`571-272-7822
`
`
`
`
`Paper No. 27
`Entered: July 30, 2021
`
`UNITED STATES PATENT AND TRADEMARK OFFICE
`____________
`
`BEFORE THE PATENT TRIAL AND APPEAL BOARD
`____________
`
`GOOGLE LLC,
`Petitioner,
`
`v.
`
`UNILOC 2017 LLC,
`Patent Owner.
`____________
`
`IPR2020-00755
`Patent 6,366,908 B1
`____________
`
`Record of Oral Hearing
`Held virtually: July 15, 2021
`
`
`
`
`Before SALLY C. MEDLEY, KRISTEN L. DROESCH, and
`SHIELA F. McSHANE, Administrative Patent Judges.
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`IPR2020-00755
`Patent 6,366,908 B1
`
`
`
`APPEARANCES:
`
`ON BEHALF OF THE PETITIONER:
`
`
`JOSHUA L. GOLDBERG, ESQ.
`ERIKA H. ARNER, ESQ.
`Finnegan, Henderson, Farabow,
`Garrett & Dunner, LLP
`901 New York Avenue, NW
`Washington, D.C. 20001-4413
`
`
`
`ON BEHALF OF THE PATENT OWNER:
`
`
`BRETT A. MANGRUM, ESQ.
`Ethridge Law Group
`1100 Queensboro Boulevard, Unit 200
`Mt. Pleasant, South Carolina 29464
`
`Also Present, Observing:
`
`JOE SHEERER
`STEVE PEDERSEN
`
`
`
`
`The above-entitled matter came on for hearing on Thursday, July 15,
`
`2021, commencing at 1:01 p.m. EST, by video/by telephone.
`
`
`
`
`
`
`
`
`
`
`2
`
`
`
`IPR2020-00755
`Patent 6,366,908 B1
`
`
`P R O C E E D I N G S
`- - - - -
` (Proceedings begin at 1:01 p.m.)
` JUDGE MEDLEY: Good afternoon.
` This is the hearing for IPR2020-00755 between
`Google LLC and Uniloc 2017 LLC involving U.S. Patent No.
`6,366,908.
` I am Judge Medley, and with me are Judges Droesch
` and McShane.
` At this time, we'd like the parties to please
` introduce counsel for the record, beginning with the
` Petitioner.
` MR. GOLDBERG: Good afternoon, Your Honor.
` This is Joshua Goldberg for Petitioner Google.
` With me I have my lead counsel Erika Arner, and on
`the phone, Joe Sheerer, IP counsel at Google.
` JUDGE MEDLEY: And Mr. Goldberg, you will be
`arguing? Presenting argument?
` MR. GOLDBERG: Yes, Your Honor.
` JUDGE MEDLEY: Okay, thank you.
` And for Patent Owner, who do we have?
` MR. MANGRUM: Good afternoon, Your Honor.
` This is Brett Mangrum for Patent Owner. I'd like to
`also announce that Steve Pedersen, in-house counsel for
`patent owner, is listening on the line. And I will be
`arguing today on behalf of Patent Owner.
`
`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
`
`3
`
`
`
`IPR2020-00755
`Patent 6,366,908 B1
`
` JUDGE MEDLEY: Okay, thanks so much.
` All right. So each party, as you know from our
` order, has 45 minutes total time to present your arguments.
` Petitioner, you'll proceed first and may reserve
` some of your argument time to respond to arguments presented
` by the Patent Owner.
` Thereafter, the Patent Owner will respond to
` Petitioner's presentation and may reserve argument time for
` surrebuttal.
` Mr. Goldberg, do you wish to reserve some of your
` time to respond, and how much?
` MR. GOLDBERG: Yes, Your Honor. I will reserve 20
`minutes.
` JUDGE MEDLEY: Thank you.
` And Mr. Mangrum, would you like to reserve time, and
`how much?
` MR. MANGRUM: Yes. I'd like to reserve time. I
`anticipate needing no longer than 10 minutes.
` JUDGE MEDLEY: Okay, thank you.
` Okay. Mr. Goldberg, when you're ready, we're ready
`for you.
` MR. GOLDBERG: Thank you, Your Honor.
` I'd like to start with Slide 2.
` As you may recall from the Institution Decision,
`Google had eight grounds in its Petition. All of them were
`instituted.
`
`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
`
`4
`
`
`
`IPR2020-00755
`Patent 6,366,908 B1
`
` There are seven of those grounds, however, that
`Uniloc has not presented any arguments for other than saying
`because Ground 1 fails, those other grounds fail, so today
`I'm going to focus on just the first ground, that is that
`Claims 6 to 12 are unpatentable as obvious over
`Braden-Harder in view of Grossman.
` Turning to Slide 3.
` There are actually a very limited number of issues
`that the Board needs to decide here.
` The level of ordinary skill in the art is not
`disputed by Uniloc. They have not presented an alternative
`so there's no need for the Board to address any dispute
`there.
` As for the step-plus-function terms, although there
`was a dispute as to whether the terms in Claim 6 are
`governed by step-plus-function or not, the outcome, as the
` Board recognized in its Institution Decision, does not turn
` on that.
` Regardless of whether the claims are interpreted
` under step-plus-function or not, Google has shown that the
` claims are unpatentable and, therefore, again, the Board
` really doesn't need to decide that issue.
` In addition, the claim construction of the term
` keyfact is another one where, either way, whether keyfact is
` interpreted under Uniloc's original litigation construction
` or interpreted under its much narrower new construction that
`
`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
`
`5
`
`
`
`IPR2020-00755
`Patent 6,366,908 B1
`
` it introduced in its Patent Owner Response, either way we've
` shown that the claims are unpatentable.
` And I just want to pause for a second. I'm hearing
` some kind of background noise. I want to make sure the
` judges can hear me okay.
` Okay, thank you.
` JUDGE DROESCH: I'm sorry. This is Judge Droesch
`speaking.
` And I believe somebody does not have themselves on
`mute. If everybody could just doublecheck that you're on
`mute. I hear some background noise.
` MR. GOLDBERG: Turning to the last point here.
` We, in our Petition, have argued that in that first
`ground, Braden-Harder in view of Grossman, that Heidorn and
`Messerly are incorporated into Braden-Harder.
` And again, there has been no dispute from Uniloc
`about whether that incorporation was proper, so again, that's
`another issue that the Board does not need to decide.
` Turning to Slide 4.
` The issues are also very limited here because
` Google's expert testimony remains unrebutted. Uniloc did
` not introduce any new evidence post Institution. The
` record, as it stood at the Institution Decision, is exactly
` the record as it stands today.
` Uniloc provided no expert declaration to rebut
` Dr. Jansen's expert testimony, Uniloc did not seek to depose
`
`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
`
`6
`
`
`
`IPR2020-00755
`Patent 6,366,908 B1
`
` or cross examine Dr. Jansen, and Uniloc did not provide any
` other evidence to change the Board's reasoning in the
` Institution Decision. All Uniloc presented was bare
` attorney arguments.
` If we turn to Slide 5, we can see here an overview
` of Claim 6, the three big steps color-coded as they were
` in the Petition; keyfact extracting step, keyfact indexing
` step, and keyfact retrieving step.
` I'm going to focus today on the keyfact extracting
` step because that's really the only place that Uniloc has
` presented any new, meaningful arguments since the
` Institution Decision.
` They did have some arguments about the keyfact
` retrieving step, but they're identical to what the Board
` already saw and addressed at its Institution Decision so
` I'll leave those for now, unless the Board has specific
` questions on them.
` Turning to Slide 6, we can see a summary here of the
` different positions that exist for the term keyfact.
` As I noted earlier, Uniloc has a litigation
` construction for this that it argued to the District Court.
` They told the District Court that, "keyfact means a fact
` contained in sentences," and Google adopted --
` JUDGE MEDLEY: Mr. Goldberg?
` MR. GOLDBERG: Yes.
` JUDGE MEDLEY: Excuse me for interrupting. I have a
`
`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
`
`7
`
`
`
`IPR2020-00755
`Patent 6,366,908 B1
`
`question before we get into keyfact and what that means and
`all of that.
` I just want to go back to the -- I believe it's your
`position that the claims invoke 112-6?
` MR. GOLDBERG: Yes, Your Honor, that is our
`position, but we don't think that it's necessary for the
`Board to decide that because the claims are actually as noted
`in the Institution Decision. It's kind of a broader view
`when you don't look at the means-plus-function, so I think
`either way the claims have been shown unpatentable and the
`Board really doesn't have to make that decision.
` JUDGE MEDLEY: Okay. I'm trying to wrap my head
`around all of the issues raised by Patent Owner.
` It seems to me they're saying, well, it does matter
`because the way that you're construing the claims then is
`that the claims just contain function and that you look to
`the spec then for the corresponding acts. That they're sayingthe claims
`actually include the acts and that you don't address
`the acts that are in the claim language. I believe that's what
`they're saying.
` Is that really kind of form over substance so
`that, in respect, you have addressed the acts, you just do so by
`then looking to the specification; is that correct?
` MR. GOLDBERG: Yeah, that's exactly correct, Your
`Honor. And let me actually jump to that and we can come back
`to the keyfact issue later.
`
`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
`
`8
`
`
`
`IPR2020-00755
`Patent 6,366,908 B1
`
` If you want to turn to our Slide 11, we have here
`Claim 6 and Claim 7 shown.
` And the main -- well, in fact, really the only part
`of the claim that Uniloc actually argues this for is for why
`our Petition is deficient in its mappings to the prior art
`as the analyzing step shown in blue, which is part of the
`keyfact extracting step of Claim 6.
` And we can see here the blue shown in Claim 6 and
`then the corresponding blue shown in Claim 7.
` And one of the key issues that I think it's
`important to recognize here is that Claim 7 does not refer
`to the act of Claim 6, and Claim 6 doesn't expressly recite
`any act.
` Claim 6 recites a series of steps, the keyfact
`extracting step is one of them.
` Claim 7 explains what some potential substeps are
`that are basically an embodiment of the keyfact extracting
`step from the specification.
` And if we turn to Slide 7, we can see how this
`actually impacts the analyzing, which, again, is the only
`one that Uniloc actually pointed to when it was disputing
`our Petition mappings.
` And on the right-hand side, you can see, at the top
`of the slide -- this is in the '908 patent at Column 5,
`lines 15 to 19 -- it talks about how analyzing a document is
`done, and it says, "The way you analyze a document is
`
`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
`
`9
`
`
`
`IPR2020-00755
`Patent 6,366,908 B1
`
`through the processes of morphological analysis."
` And then we can see in Claim 6, you've got analyzing
`the document. You go down to Claim 7, and there's the
`analyzing morphology, which is the way that you analyze a
`document.
` So by showing that these different parts of Claim 7,
`which, you know, in fact, we actually ended up mapping them
`in the context of Claim 6 as well because we were arguing
`for a means-plus-function -- a step-plus-function
`construction there-met by the Braden-Harder reference.
` And we can see further, if we want to look at the
`case law in this type of issue, if we go to Slide 13, we
`have this Realtime Data case from the Federal Circuit that
`has claims that are essentially identical to the claims that
`are at issue in the '908 patent.
` On the left side we can see an independent claim
`that recites a method with a series of steps. One of those
`steps is maintaining a dictionary, and there's a dependent
`claim that specifically talks about what that step is.
` It says, "Wherein the step of maintaining a
`dictionary comprises the steps of," and then it recites a
`series of steps.
` This is identical in form to what we're dealing in
`the '908 patent. We can see on the right side Claim 6 has a
`keyfact extracting step.
` Claim 7 using identical language to what was in
`
`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
`
`10
`
`
`
`IPR2020-00755
`Patent 6,366,908 B1
`
`Claim 4 in the Realtime case, says that, "Wherein said step
`of keyfact extracting comprises the steps of," and then it
`recites the series of steps.
` And we can see here from the Federal Circuit, the
`claim elements introduced in Dependent Claim 4 with
`comprising language are properly understood as giving
`details sufficient to constitute a particular embodiment,
`and that's exactly what we just saw when we went to the
`specification of the '908 patent. The morphological
`analysis that's happening in Claim 7 is just the way that
`you do the analyzing that is happening in the keyfact
`extracting step from Claim 6.
` And if we turn to Slide 14, we can see how there
`were some mappings done for this in the Petition and by
`Dr. Jansen. The analyzing a document, we mapped that to
`Braden-Harder. You can see, again, that the color coding,
`the blue in Figure 10A and Figure 11. Dr. Jansen's
`completely unrebutted expert testimony that Braden-Harder's
`triple generation process 1100 separately analyzes textual
`phrases in the document and constructs a corresponding
`set of logical form triples for that document within
`dataset 1030, so there we've got analyzing a document.
` Uniloc has also complained that somehow we're only
`addressing a document. They're only addressing a user
`query. Well, that's not the case, we addressed both.
` We can see on Slide 15. Again, completely
`
`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
`
`11
`
`
`
`IPR2020-00755
`Patent 6,366,908 B1
`
`unrebutted testimony from Dr. Jansen. Braden-Harder teaches
`this similar retrieval process Step 1245 for analyzing a
`user query to yield logical form triples.
` And again, we have the, in blue, Figure 12A, some
`annotations there.
` If we go to Slide 16, we can see, when you start
`bringing in the Claim 7, analyzing morphology, we
`additionally have further citations to the Dr. Jansen
`declaration, and also to Braden-Harder. That completely
`unrebutted testimony from Dr. Jansen that Braden-Harder's
` morphological analysis stems or normalizes differing word
` forms, for example, verb tense and singular-plural noun
` variations to a common morphological form for use by a
` parser.
` We can see down at the bottom right, again, in blue
` there, corresponding to the tag sequences with different
` part of speech in blue, examples there of various forms; you
` have nouns, verbs.
` So we're hitting all of the limitations of these
` claims, and therefore, I really --
` I think Your Honor said something about maybe
` Uniloc's argument was kind of a form over substance thing. I'm
` not sure whether it even rises to that. I think it's just a
` legally and factually incorrect argument that they're
` making, and under a proper analysis of the claims, and
` looking at Braden-Harder, it clearly discloses everything
`
`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
`
`12
`
`
`
`IPR2020-00755
`Patent 6,366,908 B1
`
` that is in these analyzing steps that Uniloc is mapping to.
` Returning back to --
` JUDGE MEDLEY: Okay, thank you. Just a quick
`question.
` What is your take on Claim 6, which you say invokes
` 112-6? We have keyfact extracting step for doing something,
` for example. What is your take on that, and then Claim 7
` recites "a step of". What is the difference there?
` I mean, we know sep for is --
` MR. GOLDBERG: Yeah.
` JUDGE MEDLEY: -- that's, sort of, the trigger.
` But then is this just sloppy claim drafting or is it
`that that's an intent that Claim 6 is not -- does not invoke
`112-6?
` MR. GOLDBERG: So first of all, just to make sure we
`all on the same page here, we had argued Claim 6, which
`should be interpreted as step-plus-function, but we did not
`argue that the dependent claims should be interpreted as
`step-plus-function.
` As for the difference between the use of "for" and
` the use of "of" between the two claims, I think that that is
` actually indicative of what is happening because of the
` means-plus-function in Claim 6 and not in 7.
` In Claim 6, we're talking about keyfact extracting
` step for and we're reciting functions, and then we look to the
` specification and we see what are the corresponding acts.
`
`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
`
`13
`
`
`
`IPR2020-00755
`Patent 6,366,908 B1
`
` Now, it happens to be that the acts end up matching
` up to what is in Claim 7, and therefore, that Claim 7 is --
`under a step-plus-function type of analysis, it's
`essentially reciting what the keyfact extracting step
`becomes under means-plus-function.
` But I'll note that Uniloc had argued about claim
`differentiation, and under claim differentiation, when
`you're in step-plus-function, it's not something that
`becomes an issue because you're stuck with the independent
`claim says what it says and it doesn't matter whether the
`dependent claim says something different.
` Now, as a result of Dependent Claim 7 then
`introducing what corresponds effectively to the
`means-plus-function -- sorry -- step-plus-function
`interpretation of the keyfact extracting step, we see that,
`again, it's just showing what is in the specification, it's
`an embodiment.
` And either way, Google showed that these various
`steps in Claim 7 are taught by Braden-Harder, so therefore,
`they're all met, and it really doesn't matter which way you
`look at this analysis because, either way, everything in
`Claim 6 and everything in Claim 7 are taught by
`Braden-Harder.
` Does that answer your question?
` JUDGE MEDLEY: Yes, thank you. You can continue,
`please.
`
`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
`
`14
`
`
`
`IPR2020-00755
`Patent 6,366,908 B1
`
` MR. GOLDBERG: Okay. I'll just actually note one
`more thing on that point.
` Uniloc, I think, also argued that the Board should
` essentially just treat that word "for" as "comprising", even
` though it says the word "for". If the word had been
` "comprising" then that would be a different situation, but,
` of course, that's not how the claim was actually written.
` So going back to the keyfact issue now again on
` Claim 6.
` As I mentioned earlier, Uniloc had argued in the
` litigation that, "Keyfact means a fact contained in
` sentences."
` Google adopted that construction for purposes of
` this IPR and specifically proposed using that construction
` in this IPR.
` Uniloc then, in its Patent Owner Response, came up
` with a completely new construction that is much narrower
` than what it had argued in the litigation and said that,
` "Now it needs to be a factual extraction of a sentence which
` expresses semantic relation between words and a sentence in
`the form of object, property."
` I'll note that this object, property part seems to
`be what they're mainly focusing on, that it needs to be in
`that specific format, and I also note that Uniloc did not
`present any arguments under its litigation construction.
` So in the instance that the Board was defining that
`
`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
`
`15
`
`
`
`IPR2020-00755
`Patent 6,366,908 B1
`
`keyfact means fact contained in sentences, and then it's
`Uniloc's argument supply because they're all based on their
`new narrower claim construction.
` But it doesn't really matter whether the Board
`construes it as fact contained in sentences or Uniloc's much
`narrower construction, and the Board, for that reason, in
`our view, doesn't even really have to construe it.
` Either way, Google has shown that this is taught by
`Braden-Harder.
` If we turn to Claim 7, the reason that Google got
`into this object, property form in the Petition is because,
`as we just discussed, Uniloc -- I'm sorry -- Google had
`argued that it would be a step-plus-function in its
`Petition, and under a step-plus-function analysis, you then
`need to look to the specification to see exactly what the
`acts are, and in that step-plus-function world, we would
` agree that you need to be having this object, property type
` format.
` We can see on Slide 7 the, again, color coding that
` shows how all of the different parts of even Uniloc's new
` construction were addressed in the Petition.
` We can see in green, Petition at 34, that the
` logical form triples, which, equivalent to the claimed
` keyfacts, are facts contained in sentences and provide a
` concept-based retrieval, so we've got the factual part.
` As for the extraction of the sentence part,
`
`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
`
`16
`
`
`
`IPR2020-00755
`Patent 6,366,908 B1
`
` Petition, page 45, we can see logical form triples are
` extracted based on the semantic relationships of the nouns
` with other words in the sentence, so we've got that part.
` As far as the orange goes --
` JUDGE MEDLEY: And just to let you know, you're --
`if you want to still have 20 minutes left, that time comes up
`at 1:28 according to my clock, just to let you know.
` MR. GOLDBERG: Okay. Thank you, Your Honor.
` On the orange part, I want to note a slight
` difference here and explain it.
` Uniloc had said object, property. You'll see at
` Petition, page 47 -- this is on the bottom left of the
`slide -- we had said object, modifier.
` The difference between property and modifier is
`inconsequential. We can see from the '908 patent at the
`bottom right that the property is the modifier, those are
`the same things so that is a distinction without a
`difference.
` But looking at the full construction again, we can
`see at Petition, page 47, and again, this is to the extent
`the limitation requires generating keyfacts that have the
`forms object, property under 112-6, a person of ordinary
`skill in the art would have understood that Braden-Harder
`generates logical form triples where the case of a
`constituent with noun modifiers, such as “Nadj” and “Mods”
`relations which exhibit the semantic relationship of the
`
`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
`
`17
`
`
`
`IPR2020-00755
`Patent 6,366,908 B1
`
`form object, modifier.
` It's not just Google saying this, and we also have
`support from our expert, Dr. Jansen. If we turn to Slide 8.
`And again, this is completely unrebutted. Uniloc didn't put
`in its own expert testimony, it also did not even seek to
`depose Dr. Jansen.
` He explains, as we just discussed, how a person of
`ordinary skill in the art would have understood that
` Braden-Harder generates logical form triples for the case of
` a constituent with non-modifiers, such as “Nadj” and “Mods”
` relations that exhibit the semantic relationship of form
` object, modifier, and included is relations of an express
` part of the tuple object, mods, modifier.
` Moreover --
` JUDGE MEDLEY: Can I ask a question, just to clarify
`the record --
` MR. GOLDBERG: Yes.
` JUDGE MEDLEY: -- or for me.
` On your Slide 8, out of those eight listed there,
`then what would be the object and then what's the property?
` MR. GOLDBERG: So if you look at the bottom right
`where it has the “Mods” examples, you could think about it as,
`looking at the first “Mods” example, that “BOWL” would be the
`object and “SHARK” would be the modifier for the property.
` JUDGE MEDLEY: Okay.
` MR. GOLDBERG: And --
`
`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
`
`18
`
`
`
`IPR2020-00755
`Patent 6,366,908 B1
`
` JUDGE MEDLEY: Thank you.
` MR. GOLDBERG: -- Dr. Jansen goes on to explain
`that, because Braden-Harder generates logical form triples
`for non-modifiers, Braden-Harder generates keyfacts having
`the forms object, modifier.
` Now, this is perfectly fine, and we could stop here,
` but Dr. Jansen actually went even further and explained, if
` we turn to Slide 9, how even if you needed it to literally
` be in this exact form of object and property, and that's it
` and nothing else is allowed, and there's nothing in the
` specification of the '908 patent that says anything about
` not allowing anything else to be involved in the keyfact,
` but even if that was a requirement, he testifies that it
` would have been obvious to generate other logic forms that
` do not include their relations.
` So in other words, it gets rid of that mod that's in
` the middle of the tuple.
` And from the Braden-Harder reference itself, it
` recognizes that it can readily utilize any other form that
` can portray a semantic relationship. And it talks about
` how, in doing the searching and the matching that it does,
` these things can be relaxed to encompass paraphrases.
` If we go to Slide 10, we have further testimony from
` Dr. Jansen talking about how this modification works and how
` it's so simple. He says that, "A person of ordinary skill
` in the art would have been capable of implementing
`
`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
`
`19
`
`
`
`IPR2020-00755
`Patent 6,366,908 B1
`
` Braden-Harder to extract in other logics forms because doing
` so would require minimal modification without undo
` experimentation."
` Uniloc had no arguments about this in its Patent
` Owner Response. It in no way disputed that this would be
` obvious.
` Uniloc instead, in its Sur-reply for the first time,
` improperly introduced new arguments. We don't think the
` Board should get to those because Uniloc raised them too
` late and they really shouldn't be considered, but even if
` the Board does consider Uniloc's new Sur-reply arguments,
` even in that case those arguments fail. Uniloc had argued,
` for example, in its Patent Owner response that --
` JUDGE MEDLEY: Mr. Goldberg, could you please
`identify exactly where in the Sur-reply you are asserting that
`those -- that's the new arguments, where they are?
` MR. GOLDBERG: Yes. Give me one second.
` JUDGE MEDLEY: Thank you.
` MR. GOLDBERG: So for example, Your Honor, in the
`Sur-reply, the entirety of Section III.A.4.
` JUDGE MEDLEY: Okay, thank you.
` Okay, you can continue. I'm just letting you know
`you're going into your 20 minutes.
` MR. GOLDBERG: Yeah, I realize that, Your Honor. I
`will be quick about it.
` So one of their arguments was that somehow doing
`
`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
`
`20
`
`
`
`IPR2020-00755
`Patent 6,366,908 B1
`
` this modification would render Braden-Harder unable to
` fulfill its intended purpose.
` First of all, that's just bare attorney argument.
` They have no expert testimony about this.
` Secondly, they don't explain why that's the case.
` And thirdly, we have, as we just went over -- in
` Braden-Harder itself it recognizes that other forms are
` possible, and we have the testimony from Dr. Jansen
` completely unrebutted explaining how simple this would be so
` there's no issue with it rendering it inoperable for its
` intended purpose.
` Uniloc also made an argument that somehow this would
` make it so that other Google arguments related to other
` limitations of the claims would no longer be feasible
` because we relied on that relationship label.
` We did not rely on that relationship label for a
` semantic relationship, which is what they're saying we
` relied upon it for.
` I'll note that the claims don't recite semantic
`relation. Even if you go by Uniloc's narrow construction
`and say that maybe they do, we still didn't point to that
`label in making our arguments.
` Instead, the Petition explains that, "Semantic rules
`are used to provide a pattern representing the sentence in
`logical form." This is at Petition 45. "And graph walk
`rules are used to extract the triples based on semantic
`
`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
`
`21
`
`
`
`IPR2020-00755
`Patent 6,366,908 B1
`
`relationships of the nouns with other words in the
`sentence," still in Petition, 45.
` And what this effectively means is that these rules,
`these patterns, it's the act of doing the process, that is
`what’s making it so that there are semantic relationships that
`are being analyzed, they're semantic relationships that are
`being represented by the triples. That's what Braden-Harder
`is all about.
` And with that, unless there's any questions, I'll
`reserve the remainder of my time for rebuttal.
` JUDGE MEDLEY: Okay, thank you.
` You have approximately 17 minutes left.
` And just one moment. Let me reset.
` Okay. Mr. Mangrum, when you're ready, you may
` begin.
` MR. MANGRUM: Thank you. Brett Mangrum for Patent
`Owner.
` And I'm prepared to take our positions and
` highlights from our demonstratives in order. However, in
` view of Judge Medley's questioning, it might make sense to
` start with the step-plus-function issue, and I'm happy to do
` it in whatever way would be the most helpful for Your
` Honors.
` JUDGE MEDLEY: That's fine. It's your time.
`However you want to divide it up is fine.
` MR. MANGRUM: Okay. Let me do it in reverse order
`
`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
`
`22
`
`
`
`IPR2020-00755
`Patent 6,366,908 B1
`
`from the slides then in view of the questioning.
` I'm going to first address the step-plus-function
` issue. I don't believe it was accurately characterized, the
` issue itself, by opposing counsel.
` And I do think -- it is important to ask whether or
` not this is a form-over-substance question -- that was Judge
` Medley's question -- and it's not.
` And in fact, when we heard arguments today, it only
` underscored the problems with the constructions applied in
` the Petition, and that was specifically in relation to a
`comparison of Claims 6 and Dependent Claim 7, and I'm going
`to emphasize that from our slides.
` So if Your Honors can turn with me to Slide 7,
`that's where we start addressing -- I'm sorry, Slide 7 of
`Patent Owner's demonstratives is where we start addressing
`the step-plus-function issues.
` We highlighted some points from the brief in
`Slide 7, and the important takeaways I want Your Honors to
`focus on here is that, of course, the preliminary holding
`that Claim 6 recites acts in connection with each of the
`three fundamental, what are introduced as, steps.
` So for context, we'll go back to Slide 7, but why
`don't you turn with me to Slide 1 where the claim language
`is reproduced.
` What we've done here is indented essentially the
`three paragraphs of Claim 6 in Slide 2.
`
`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
`
`23
`
`
`
`IPR2020-00755
`Patent 6,366,908 B1
`
` You see a keyfact extracting step for, and then there
`are a series of acts that are recited.
` A keyfact indexing step for, and that's followed by a
`series of acts. And each of these acts are introduced by
`verbage and, like, such as extracting, analyzing,
`et cetera.
` And we agree with the Board's preliminary assessment
`that those are acts and that that takes us out of the mean
`or step-plus-function rubric, or law, case law.
` What's impor