[00:00:00] Speaker 03: This case is Vivio, Incorporated versus Comcast Cable Communications, 2018-24-22. [00:00:08] Speaker 03: When everyone has sat down, Ms. [00:00:13] Speaker 03: Kelly can begin. [00:00:15] Speaker 03: Please proceed. [00:00:20] Speaker 01: Thank you. [00:00:21] Speaker 01: Good morning, Your Honors. [00:00:22] Speaker 01: May it please the Court. [00:00:23] Speaker 01: Christina Caggiano Kelly on behalf of Appellant Vivio. [00:00:28] Speaker 01: The board committed legal error in this case by applying a hindsight bias in its obviousness analysis. [00:00:37] Speaker 01: The board essentially took two divergent prior art references, Gross and Smith. [00:00:45] Speaker 01: It deconstructed these references into pieces of their functionality and then reassembled those pieces to create an entirely new algorithm that isn't disclosed in either of these references or their combination. [00:00:59] Speaker 01: It used the 696 patents claimed associating step as a blueprint for doing so. [00:01:07] Speaker 01: The board essentially took [00:01:10] Speaker 01: the content item descriptors out of Gross. [00:01:13] Speaker 01: It put them into the mapping index in Smith. [00:01:18] Speaker 01: But in order to do so, had to use Gross' concept of incremental searching in order to further deconstruct those descriptors into prefixes. [00:01:30] Speaker 03: Findings of obviousness based on combinations of references do. [00:01:37] Speaker 03: One looks at what one reference teaches and what the other teaches and whether there's a motivation to combine them. [00:01:45] Speaker 01: Yes, that is the obviousness analysis and the law of obviousness governs basically where you draw the line between properly combining references and improperly using hindsight bias and the claim as a blueprint to essentially re-engineer the prior art and arrive at the disputed claims without any particular motivation or teaching in the art that would tell you to do so. [00:02:13] Speaker 01: That's what makes this an ultimate legal conclusion. [00:02:17] Speaker 01: Because in this case, there are some small disputes, but there isn't much dispute over what the prior art discloses. [00:02:25] Speaker 01: There isn't a battle of the experts. [00:02:28] Speaker 01: There isn't a whole lot that we need to look at as far as underlying findings of fact to which this court should defer. [00:02:35] Speaker 01: to what the board found below. [00:02:37] Speaker 01: This is an exercise in what does it mean to properly conduct an obviousness analysis. [00:02:44] Speaker 04: Do I understand correctly, as you said, you don't dispute the board's findings and what the prior art teaches. [00:02:49] Speaker 04: You don't dispute the board's finding if there's a motivation to combine. [00:02:54] Speaker 01: Well, no. [00:02:54] Speaker 01: We do dispute the board's finding that there is a motivation to combine. [00:02:58] Speaker 01: And we do dispute, for instance, that the board found that the Smith reference anticipates or teaches prefixes. [00:03:07] Speaker 01: We think that a reading of that reference, that's not [00:03:11] Speaker 01: That's not a correct reading of that reference. [00:03:14] Speaker 01: But ultimately, even if you were to accept what the board did here, there wasn't a motivation to combine these references in the way that the board combined them. [00:03:25] Speaker 01: We don't dispute, for instance, that [00:03:27] Speaker 01: Gross mentions performing its algorithm on a wireless telephone, and that Smith discloses a wireless telephone. [00:03:35] Speaker 01: But if one were to look at Gross and say, oh, here's my motivation to combine it with Smith, as both experts below admitted, the way you would combine them is by using methods that were known at the time of disambiguation, such as multitap, [00:03:52] Speaker 01: or T9, and that would be how you could combine those references. [00:03:57] Speaker 01: Essentially, you would use a disambiguation on Smith's keypad, you would get to your search term, and then if you put that search term into Gross, you could then use Gross's index, and it would work. [00:04:11] Speaker 01: That's not the algorithm that is claimed here. [00:04:15] Speaker 01: Instead, in order to combine them the way the board combined them, which was improper and for which there was no motivation, is the board deconstructed everything and re-engineered an entirely new direct mapping algorithm. [00:04:30] Speaker 01: And nothing in Gross or Smith discloses anything that would teach you a direct mapping algorithm that takes you from the starting point of ambiguous prefixes, goes through descriptor searching on an incremental basis, and then gets you to content items. [00:04:47] Speaker 04: I have a question just about our standard of review. [00:04:50] Speaker 04: I noticed that your brief didn't refer to substantial evidence anywhere. [00:04:54] Speaker 04: What do you think is our standard of review? [00:04:56] Speaker 04: You say this is a legal issue. [00:04:58] Speaker 04: But for us to dispute things like the underlying, or for you to dispute the underlying findings of fact about what the prior teaches, and even if there's a motivation to combine, those are questions of fact reviewed for substantial evidence. [00:05:12] Speaker 04: How do you deal with that? [00:05:14] Speaker 01: So to the extent that the court [00:05:18] Speaker 01: feels that the only way to reach that this ultimate legal issue or to find in Vivio's favor on the ultimate legal issue would be to reverse specific factual findings of the board, then that part would be substantial evidence with you, but we don't. [00:05:33] Speaker 01: Well, because in Vivio's articulation of what this argument is, this is an entirely legal analysis of what is a hindsight bias. [00:05:44] Speaker 01: And I think if you look at hindsight bias, and you look at what cases like Cyclobenzaprine and Yamanouchi tell us about using the claims as a blueprint, that is all this court needs to reverse what the board did here. [00:06:00] Speaker 01: Nothing about what the board did is a proper obviousness analysis and though there are some [00:06:08] Speaker 01: underlying disputes like whether Smith discloses a prefix, ultimately that doesn't matter. [00:06:13] Speaker 01: Because even if you interpret the Smith reference the way the board interprets it, the obviousness analysis ultimately is still improper because they are still using the claim as a blueprint and re-engineering the prior art in order to make it fit, in order to arrive at the claims. [00:06:31] Speaker 01: And I don't think we need to overturn any particular underlying findings in order to do that. [00:06:39] Speaker 01: But once you look at the hindsight bias that the board had in piecing together all of the divergent elements that arrived at the claim, [00:06:51] Speaker 01: you can see that it is a very strained fit for certain elements. [00:06:56] Speaker 01: And I think that's really the point that we're making. [00:06:59] Speaker 01: It is a very strained fit. [00:07:00] Speaker 02: Well, a strained fit doesn't overcome a substantial evidence review. [00:07:04] Speaker 01: And that's why we're not advocating that the court reverse anything on substantial evidence. [00:07:10] Speaker 02: Honestly, I don't understand your argument. [00:07:12] Speaker 02: Your argument to me is we don't think there's substantial evidence for the motivation combined here. [00:07:16] Speaker 02: I understand that you're attempting to morph that into some sort of legal argument, but I truly don't see any legal argument embedded in what you're arguing. [00:07:24] Speaker 02: Well, ultimately, you throw away the relative hindsight bias. [00:07:28] Speaker 02: Hindsight bias is the reason that we have the obviousness test, because we're concerned about hindsight bias. [00:07:35] Speaker 02: It's not somehow a separate component of obviousness that we independently analyze. [00:07:40] Speaker 02: It is what the four parts of the Graham versus John Deere test exist to hopefully prevent. [00:07:46] Speaker 02: And so I just don't understand. [00:07:48] Speaker 02: You're acting like hindsight bias is the fifth prong of the obviousness test. [00:07:52] Speaker 02: I really don't get it. [00:07:53] Speaker 01: Hindsight bias is not itself a prong of the test. [00:07:57] Speaker 01: But hindsight bias, again, as you said, it is what the grand factors exist to prevent. [00:08:03] Speaker 01: It is what makes obviousness ultimately a question of law. [00:08:07] Speaker 02: So if after resolving... Hindsight bias is not what makes obviousness ultimately a question of law. [00:08:14] Speaker 02: I don't even know where you got that from. [00:08:16] Speaker 01: It's just made up. [00:08:18] Speaker 01: Well, the ultimate legal analysis in obviousness [00:08:22] Speaker 01: It is an ultimately illegal conclusion. [00:08:25] Speaker 01: It is based on underlying, but it's not because oh, let's make it a question of law because of hindsight bias That's part of so avoiding hindsight bias is part of what the underlying findings of fact and Graham are designed to avoid as you said, which is which is correct, but ultimately [00:08:49] Speaker 01: What constitutes a proper obviousness analysis and what doesn't is a legal question. [00:08:54] Speaker 01: And hindsight bias informs that legal question. [00:08:58] Speaker 02: No, they applied the four Graham versus John Deere factors in this case in a pretty straightforward fashion. [00:09:03] Speaker 02: They found a motivation to combine. [00:09:05] Speaker 02: You don't like the results. [00:09:07] Speaker 02: I mean, you have a shot at convincing me there's no substantial evidence for a motivation to combine here, but I really don't follow or understand your argument that this is somehow a legal question that goes to the heart of whether they applied the wrong obviousness test. [00:09:20] Speaker 02: I don't see any misapplication of the law here. [00:09:25] Speaker 02: Well, if there is nothing beyond... You know, like the result they reached based on it, but I don't see how they stated the test wrong or applied it improperly. [00:09:34] Speaker 01: Well, if once you go through all of the Graham factors, there is nothing else to do but conclude obviousness, then obviousness isn't a question of law anymore. [00:09:47] Speaker 01: Obviousness becomes simply a question of fact, and Graham factors are completely determinative. [00:09:53] Speaker 01: And this court would never have de novo review over obviousness, which I don't think [00:09:57] Speaker 01: is what the case law tells us. [00:09:59] Speaker 01: Yes, the Graham factors are there. [00:10:01] Speaker 01: They are the underlying findings of fact that ultimately inform the legal question of obviousness. [00:10:07] Speaker 01: And they are designed to help avoid slipping into hindsight bias, which would be a legal error. [00:10:13] Speaker 02: Are you trying to suggest to me that if all four Graham factors point towards obviousness, a court is free to nonetheless conclude there's no obviousness? [00:10:24] Speaker 02: Because it's a question of law, that seems to be what you're saying. [00:10:27] Speaker 02: Well, there is an analytical... If any district court ever listens to this, I will reverse you if you do that. [00:10:33] Speaker 01: Well, in order then for obviousness to be considered a question of law, there needs to still be some sort of legal analysis as far as when does one [00:10:46] Speaker 01: conclude that the question was answered correctly and that the Graham factors were properly employed. [00:10:53] Speaker 01: So that is what it means to be a mixed question of fact and law, which is what obviousness is. [00:11:00] Speaker 01: And so once you resolve the underlying factual issues, if there's nothing left to do, then obvious is not a question of law at all. [00:11:10] Speaker 01: There has to be some legal oversight to, well, we've looked at the underlying factors. [00:11:16] Speaker 01: How are we going to apply the knowledge that we gain from answering those underlying questions into answering what is ultimately a legal question? [00:11:28] Speaker 01: And the case law tells us that using the claim as a blueprint in order to find all of the elements in the prior art and combine them in the way that the claim combines them, that is an improper legal analysis. [00:11:46] Speaker 01: So I think if the decision, then, is that once the underlying findings of fact have been found, there is nothing more for the court to do, then obviousness wouldn't be a question of law, which was not what the case law tells us it is. [00:12:02] Speaker 03: You're into your rebuttal time. [00:12:03] Speaker 03: You can continue, or we'll save it for you. [00:12:07] Speaker 01: I'd like to reserve it. [00:12:08] Speaker 01: Thank you. [00:12:11] Speaker 03: Mr. Meeker. [00:12:19] Speaker 00: May it please the court, Fred Meeker with the law firm of Banner-Witkoff, representing Comcast Cable Communications, LLC. [00:12:28] Speaker 00: The sole issue appealed by Vevio is whether there is substantial evidence to support the board's conclusion that the claims are obvious in view of the Gross-Smith combination with respect to the associating limitation. [00:12:46] Speaker 00: The associating limitation in the claims associate strings of overloaded keys with content items. [00:12:55] Speaker 04: Mr. Meeker, I think I heard the appellant say that they're not relying on substantial evidence. [00:13:01] Speaker 04: They're challenging any of the fact findings made by the board. [00:13:06] Speaker 04: Do you have a different understanding? [00:13:09] Speaker 00: You know, I heard that as well, Judge Moore. [00:13:11] Speaker 00: And I'm not sure. [00:13:12] Speaker 00: That's Judge Stoll. [00:13:13] Speaker 00: I'm sorry, Judge Stoll. [00:13:14] Speaker 00: I apologize, Your Honor. [00:13:17] Speaker 00: I heard that as well. [00:13:21] Speaker 00: And it's not clear to me that [00:13:25] Speaker 00: As I understand her single issue on appeal, if you look at their brief, the single issue on appeal is an issue of whether the associating step is met by the combination [00:13:39] Speaker 00: And in their reply, they confirmed that that is the sole issue on appeal. [00:13:44] Speaker 00: And that associating step, whether it's met by the combination, is a question of substantial evidence, as Your Honor, I think, indicated in your questioning of my [00:13:55] Speaker 00: a friend on the other side. [00:13:57] Speaker 00: I think that is an issue of substantial evidence. [00:14:00] Speaker 00: The ultimate conclusion may be a matter of law, but the review here is on substantial evidence. [00:14:07] Speaker 00: Is the combination met? [00:14:10] Speaker 00: And is there substantial evidence in the record that that combination is met? [00:14:15] Speaker 00: Even though the legal [00:14:17] Speaker 00: conclusion is a question of law. [00:14:20] Speaker 00: It's based on underlying fact issues. [00:14:22] Speaker 00: And in this case, the only underlying facts are submitted by our expert. [00:14:28] Speaker 00: Dr. Fox's original testimony on this issue was provided [00:14:37] Speaker 00: And that's starting at 1006. [00:14:41] Speaker 00: And in particular, the paragraphs related to the motivation to combine are 92 at appendix, I'm sorry, 878, appendix 879, paragraphs 92 to 94 in Fox's rebuttal declaration. [00:15:04] Speaker 00: There's also substantial evidence 1191 through 1194 discuss that motivation to combine. [00:15:17] Speaker 00: And I would say there is no evidence on the other side. [00:15:20] Speaker 00: There's no declaration. [00:15:23] Speaker 00: support on this issue supplied opposing that. [00:15:29] Speaker 00: In the opening brief, they didn't cite any declaration, any expert testimony. [00:15:33] Speaker 04: They don't need to have an expert, though. [00:15:36] Speaker 00: They don't. [00:15:36] Speaker 00: They can cite to the two references, which they did. [00:15:39] Speaker 04: Your point is that you're identifying substantial evidence in support of the board's fact determination. [00:15:45] Speaker 04: But just to clarify, I'm not aware of anything that says that you have to have an expert. [00:15:51] Speaker 00: I agree. [00:15:51] Speaker 04: There could be a proof of an absence of a claim element just through the references themselves, not teaching it. [00:15:58] Speaker 00: I agree. [00:15:58] Speaker 00: But in this case, normally you would identify a missing limitation, something missing in the record between those two references. [00:16:11] Speaker 00: And in this case, we have [00:16:14] Speaker 00: we have, I believe, substantial evidence to support that. [00:16:19] Speaker 00: And if you look at their expert they cited in their reply, he actually addresses different issues than hindsight. [00:16:27] Speaker 00: None of that is addressed in their expert report. [00:16:30] Speaker 00: If you read their expert report, it doesn't refer to hindsight. [00:16:34] Speaker 00: It refers to two things which are [00:16:37] Speaker 00: First two are criticisms of their invention. [00:16:40] Speaker 00: The invention operates with, you know, more imprecision, and the second thing is it operates slower, and then they say, well, you would have used multi-tap, which is one of the arguments they made here today, or T9. [00:16:55] Speaker 00: You know, just because you could have done it differently could have come up with a different combination that doesn't render the combination we provided inappropriate or wrong. [00:17:05] Speaker 00: And then [00:17:08] Speaker 00: Their third thing was really a mischaracterization of the combination. [00:17:12] Speaker 00: So they said you could have replaced Gross's index with Smith's index, and that's not what we argued. [00:17:20] Speaker 00: We argued that you would use Smith to modify the index of Gross. [00:17:25] Speaker 00: So Gross teaches an index for completed words, car, [00:17:31] Speaker 00: or dog, and an alphanumeric index that translates, and it can be an incremental search, but it's an alphanumeric index, whereas Smith teaches a numeric index. [00:17:44] Speaker 00: It teaches how do you convert something that's an alphanumeric index into a numeric index. [00:17:51] Speaker 00: And the court below did hold that Smith teaches prefix as well because it has single character search. [00:17:59] Speaker 00: If you look at the Smith reference, [00:18:04] Speaker 00: It does teach single-character search, even though we didn't rely on it, and it's not something that was necessary for the decision. [00:18:15] Speaker 00: But there is the number three as one of the terms, and that is a single-character incremental search. [00:18:20] Speaker 00: And that's what the board below found as a backup position. [00:18:24] Speaker 00: So they said, even if we had made that combination, which we didn't make and the board found we didn't make, [00:18:31] Speaker 00: They said that Smith still teaches incremental searching and I think that's what they challenge. [00:18:49] Speaker 00: So Judge Moore, I think you had mentioned that hindsight bias is not a separate test for obviousness. [00:18:57] Speaker 00: And we agree with that. [00:19:00] Speaker 00: The hindsight in this case, A, wasn't really addressed below, other than as that is the burden of our combination, not to use hindsight in the combination. [00:19:15] Speaker 00: And secondly, the entire teaching of Smith is how to convert an alphanumeric index into a numeric index. [00:19:25] Speaker 00: And it's literally throughout Smith, but particularly in column five of Smith. [00:19:34] Speaker 00: Okay, so getting back to the sole issue on appeal, which is the associating stuff, associating overloaded keys to content items, we think that is taught in the Gross-Smith combination. [00:19:55] Speaker 00: The arguments with respect to whether or not content items are [00:20:08] Speaker 00: are the related or cannot be in the content itself, where the descriptors can't be in the content. [00:20:15] Speaker 00: So first of all, it doesn't matter for our combination. [00:20:18] Speaker 00: Growth has descriptors in the content and outside of the content. [00:20:23] Speaker 00: Growth has two search paths. [00:20:27] Speaker 00: And one of the search paths is for descriptors that are inside the content. [00:20:34] Speaker 00: And one of the search paths is for descriptors outside of the content. [00:20:38] Speaker 00: And in fact, if you go through all of the various [00:20:43] Speaker 00: flow diagrams in gross, you'll see that it in fact uses two different determinations, one for descriptors inside the and one for descriptors outside. [00:20:58] Speaker 00: So if you go to, for example, [00:21:02] Speaker 00: Figure 4B, which is Appendix 767, there are actually two search paths. [00:21:09] Speaker 00: So receive search string entered into files main search field. [00:21:14] Speaker 00: And so that is an incremental search with respect to the content of the files. [00:21:20] Speaker 00: And it says incremental search below that. [00:21:22] Speaker 00: And then Figure 4B, a separate path, is receive search string [00:21:28] Speaker 00: in attributes field and an incrementally-performed search in the attributes field. [00:21:33] Speaker 00: So GROSS allows you to search on things like date, type of file, as well as the content of the file. [00:21:41] Speaker 00: So GROSS actually teaches both. [00:21:44] Speaker 00: GROSS also teaches prefix strings. [00:21:48] Speaker 00: So even under Roevey's constructions or Roevey's argument of the interpretation of the board's construction, we still win under these references. [00:21:57] Speaker 03: You mean should win. [00:22:01] Speaker 00: We should win. [00:22:03] Speaker 00: That's exactly right, Ms. [00:22:05] Speaker 00: Laurie. [00:22:07] Speaker 02: It seems like you've covered all your points. [00:22:11] Speaker 00: I really have, Your Honor. [00:22:12] Speaker 00: I mean, honestly, if there's something else that I can answer, any questions, I'd be happy to answer them. [00:22:18] Speaker 00: But I think we've covered it all. [00:22:21] Speaker 03: No demerits from not using up all your time. [00:22:24] Speaker 00: Thank you. [00:22:26] Speaker 03: Miss Kelly, you have some rebuttal time. [00:22:33] Speaker 01: Thank you, Your Honor. [00:22:35] Speaker 01: So the parties [00:22:37] Speaker 01: can dispute the standard of review. [00:22:39] Speaker 01: Both standards, substantial evidence and de novo, are listed in the standard of review section in our brief. [00:22:45] Speaker 01: We advocate that this is ultimately a legal error. [00:22:49] Speaker 01: We use the terminology of de novo review throughout our brief. [00:22:55] Speaker 01: But even if this court were to find that ultimately it must apply substantial evidence review, our question presented remains the same. [00:23:03] Speaker 01: Our arguments remain the same. [00:23:06] Speaker 01: We characterize them as legal arguments because they go towards the ultimate issue of obviousness as opposed to the underlying Graham factors. [00:23:17] Speaker 01: But either way, the board committed an error in the way it conducted its analysis. [00:23:22] Speaker 01: And I think that argument basically remains the same regardless of whether you want to characterize it as substantial evidence review or de novo. [00:23:33] Speaker 01: Comcast. [00:23:35] Speaker 01: attorney mentioned that the only facts in identifying substantial evidence, the only facts in the record were those submitted by their expert, Dr. Fox. [00:23:45] Speaker 01: But Dr. Fox's testimony relies on his opinion that content items and descriptors are the same. [00:23:53] Speaker 01: And the board expressly rejected that opinion and stated in its final written decision that it was not relying on Dr. Fox's testimony in that regard. [00:24:02] Speaker 01: So again, what we're left with [00:24:05] Speaker 01: is the board essentially setting aside the opinions of both experts and conducting its own review, but doing so under an improper rubric. [00:24:15] Speaker 01: The Comcast attorney also said that we did not identify anything in the claims that was actually missing from the combination, but we did. [00:24:22] Speaker 01: That was the associating step, the step that we discussed throughout our brief. [00:24:26] Speaker 01: That is the step that is missing even from the combination. [00:24:30] Speaker 01: And then also he mentioned that Smith discloses a single character, which is the [00:24:38] Speaker 01: one of the what could be characterized as underlying facts. [00:24:41] Speaker 01: But again, it doesn't matter to the ultimate analysis. [00:24:45] Speaker 01: But the single character discussed in Smith's index is not a prefix. [00:24:50] Speaker 01: It is a situation where a full search term happens to be a single character long. [00:24:55] Speaker 01: And we made that distinction again below. [00:24:59] Speaker 01: So if the court has any other questions. [00:25:02] Speaker 03: Thank you, counsel. [00:25:03] Speaker 03: The case is submitted.