particular machine must be a computer.  In addition, you normally have to physically transform matter by moving the matter into the MRI machine in order to acquire an MRI of the region of interest of the matter.

However, I'm sure there are examiners in certain technology centers that can take MRIs in their heads.

Regards.

[khazzah]

However, I'm sure there are examiners in certain technology centers that can take MRIs in their heads.

No doubt. We did just talk to an Examiner who could, without a computer, "mentally/visually/audibly monitor computer processing resources used by one or more computer processes executing in a computer". Discussed here http://www.intelproplaw.com/ip_forum/index.php/topic,16941.msg85365.html#msg85365

[Isaac]

I don't believe the law is that the particular machine must be a computer. 

Exactly so.  The commentator's remarks are a bit goofy.

The examiner actually alleged that the claim did not require a machine or transformation and did not require a computer.  The BPAI could have found that the claim did require a MRI.  Alternatively the Board could have found that the claim required transforming data that represented a physical object. 

But the board did a little better than that in my opinion.   They pointed out that the examiner did not establish that the claim was not abstract despite the allegation that the claim did not pass the MOT test.  The Board then found the claim not to be abstract without resorting to MOT.

Bilski was cited for the following proposition:

...for a claimed method to be disqualified under § 101 because it is too abstract, “this disqualifying characteristic should exhibit itself so manifestly as to override the broad statutory categories of eligible subject matter.”

Examiners should not pretend that the SCt never decided Bilski v. Kappos by simply relying on non compliance with the machine or transform test.  It would be nice is some decisions like this were made precedential.