Patent Eligibility: Distinguishing Collecting Information from Using Information

by Dennis Crouch

I’m always excited to read a decision that splits hairs–finding some claims in a patent valid and others invalid.  My hope is that the explanation will really get to the crux of the issues and help me to better understand how the law works. I’m often disappointed, but the Federal Circuit’s newest decision in Weisner v. Google LLC, — F.4th — (Fed. Cir. 2022) does offer some good clues.  The basic outcome here: Weisner’s claims directed toward collecting information are abstract ideas; those directed toward using the information are patent eligible.  This outcome is consistent with what we are seeing in biotech as well: diagnostics get a thumbs down; therapeutics get a thumbs up.

Weisner sued Google for patent infringement back in 2020, asserting infringement of a family of four patents. U.S. Patent Nos. 10,380,202, 10,642,910, 10,394,905 and 10,642,911.  But, Google won the case fairly quickly on a R.12(b)(6) motion to dismiss with a holding from the district court that the asserted claims are ineligible under 35 U.S.C. § 101 (abstract ideas).  On appeal, a divided Federal Circuit has reversed-in-part, holding that some of the claims are patent eligible because they implement “a specific solution to a problem rooted in computer technology.”  Alice Step 2.

Judge Stoll wrote the majority opinion joined by Judge Reyna. Judge Hughes dissented, arguing that all the claims are invalid.

Weisner’s four patents are all part of the same family and all relate to ways of recording a person’s physical location history in ways that correspond to certain transactions. In the image (Fig.3 above), you can see the dude interacting with a Macy’s story device. Dude’s phone is configured to record the transaction along with the location history and push it to the cloud.  Dude is doing something similar at Benson’s (Fig.4 below), but in that one he is entering the transaction details manually since Benson’s is a Brooklyn institution and isn’t high-tech like Macy’s.

The Federal Circuit divided the claims into two categories:

  • Claim Set A: Accumulating location histories for later use.
  • Claim Set B: Accumulating the location history and also using hose histories to improve search results (e.g., give different search results if someone has been to a Brooklyn Benson’s before).

In the appeal, the Federal Circuit agreed that all of the claims fail Alice step 1. In particular, the court found that they were all directed to the unpatentable idea of “collecting information on a user’s movements and location history and electronically recording that data.”  Essentially, this is an electronic travel log that is effectively the same as what humans have been doing for all of history.  As the district court wrote: “Humans have consistently kept records of a person’s location and travel in the form of travel logs, diaries, journals, and calendars, which compile information such as time and location.” Weisner v. Google LLC, 551 F. Supp. 3d 334 (S.D.N.Y. 2021).  Weisner argued that the claims were limited to keeping travel history of members only — a feature that improves data integrity.  On appeal though, the Federal Circuit rejected this claim as “attorney argument, unlinked to the complaint or the patent claims or specifications.”  Even though eligibility is a question of law, it must still be based upon something more than simply attorney argument.

At step 2, the court also agreed that Claim Set A was not salvageable because the recited innovations are all merely generically claimed components.

But, Claim Set B is different. First, because Set B adds the search results uses, the Federal Circuit found it to be a “much closer question” as to whether they are directed to an abstract idea.  In the end though the court concluded that the claims are directed to an abstract idea at Alice Step 1.  I note here that the appellate court did not use any real analysis to decide this issue other than seemingly attorney judgely  argument.  Perhaps the court should have checked the prior portion of this very opinion where it discounted mere attorney argument on the same topic.

At Alice Step 2, the court concluded that the claims recite a specific implementation that – at least at this stage – should have survived a motion to dismiss because they “plausibly capture an inventive concept.  Here, the court notes that the inventive concept comes from tying searches to a third party “reference individual” whose location history is similar to the party doing an internet search. “The system then prioritizes search results that the reference individual has visited.”  In its decision, the Federal Circuit explained that this linkage is more than simply “improving web search using location history.” Rather, it is a specific implementation of that concept.

Judge Hughes wrote in dissent that the claimed search improvements are just like any old search with the addition of new searchable data — location history.  The Stoll majority responds in a footnote that Judge Hughes opinion “misses the point” by discounting the importance of prioritizing travel histories in the search.  As I mentioned in the intro, this outcome fits a standard approach that we are seeing in eligibility cases — it is much easier to protect methods of using information than it is to protect methods of collecting information.  In the use-case, the information itself combined with some use will regularly be seen as an inventive concept.  On the other hand, the courts have been less willing to say that collecting information is patent eligible.

= = = =

The patents here are co-owned by the inventors, Sholem Weisner and Schmuel Nemanov.   Weisner wanted to sue Google, but Nemanov refused.   Weisner went ahead and sued Google and also joined Namanov as an involuntary party under FRCP 19(a).  See, AsymmetRx, Inv. v. Biocare Med., 582 F.3d 1314, 1322 (Fed. Cir. 2009)(“A patentee that does not voluntarily join an action prosecuted by its exclusive licensee can be joined as a defendant….”).   The agreement between the two parties is a bit interesting. Weisner owns 78% of the rights, and the parties agreed that he would control exploitation of the patent “and will not be undercut by a minority owner.” WeisnerNemanovAgreement. The two battled these issues in a parallel lawsuit in NY State Court. WEISNER, SHOLEM vs. NEMANOV, SHMUEL, Docket No. 502269/2020 (N.Y. Sup Ct. Jan 29, 2020).

12 thoughts on “Patent Eligibility: Distinguishing Collecting Information from Using Information

  1. 5

    “The basic outcome here: Weisner’s claims directed toward collecting information are abstract ideas; those directed toward using the information are patent eligible.”

    I’m sorry, but this is absolute nonsense. Every application that collects data uses the data somehow.

    It’s the right outcome, but the legal theory iss trash

    1. 5.2

      It’s the right outcome, but the legal theory [is] trash[.]

      Agreed. I am happy to see the patentee win a subject matter eligibility challenge (most such decisions are nonsense). Still and all, the idea that the precedent leads to this outcome in this case, but the exact opposite outcome in three dozen other, materially identical cases is a bit hard to swallow.

      1. 5.2.1

        Stoll can be capricious, but today she’s capricious (semi) on the correct side.

        If only someone in a black robe would realize that s 103 addresses the preemption issue

  2. 4

    A farce, exactly. The utility to the end-user (the person doing the search) is entirely, 1000% abstract. Maybe the resulting suggestions are good, maybe they’re not. Patentable utility in purely human affairs is an affront to liberty.

    It’s definitely “technology” to define consumer personae by shared attributes of individuals. Wait, what?

    Meanwhile, even at the sole appellate court, no outside observer can predict which inventions will be eligible and which will not be with better than coin-flip accuracy.

    Meanwhile Shmuel and Sholem may as well be Bialystock and Bloom in revealing how things really work….

    1. 4.1

      Patentable utility in purely human affairs is an affront to liberty.

      Damm those traffic lights, meddling with purely human affairs.

  3. 3

    “ the Federal Circuit explained that this linkage is more than simply “improving web search using location history.” Rather, it is a specific implementation of that concept.”

    And do the farce continues. An ineligible abstraction (logic) mysteriously becomes non-abstract when more words (or more specific words) are used to describe it. It makes no sense and it never will but if the end result (some eligible data processing claims in the patent system) is all that matters (hint: it’s all matters to some self-interested folks) then we are stuck with the game playing.

    It’d be nice if more people just admitted this but oh well.

    1. 3.1

      “And do the farce continues. An ineligible abstraction (logic) mysteriously becomes non-abstract when more words (or more specific words) are used to describe it.”

      I don’t think I’ve ever agreed with you on anything, but here you’re correct

  4. 1

    Here, the court notes that the inventive concept comes from tying searches to a third party “reference individual” whose location history is similar to the party doing an internet search. “The system then prioritizes search results that the reference individual has visited.”

    On appeal, a divided Federal Circuit has reversed-in-part, holding that some of the claims are patent eligible because they implement “a specific solution to a problem rooted in computer technology.”

    Reading these two statements I thought the court had erred, as obviously recommending something based upon a similarly situated individual is both conventional and not a problem rooted in computer technology. But upon reading the decision the court emphasizes a lot of particular computer limitations in a specific method. I would say that the quoted description improperly describes the claims in too generic a fashion.

    That being said, the ‘911 claims about the geographic location of the user are neither inventive nor a computer program. Virtually all conventional recommendation is geographically limited. If you tell a person that you like Local Steakhouse X and ask for recommendations, you’re still going to get Local Establishment Y. Nobody responds to a request for food by saying “You should really try this little place in Berlin.”

    I understand why the district court would find it ineligible – because the best that can be said of any claim is that it is using conventionally discoverable information to solve a conventional recommendation question based on conventional similarity concerns. All recommendation is based upon analogizing some feature of the requestor to historically similar situations. There’s a big difference between no use, a use of having a database of where other people visited (use type A), and a use of having a database of local items to recommend (use type B). B is clearly more conventional than A.

    1. 1.1

      All of this is as much as to say that there should probably be a sound and solid §103 challenge if Google took the care to file the IPR or raise such a defense in its answer. I am pleased to see a court decline to treat a §103 issue as if it were a §101 issue, although doing so runs rather counter to the last decade’s worth of case law.

      1. 1.1.1

        Corrected:

        although doing so runs rather counter to some of the last decade’s worth of case law.

        Pretending that there is no Gordian Knot is just not helpful.

Comments are closed.