Trials@uspto.gov
`571-272-7822
`
`
`Paper 34
`Date: September 27, 2023
`
`
`
`
`
`UNITED STATES PATENT AND TRADEMARK OFFICEp
`__________
`
`BEFORE THE PATENT TRIAL AND APPEAL BOARD
`
`GOOGLE LLC,
`Petitioner
`v.
`NOBOTS LLC,
`Patent Owner.
`__________
`
`IPR2022-00940
`Patent 9,595,008 B1
`__________
`
`Record of Oral Hearing
`Held: August 31, 2023
`__________
`
`Before JOSIAH C. COCKS, BARBARA A. PARVIS, and
`ROBERT J. WEINSCHENK, Administrative Patent Judges.
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`

`

`IPR 2022-00940
`Patent 9,595,008 B1
`
`
`
`APPEARANCES:
`
`ON BEHALF OF THE PETITIONER:
`
`
`ELIZABETH HUNT, ESQ.
`HUNTER KEETON, ESQ.
`Wolf, Greenfield & Sacks, P.C.
`600 Atlantic Avenue
`23rd Floor
`Boston, MA 02210
`(617) 646-8443
`ehunt-ptab@wolfgreenfield.com
`hunter.keeton@wolfgreenfield.com
`
`
`ON BEHALF OF THE PATENT OWNER:
`
`
`STEPHEN ZINDA, ESQ.
`JAMES HALL, ESQ.
`Cabello Hall Zinda, PLLC
`801 Travis Street
`Suite 1610
`Houston, Texas 77002
`(832) 631-9994
`stephen@chzfirm.com
`james@chzfirm.com
`
`
`The above-entitled matter came on for hearing Thursday,
`
`August 31, 2023, commencing at 1:00 p.m. EDT, via Video-conference.
`
`
`
`
`
`
`2
`
`

`

`IPR2022-00940
`Patent 9,595,008 B1
`
`
`P-R-O-C-E-E-D-I-N-G-S
`
`1:00 p.m.
`
`
`
`JUDGE COCKS: Good afternoon, everyone. Welcome to the Board.
`
`We're here today for two oral hearings in connection with two related
`
`proceedings, IPR 2022-00940 and 00941.
`
`As the parties requested, we will have separate hearings for these two
`
`proceedings. They will be heard consecutively. We will begin with the
`
`00940 case, which involves U.S. Patent Number 9,595,008. We will have a
`
`separate transcript generated for each proceeding.
`
`
`
`I am Judge Cocks. I am joined on the Panel by Judges Weinschenk
`
`and Parvis. Let's begin with the introduction of counsel. Counsel for
`
`Petitioner, please state their appearance today.
`
`
`
`MS. HUNT: Good afternoon, Your Honors. I am Elisabeth Hunt
`
`from the firm Wolf, Greenfield & Sacks representing the Petitioner Google.
`
`With me is my colleague, Hunter Keeton, also from Wolf, Greenfield &
`
`Sacks. And I will be presenting the argument today.
`
`
`
`JUDGE COCKS: All right. Thank you, Ms. Hunt. And would
`
`counsel for Patent Owner please state their appearance?
`
`
`
`MR. ZINDA: Yes, Your Honor. Stephen Zinda from the law firm of
`
`Cabello Hall Zinda for Patent Owner Nobots. I am also joined by my
`
`counsel, James Hall. And we also have on the line Andres Healy, counsel in
`
`the underlying litigation, and the inventor of the patent at issue, Tim Heikell,
`
`who is also a founder of Nobots. And I will be doing the arguments in the
`
`940 proceeding, Your Honor.
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`
`
`
`
`
`3
`
`

`

`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
`
`IPR2022-00940
`Patent 9,595,008 B1
`
`
`
`
`JUDGE COCKS: Thank you, Mr. Zinda. Okay. So as we set forth in
`
`the order setting oral argument, each side has 50 minutes, up to 50 minutes
`
`of argument time.
`
`Petitioner will begin and argue their case and may reserve rebuttal
`
`time. The Patent Owner will then argue their opposition to Petitioner's case
`
`and may reserve surrebuttal time. The Petitioner will then argue any
`
`rebuttal, and we will conclude with Patent Owner arguing any surrebuttal.
`
`
`
`A couple of housekeeping items. If at any time you experience
`
`technical difficulties, please notify us immediately for instance, by
`
`contacting the team member who provided you with today's connection
`
`information.
`
`
`
`So the Panel has the entire record before us, including the
`
`demonstratives. If you could, please refer to any slide clearly that you are
`
`referring to or paper number or exhibit number of the proceeding.
`
`
`
`Lastly, please activate yourself -- please activate mute on yourself
`
`when you are not speaking. With that being said, that's all I have for
`
`beginning. Ms. Hunt, you may begin whenever you are ready.
`
`
`
`MS. HUNT: Thank you, Your Honor, and I am hoping to reserve 10
`
`minutes of rebuttal today.
`
`
`
`
`
`JUDGE COCKS: Okay. Thank you.
`
`MS. HUNT: All right. So this case is about detecting bots in internet
`
`interactions on web pages and web pages hosted on a server. And anybody
`
`who views that web page is accessing the server from a client computer over
`
`the internet.
`
`
`
`Now if that user isn't a human but rather an automated program, a bot,
`
`that can be bad because bots are usually there to do something malicious on
`
`
`
`
`
`
`4
`
`

`

`IPR2022-00940
`Patent 9,595,008 B1
`
`
`the web page. And bots are really fast so they can do that malicious thing
`
`thousands of times in a matter of seconds.
`
`
`
`So there are various tools that are used to detect and block these bots.
`
`Most of us are familiar with Captcha where you have to answer some test on
`
`the web page to prove you're not a robot.
`
`
`
`The Nobots' patents relate to detecting bots by looking at other types
`
`of data rather than making the user take a test. So the user can just do what
`
`they normally do, and the system can tell whether they are human. Now this
`
`was all known in the art before Nobots' patents.
`
`
`
`Looking at Slide 3, an example is the Willner reference. It detects
`
`bots to prevent click fraud, which is when a bot clicks on an ad on a web
`
`page. And this is bad because it makes the advertiser pay the website owner
`
`as if the click was from a human customer so you want to prevent that.
`
`
`
`Now Willner says you can tell whether the user is a human or not by
`
`looking at how they move their mouse, which is also called their cursor. If
`
`they're human, the movement isn't going to be a perfect straight line. It's
`
`going to have some randomness to it. But bots might pretend to be human
`
`by reproducing a recording of human movement. So Willner also says that
`
`you can tell it's a bot if the movement is too much the same as a previous
`
`movement.
`
`
`
`Now Nobots doesn't dispute that Willner anticipates most of the '008
`
`patent's claims. Looking back at Slide 2, Ground 1 based on Willner is
`
`undisputed on Claims 1 through 3 and 6 through 17. So unless Your Honors
`
`have questions on those claims, I plan to discuss just the other five claims
`
`that are disputed today.
`
`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
`
`
`
`
`
`
`5
`
`

`

`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
`
`IPR2022-00940
`Patent 9,595,008 B1
`
`
`
`
`But I want to briefly look at Claim 1 first just for background, so back
`
`on Slide 3. Claim 1 is about one way of doing bot detection without making
`
`the user do anything special. So the user is using a web page, and you
`
`monitor their interaction with the web page like how their mouse moves.
`
`
`
`So Claim 1 says first, there is some issue data that goes to the client
`
`device from the server. That could be a web page for example. And then
`
`you monitor and analyze what the user does in response to that issued data.
`
`
`
`So things like mouse movements are called active data in Nobots'
`
`patents. And the way you collect is you have some monitoring code that
`
`comes with the issue data from the server, and it monitors the cursor's
`
`movements. So it makes sense for active data to be collected after
`
`presenting the issued data in Claim 1.
`
`
`
`Now another way of doing bot detection that was known in the art is
`
`discussed in the other independent claim, Claim 19, and that's where I would
`
`like to start in terms of disputed issues today.
`
`
`
`So Slide 35. In Claim 19, you look at interest data from before the
`
`issued data, like a web page, is delivered to the client computer. And this is
`
`taught by the Kitts' reference.
`
`So Kitts' server has a web page, which is issued data that the client
`
`computer wants to download. But before delivering that web page, Kitts
`
`looks at data attributes that just normally come with a client device's web
`
`page request to detect if the user is a BOT. And these are things like the
`
`client device's IP address. Its geographic location. Its user profile.
`
`And these are called passive data in the '008 patent because it's data
`
`that is normally there. You don't need to send special monitoring software
`
`to the client to get passive data. So it makes sense that you can get passive
`
`
`
`
`
`
`6
`
`

`

`IPR2022-00940
`Patent 9,595,008 B1
`
`
`data prior to delivery of issued data from the server as in Claim 19. And it
`
`also makes sense together with the --
`
`
`
`
`
`
`
`JUDGE WEINSCHENK: Ms. Hunt?
`
`MS. HUNT: Yes?
`
`JUDGE WEINSCHENK: This is Judge Weinschenk. I think Patent
`
`Owner is arguing here that the interest data has to be active data. Can you --
`
`I think you're arguing that it's passive data because it's received prior to
`
`delivery of the issued data. Is it possible to acquire active data prior to
`
`delivery of the issued data or is that not possible?
`
`
`
`MS. HUNT: So what I'm -- I was giving some context of how it
`
`makes sense to acquire passive data prior to delivery of issued data. That's
`
`not the reason for the argument here. But I will say that we haven't found
`
`description in the specification for how you would acquire active data prior
`
`to delivery of issued data from the server.
`
`
`
`JUDGE WEINSCHENK: So there's no example of that, but we're not
`
`sure. It could be possible.
`
`
`
`MS. HUNT: I don't believe there's an example so that's not discussed
`
`in the record of how to do that.
`
`
`
`Okay. So I was just mentioning Claim 20. We're going to talk in a
`
`minute about what's the correct dependency of Claim 20. But just looking at
`
`what it says, it says the interest data consists of available data. And the
`
`patent says available data is passive data. So there's no dispute that the
`
`passive data that Kitts gets and analyzes is available data and that it's interest
`
`data because the patent says interest data can be active or passive data.
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`
`
`
`
`
`7
`
`

`

`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
`
`IPR2022-00940
`Patent 9,595,008 B1
`
`
`
`
`But with Claim 20, the patent is specifically targeting interest data
`
`that's passive data. And there's no dispute that Kitts acquires this interest
`
`data prior to delivery of issued data.
`
`
`
`Now as Your Honor just said, Nobots' only argument is that where
`
`Claim 19 says acquiring interest data prior to delivery of issued data, Nobots
`
`alleges this word acquiring limits the interest data so it has to include active
`
`data. And that's not true for several reasons, which I'll walk through.
`
`
`
`Slide 36. The first reason is the word acquiring is just an individual
`
`word in Claim 19. It has its ordinary meaning. Nobots is trying to rely on
`
`the phrase term acquired data in the spec. But that phrase term isn't in Claim
`
`19. The phrase term in the claim is interest data. That's what limits the type
`
`of data, and it's defined as active or passive data.
`
`
`
`The word acquiring in the claim isn't part of a phrase about the type of
`
`data. It's a word --
`
`
`
`JUDGE COCKS: Ms. Hunt, a question. You just said it is defined as
`
`both those types of, I guess, available or acquired. Are you referring to that
`
`single paragraph in the specification, I think, in Column 3?
`
`
`
`MS. HUNT: The description of interest data, Column 3, Lines 40 to
`
`45, which is on our Slide 36.
`
`
`
`
`
`JUDGE COCKS: Okay.
`
`MS. HUNT: So that explicitly says that active data or passive data
`
`can meet the term interest data. Okay?
`
`
`
`So separately the word acquiring is just a word in the clause that says
`
`acquiring from the client computing device prior to delivery. So the
`
`acquiring clause limits when and where to get the interest data. It's not
`
`about the type of data.
`
`
`
`
`
`
`8
`
`

`

`IPR2022-00940
`Patent 9,595,008 B1
`
`
`
`
`And the patent makes clear that when it uses an individual word, it
`
`doesn't have to have the same definition as when that word is used inside a
`
`defined phrase term. And there are several places that demonstrate this.
`
`
`
`So one place is on Slide 36 at the bottom right. Here the spec is
`
`talking about active data. And it says a program makes the data available.
`
`Now available data is a phrase term that is defined as passive data. But
`
`clearly that doesn't mean the word available always has to be about passive
`
`data because in that bottom right passage on Slide 36, the word available
`
`gets used to talk about active data.
`
`And the same thing applies to the word acquiring. It's a word with an
`
`ordinary meaning that can be used about any kind of data. And we can see
`
`this on Slide 37 as well.
`
`
`
`So on this slide is the specifications only discussion of the process in
`
`Claim 19. It says in a first series of embodiments, there is a comparison
`
`between interest data acquired prior to delivery of issued data to the client
`
`computing device and model data. That's what is recited in Claim 19 using
`
`the plain word acquired or acquiring.
`
`
`
`And then the spec has a second series of embodiments where there is
`
`a comparison between monitored data acquired after delivery of issued data
`
`to the client computing device and model data. That's what is recited in
`
`Claim 1.
`
`
`
`And then the spec says in both series of embodiments, acquired and/or
`
`available data may be used for comparison with the model data. So even
`
`though the spec has used that word acquired, just like Claim 19 uses the
`
`word acquiring, as a plain word to talk about when to get the interest data or
`
`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
`
`
`
`
`
`
`9
`
`

`

`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
`
`IPR2022-00940
`Patent 9,595,008 B1
`
`
`the model data, the spec explicitly says that the type of data can be acquired
`
`data or it can be available data.
`
`
`
`And you know that it's talking about the interest data and the
`
`monitored data that were in the prior two sentences because those are the
`
`data that are used for comparison with model data in both series of
`
`embodiments.
`
`
`
`Okay. So Nobots' surreply at 11 that argues that you don't know what
`
`data is being discussed in both series of embodiment statement is wrong.
`
`And the and/or in that statement about both series of embodiments means
`
`the data can include both types of data, but it's also allowed to include just
`
`one or the other. So it can include just available data. And that aligns with
`
`what Claim 20 says, the interest data consists of available data.
`
`
`
`The consists of means the interest data can be only available data,
`
`which is passive data like in Kitts.
`
`
`
`The other Nobots' surreply at Page 11 alleges is that both series of
`
`embodiment statement is somehow discussing additional embodiments that
`
`are alternatives to the first and second series as they are described in the
`
`prior two sentences. But that's wrong because this last sentence says both
`
`series, which is clearly describing what can be used in both the first and
`
`second series just discussed above where the data were acquired before or
`
`acquired after delivery of issued data.
`
`
`
`
`
`
`
`JUDGE WEINSCHENK: Ms. Hunt?
`
`MS. HUNT: Yes?
`
`JUDGE WEINSCHENK: This is Judge Weinschenk again. If we
`
`look at the claim language where it says acquiring interest data, doesn't that
`
`tell us though that it has to be active data because passive data would already
`
`
`
`
`
`
`10
`
`

`

`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
`
`IPR2022-00940
`Patent 9,595,008 B1
`
`
`be available to you so you wouldn't need to acquire it, right? Isn't that the
`
`point of the specification that says passive data is available so you wouldn't
`
`need to acquire passive data. You would only need to acquire active data,
`
`right?
`
`
`
`MS. HUNT: I don't believe the specification says passive data is
`
`available. Available data is another type of data. But it says -- okay, sorry.
`
`Let me clarify. It says passive model data comprises data available from a
`
`computing device that users interactions therewith normally stored, logged
`
`or transmitted to a remote location.
`
`
`
`So it does use the word available to talk about passive data, but that --
`
`I don't see why that wouldn't mean that you can acquire it because I mean
`
`something being available doesn't necessarily mean you already have it. It
`
`just means it is present. You can get it. And what this is saying here is that
`
`you can get it from normal storage, logging and transmitting to a remote
`
`location. That is the way you can get it. You don't have to, for instance,
`
`have some special monitoring software that you send to the client device and
`
`have to install there in order to get it. You can get it from normal
`
`transmissions, but you still have to get it from those transmissions. You still
`
`have to like do that stuff. Does that answer the question?
`
`
`
`
`
`JUDGE WEINSCHENK: It does. Thank you.
`
`MS. HUNT: All right. Okay. So moving to Slide 38. Another
`
`reason Claim 19 covers Kitts' passive data is Claim 20. Now Claim 20 has
`
`an error in its dependency, but as we'll discuss on the next slide, it should
`
`depend from Claim 19. And if it is properly interpreted that way, it is
`
`dispositive on the dispute on Claim 19 because Claim 20 requires the
`
`interest data in 19A to consist of available data, which is passive data.
`
`
`
`
`
`
`11
`
`

`

`IPR2022-00940
`Patent 9,595,008 B1
`
`
`
`
`Now contrary to Nobots' surreply, this is not a claim differentiation
`
`argument. We're not arguing about how much broader than Claim 20 Claim
`
`19 is. We're saying Claim 19 has to at least encompass Claim 20's
`
`embodiment. It has to be able to be met by the embodiment in Claim 20
`
`where interest data consists of available data only.
`
`
`
`Now it makes by far the most sense for Claim 20 to depend from
`
`Claim 19. Claim 20 follows directly after Claim 19. It's not listed together
`
`with Claim 1's independent claims even though all of these claims were
`
`added in prosecution at the same time.
`
`
`
`And the interest data in Claim 20 has antecedent basis in Claim
`
`18 and not in Claim 1. And the prosecution history shows that the applicant
`
`likely just accidentally omitted a numeral when they wrote this claim and its
`
`dependency. So Slide 39 --
`
`
`
`JUDGE WEINSCHENK: Ms. Hunt?
`
`MS. HUNT: Yes?
`
`JUDGE WEINSCHENK: Do we have authority to change the
`
`dependency here in this case?
`
`
`
`MS. HUNT: I think Your Honors need to -- because there is a dispute
`
`about the construction of Claim 19 and the scope of these claims, Your
`
`Honors need to construe what Claim 20 means. And the dependency here is
`
`clearly not what the applicant intended. Clearly there is something wrong in
`
`Claim 20. And the parties have presented their arguments about that is.
`
`I think Your Honors similarly looked at there are other claims in this
`
`patent where there were errors made in prosecution. And Nobots has said
`
`that you should interpret the claims as saying, for example, "of e)" instead of
`
`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
`
`
`
`
`
`
`12
`
`

`

`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
`
`IPR2022-00940
`Patent 9,595,008 B1
`
`
`"of d)" and things like that. And that would be the reason for being able to
`
`construe this claim.
`
`
`
`But I do want to make the point that even if Your Honors don't find
`
`that this rises to the level that you would construe or correct the claim to
`
`depend from Claim 19 and be dispositive in that way, it is still evidence that
`
`is consistent with and supports the interpretation of Claim 19 as being
`
`intended to have the interest data to be able to be met by available data only
`
`because that is what was intended in prosecution what the applicant was
`
`showing there.
`
`
`
`JUDGE COCKS: Ms. Hunt, one question -- well, two questions on
`
`that. You say what was intended in prosecution. The other side, we will no
`
`doubt hear from them, but they have pointed to evidence from the
`
`prosecution that separated out what was in Claim 21, which became Claim
`
`20 here, is specifically intended to be dependent on Claim 1. They've made
`
`that point.
`
`I believe they've also made the point that in reference to interest data
`
`here was a typographical error, and it should be a reference to model data.
`
`Those two points would somehow align, I think, certain aspects of the
`
`claim. I would just like to hear your thoughts preliminarily on those two
`
`points.
`
`
`
`MS. HUNT: Yes, Your Honor. So Slide 39, so on the point of that
`
`interest data refers to the model data. So I think everyone agrees, there is no
`
`interest data recited in Claim 1. So this term interest data in Claim 20 has no
`
`antecedent basis and that can't be right.
`
`
`
`Now the Patent Owner is saying that it should be model data. There is
`
`no reason for it to be model data as opposed to monitored data or neither of
`
`
`
`
`
`
`13
`
`

`

`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
`
`IPR2022-00940
`Patent 9,595,008 B1
`
`
`those, some other kind of data, in Claim 1. There is absolutely not a clear
`
`correction there.
`
`
`
`So the idea that the applicant deliberately put Claim 20 in a separate
`
`set from all of Claim 1's dependents, put it right after Claim 19, gave it the
`
`term that has the same term that is in Claim 19 and not in Claim 1 and
`
`accidentally switched an entire word, that that is more plausible then the fact
`
`that the applicant simply left off a zero when they were writing 20 and wrote
`
`2 instead. I just think it strains credibility. So we've shown this here on
`
`Slides 39 and 40.
`
`
`
`JUDGE COCKS: Okay. And just so I'm clear, have you discussed in
`
`your briefing this potential that the reference to Claim 2 of what was then
`
`Claim 21 during prosecution likely should have been what was intended to
`
`be a reference to Claim 20, such that the zero got dropped off?
`
`
`
`
`
`
`
`MS. HUNT: Yes, Your Honor. This is discussed in our reply --
`
`JUDGE COCKS: Thank you.
`
`MS. HUNT: -- in the discussion of this claim and also in our expert's
`
`declaration.
`
`
`
`All right. So, again, all of this evidence -- so if you interpret Claim 20
`
`this way, it is dispositive on the issue. But even if you did not, it is part of
`
`the totality of the evidence together with what we additionally discussed in
`
`the specification that when Claim 19 is talking about interest data and then
`
`talking about acquiring it at a particular time and place, all of this evidence
`
`together supports that Claim 19 could be by passive data.
`
`And there is another piece of evidence in Nobots' own assertion in the
`
`parallel district court litigation, which is on Slide 41. In their infringement
`
`contentions, Nobots alleged that Google's accused product meets Claim 19
`
`
`
`
`
`
`14
`
`

`

`IPR2022-00940
`Patent 9,595,008 B1
`
`
`by obtaining interest data in the form of passive data. So browser cookies,
`
`user agent, IP address, all of those are only passive data. And this is the only
`
`data that their contentions alleged was obtained prior to delivery of issued
`
`data as 19A requires.
`
`
`
`Now when they get up and speak today, there is some stuff they are
`
`now highlighting in their Slide 33 of their infringement contention that is
`
`buried in a snip from some material on user's cursor movements. But that
`
`happens -- the user's cursor movements happen after delivery of issued data.
`
`And it's not what they were pointing to in their infringement contentions
`
`right above that snip. They were pointing to device history, which is passive
`
`data.
`
`
`
`So, again, this added to the rest of the evidence that shows that
`
`Nobots' argument here is not credible. The idea that the Patent Owner itself,
`
`on their own patent, had a sense that 19A covers obtaining only passive data
`
`in their infringement contentions and then changed their mind once they saw
`
`Google's prior art in the IPR is not credible and shouldn't be allowed.
`
`
`
`So in summary, there are multiple reasons why the evidence shows
`
`Claim 19 reads on Kitts' passive data.
`
`
`
`JUDGE PARVIS: This is Judge Parvis. Do you have -- can you point
`
`us to an embodiment in the specification in which available data is received -
`
`- is acquired from the client computing device?
`
`
`
`MS. HUNT: So we have pointed to Slide 37. This is a very thin
`
`specification. It's very short. So the segment on Slide 37 is the only place in
`
`the specification where it discusses getting interest data, getting monitored
`
`data prior to or after delivery of issued data from the client computing
`
`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
`
`
`
`
`
`
`15
`
`

`

`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
`
`IPR2022-00940
`Patent 9,595,008 B1
`
`
`device. And this is where it says that interest data acquired can be in that
`
`series of embodiments the acquired data type and/or the available data type.
`
`
`
`JUDGE PARVIS: So this is -- the portion of the specification, the
`
`only portion of the specification that Petitioner is relying on is in Column 3,
`
`Lines 47 to 60, is that correct?
`
`
`
`
`
`
`
`MS. HUNT: That's correct. That's the only --
`
`JUDGE PARVIS: (Simultaneous speaking) -- sorry. Go ahead.
`
`MS. HUNT: Sorry. That's the only description of the process that
`
`happens in these claims. Now I do want to clarify, if your question was does
`
`it discuss obtaining available data from the client computing device, I guess,
`
`as opposed to from somewhere else, I mean, all of this data comes from the
`
`client computing device. It's always generated at the client com puting
`
`device from the user's interaction.
`
`For example, it's the client's IP address, the client's browser history.
`
`Those are the client's geographic location. All of these are passive data that
`
`come from the client computing device so they would need to be gone
`
`through.
`
`
`
`JUDGE PARVIS: So I guess it would be helpful then in this portion
`
`of the specification to understand, is Petitioner relying on available data
`
`being acquired by the client computing device?
`
`
`
`MS. HUNT: Yes, Your Honor, because when you put these sentences
`
`together, it says in a first series of embodiments, there is interest data
`
`acquired prior to delivery of issued data. And then in the last sentence, it
`
`says in both series of embodiments, so that includes the first series of
`
`embodiments, acquired data and/or available data may be used for that
`
`comparison with the model data.
`
`
`
`
`
`
`16
`
`

`

`IPR2022-00940
`Patent 9,595,008 B1
`
`
`
`
`So putting these two sentences together, it is saying available data in
`
`the first series of embodiments can be acquired prior to delivery of issued
`
`data to the client computing device. Does that answer your question?
`
`
`
`JUDGE PARVIS: Is there another way of reading that sentence that it
`
`is just talking about the fact that perhaps available data is available but not
`
`necessarily acquired in the first embodiment?
`
`
`
`MS. HUNT: But the first -- the sentence about the first series of
`
`embodiments is defining that first series of embodiments as embodiments in
`
`which a comparison between interest data acquired prior to delivery and
`
`model data is performed.
`
`
`
`So the interest data is being acquired prior to delivery of issued data.
`
`And then that last sentence is saying, the type of data you can get in both of
`
`these embodiments can be acquired data and/or it can be available data.
`
`So the data in both series of embodiments is being acquired. In one
`
`series, it is being acquired prior to delivery. In the other series, it's being
`
`acquired after delivery under the plain meaning of the word acquired. And
`
`then in both those series of embodiments, described in just that way, the type
`
`of data that can be acquired prior or acquired after can be acquired data as
`
`defined and/or available data as defined.
`
`
`
`JUDGE PARVIS: I think I understand Petitioner's position. Thank
`
`you.
`
`
`
`
`
`MS. HUNT: All right. Thank you.
`
`JUDGE WEINSCHENK: Ms. Hunt, I have a question on this issue.
`
`Do we need to resolve this dispute? Do you have a reference that teaches
`
`active data or do you only have references that teach acquiring passive data?
`
`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
`
`
`
`
`
`
`17
`
`

`

`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
`
`IPR2022-00940
`Patent 9,595,008 B1
`
`
`
`
`MS. HUNT: In the ground for Claim 19, the ground is based on Kitts,
`
`which is acquiring passive data. And part of the reason is that it's also
`
`meeting Claim 20, which says the interest data has to consist of available
`
`data, which would only be passive data.
`
`
`
`
`
`JUDGE WEINSCHENK: Okay.
`
`MS. HUNT: All right. So if there aren't further questions on that, I
`
`will move on to the disputed claims that depend from Claim 1.
`
`
`
`All right. So going back to Slide 2, I'm going to talk about Claims 4
`
`and 5. Now there are three grounds for these claims. Two of them are
`
`combinations with Kitts and then there is also an alternative ground with
`
`O'Connell alone.
`
`So I will talk first about the Kitts combinations. The only disputes on
`
`those grounds are motivation and expectation of success. There is no
`
`dispute that the combinations meet the claim limitations.
`
`
`
`So Slide 3 again. Starting with the Willner-Kitts ground, once again
`
`Nobots doesn't dispute that Willner meets Claim 1. And as I said earlier,
`
`Willner is detecting click fraud on web pages with advertisements.
`
`
`
`So Slide 5, Claims 4 and 5 add limitations on using passive model
`
`data. So Ground 2 maps these limitations to Kitts and adds Kitts' use of
`
`passive data to Willner's system that already uses active data. So the
`
`combined system will look at both types of data to detect bots.
`
`
`
`Now like Willner, Kitts is also collecting and analyzing data to detect
`
`click fraud. And it is determining a likelihood score that the user is human
`
`based on whatever data is analyzed just like Willner. So it makes sense that
`
`Willner's system can and would make use of the data that Kitts teaches to
`
`use in the same type of system.
`
`
`
`
`
`
`18
`
`

`

`IPR2022-00940
`Patent 9,595,008 B1
`
`
`
`
`Similarly, Ground 4 maps most of Claim 1 to O'Connell and then
`
`modifies it with Kitts for Claims 4 and 5. O'Connell, again, is in the context
`
`of detecting fraud in a web page interaction by analyzing
`
`various data and determining a likelihood score that the user is a bot.
`
`And in combination with O'Connell, Kitts also teaches the model data
`
`of Limitation 1D. So Nobots in Ground 3 that applies O'Connell alone,
`
`Nobots has an argument that O'Connell doesn't have model data or passive
`
`data. Those arguments don't apply to Ground 4 where Kitts undisputedly
`
`supplies the model data and the passive data.
`
`
`
`And, again, for the Kitts combination Grounds 2 and 4, Nobots only
`
`disputes whether there is motivation and reasonable expectation of success
`
`in the combinations. So I'll talk about motivation first.
`
`
`
`Slide 6. Kitts says that passive data can be an important tool in
`
`determining whether click fraud is occurring and says you can use the
`
`passive attributes in conjunction with other user or program parameters.
`
`And that provides motivation to use passive data in conjunction with the
`
`active data that Willner and O'Connell analyze.
`
`
`
`Slide 7. So the petition cited those teachings in Kitts. In addition, the
`
`petition and Google's expert also cited multiple other references showing
`
`that it was well-known that passive data helps differentiate humans from
`
`bots. And Your Honors correctly agreed at institution that all of these
`
`teachings