# Oral argument — Arendi S.A.R.L. v. Google Inc. > Briefs, arguments, decisions, and more. URL: https://www.frixlaw.com/law-library/documents/argument%3A26723 ## Record - **Collection:** Oral argument - **Document type:** Oral argument - **Published:** November 7, 2016 ## Text The first argued case this morning is No. 16-1249, Arendi S.A.R.L. v. Google Inc., Mr. Sunstein. Proceed. Honorable Judges, good morning. The Court should reverse the Board's decision invalidating the claims because the Board adopted a claim construction contrary to a clear prosecution history disclaimer. To distinguish the prior art, the applicant amended the claims to add an analyzing step that occurs automatically upon a single entry of the execute command. The amendment and accompanying remarks expressly disclaimed user assistance, such as by selection of the first information or the text to be analyzed as a condition for the analyzing. The Board's disregard of this disclaimer in its claim construction was error under Proxicon and Omega Engineering. Under the proper construction, the claims are valid. The Court should reverse. So the prosecution history contains a clear and unmistakable disclaimer that was required to obtain allowance over the so prior art. The amendment added the requirement of a computer process for, this is a quote, analyzing the document to determine if the first information is contained therein. Because the claim covers a computerized method, the amendment means that the determination if the information is contained therein is a step carried out by a computer, not by a human. So the analyzing happens by the computer in contrast to the manual text selection by a user that is required by the so prior art. In making the amendment, the applicant made this very point and further emphasized that the computerized method executes this step automatically based on a single click of the input device. I quote. Mr. Sensi, a question that arose as I looked at this, if this were just an ordinary routine prosecution before the patent had issued and this issue had arisen, then saying that you can read the claim so broadly and so on, the whole broadest reasonable interpretation, the applicant has an opportunity to narrow the claim to respond to that objection. Now, as I saw in the record, and perhaps it was too vague to be clear, in this case the board authorized amendment that might have responded, might or might not, but that might have responded to this objection, and the applicant did not make such an amendment. Is that correct? I believe it is not correct, Your Honor. In fact, the additional claim limitation, the analyzing limitation, was added in the prosecution following an interview with the examiner, and the examiner invited just this amendment, which the applicant made. And we have a computerized process. We know it's computerized because the preamble says that it's a computer method. So the process that was added is analyzing the document to determine if the first information is contained therein. So the applicant added that. The applicant added that after the examiner, using BRI, said, applicant, we invite you to add this limitation. The applicant adds that limitation, explains why it's added, and the examiner, so that amendment was made on December 18, 2000. In January 2001, the examiner notes that the applicant did just what the examiner requested, and says that distinguishes over its own in the reasons for allowability. So, in fact, the amendment follows an invitation by the examiner, and the examiner, using BRI, determined that that limitation distinguishes over the so prior art. And the reason is the analyzing happens automatically. You have to understand that one of the things that's clever about this technology is that it mediates between two application programs. There's the application program in the preamble, in which the document is open, and there's text in that document. And there's another thing called a record retrieval program, which could be, for example, Outlook, where information is stored in records and information can be retrieved from it, contact information. So within the application program that is open, one can access information that's in the Outlook program and view it and insert it and do things like that that are reflected in the claim. So the interaction is very clever. So the functionality that actually was developed by my client was to be able, within an application program such as Word, to access address information and other kinds of information that pertain to, for example, a name you might have typed into the document. So you can insert the address into the Word document without having to flip into a new application. That was the invention. And so the invention discloses analyzing the text. This is in the description in the patent, analyzing the text in the document to figure out what's contact information. And what do you have to do? You simply push the one-click button that was developed by my client, and then the analyzing occurs. It's automatic. And this claim limitation to distinguish over in SO, there's some analyzing going on, but first you had to select that which is going to be the subject of the analyzing. Unlike SO, all you do is push this little button and it happens. So the limitation was added at the invitation of the examiner, and the examiner says now that you've added this, you've distinguished over at SO. And the reason you have distinguished over at SO is that you've added this limitation that allows the analyzing to occur automatically. In fact, in January 2nd, the examiner says, this is a quote from the examiner's reasons for allowance, however in SO the text string to be processed is determined by the current cursor position as specified by the user, and then there's a citation into SO, whereas the present invention, and this is a quote, an insight quote, does not require the user to select the text string to be processed since it functions automatically upon a single click of an input device. But this board was of a different view, and the whole structure of the America Invents Act is that three heads may be better than one, and so now we have what looked to me, and this is the clarification I was looking for, like an invitation and opportunity to amend. We know that there's a certain controversy about how many requests for amendments are granted and aren't, but that this was a situation in which it looks from the record as if the opportunity to amend was offered but was not accepted. And my question is, do I read that record correctly? I see that the board had done so. I think the board certainly said that the disclaimer was not evident to the board, that the board failed to see that, and they certainly said that, in their opinion, that the applicant has an opportunity to amend. Certainly Google said that. On the other hand, we think the disclaimer record is really clear because the examiner was quoting from the applicant. But even if you're correct in your disclaimer argument, on pages 38 and 39 of the board decision, I understand it analyzing the references and their disclosures, which is a fact question, and concluding, nonetheless, what's disclosed doesn't constitute selection, even under your construction. Am I missing something? That's one argument. Tell me why we should not only change the claim construction, which I think you've already done with your disclaimer argument, but in addition to overcoming that hurdle, you also need to overcome the hurdle that the board seems to have made an alternative fact-finding about the disclosure. The board did make that alternative fact-finding. They did or did not? They did. But I believe that finding is not supported by the evidence. The reason is the effect of the disclaimer was to require that the analyzing occurs automatically. If the analyzing occurs automatically, what is ruled out is not simply selection, but any form of assistance by the user once there's been an execute command entered. If there's no assistance by the user, then clearly having to put the information in the selection box of good hand is a kind of assistance, dramatic assistance that is not present in the RENDI solution, and the claims rule it out because it says it must happen automatically upon entry of the single click. So the board got to that position by ignoring the full impact of the disclaimer. So the disclaimer was achieved. Yes, it rules out selection, but it also rules out any assistance beyond entry of that single execute command. That's what really got ruled out. And in that sense, once you look at the full power of the disclaimer, and the examiner embraced it, the examiner adopted it, the board, by the way, says it does not need to pay attention to the examiner because it's just what the examiner said, but they failed to pay attention to the fact that the examiner was quoting from the applicant. The examiner on January 2001 was quoting from exactly what the applicant put in the remarks in December 2000. So the record was selectively regarded by the, may I have the water please? It was viewed in a very selective way by the board, and that's not proper. It has to look at the whole record. It can't pick the parts that it likes. In our opening brief, we pointed out in a passage what was said by the applicant and what was quoted by the board, and they missed all the stuff that was most germane. And similarly, they missed that the examiner had quoted from the applicant. So the record, when it's viewed in its entirety, it's very clear about the scope of this disclaimer, and the effect of that disclaimer is to rule out assistance beyond entry of that click. And I think when one really focuses on this ruling out of assistance, then the record is, okay, the board was selective in its consideration of the record. It didn't look at the whole record. I don't think there's any evidence to support the board's conclusion on the basis of the record taken in full. Okay. Thank you. Let's hear from the other side, and we'll save you rebuttal time. Thank you so much. Mr. Smith. Thank you, Judge Newman, Matthew Smith for the Appellee School and Motorola Mobility. Let me address your question, Judge Newman, then the alternative holding, and then get into the disclaimer argument. So I think there was some confusion about that question. The board did invite the appellant here, the patent owner, to file a motion to amend that's in the appendix at pages 1322 to 1323. Gave them, in fact, a special dispensation with the page limits to file amendments to all 79 claims, and they just didn't take advantage of that. So there was the opportunity to amend in this case, and the patent owners didn't do that. Regarding the alternative holding, which Judge Moore referred to, all of Arendi's claim construction arguments culminate in a proposed construction, which is found in Arendi's opening brief towards the end of the claim construction section at the bottom of page 50. And that proposed construction says, in relevant parts, I'm paraphrasing it, that the single entry and analyzing limitations of the claims must carry out an examining of the document that is inconsistent with user selection of the text. To be sure, the board rejected that as a matter of law, and I'll talk about that in a couple of minutes. But the board did not stop there. The board has an alternative holding, which was referred to in the colloquy with Mr. Sunstein, that occurs on pages 36 to 40 of the board's opinion, where the board takes Arendi's claim construction, assumes that that's the proper construction, and applies it to the prior art of record, good hand. And in so doing, the board made two, I think, key fact findings that are supported by substantial evidence, and that aren't disputed on a substantial evidence basis by Arendi. The first being that the good hand prior art performs analysis that is essentially the same as the 853 patent. The board uses that language four times in its opinion. And that's the analysis of identifying display names that will later be searched, distinguishing those display names from other material in the address field of the email document, including semicolons and spaces, and also properly formatted email addresses. That finding by the board, that set of findings by the board, was supported by the testimony of Mr. Allison, Google's expert, was supported by the deposition admissions of Dr. Levy, and supported by the specification of the good hand prior art itself. Arendi doesn't attack that by saying the evidence wasn't substantial, or that no reasonable fact finder could have relied on it. Arendi spends most of its argument simply re-arguing the claim construction. The second important finding by the board in that alternative holding was that good hand does not perform user selection of text. And that finding was supported by substantial evidence in the form of good hand's disclosure itself, and Mr. Allison's testimony about how good hand operated, as well as the examples of selection that are found in the 853 patent, at column 10, lines 5 to 10, and in the SO patent at column 4, lines 31 to 43, which describes selection in the way that users of graphical user interface systems are familiar with, highlighting, clicking on, selecting, italicizing, things like that. Good hand doesn't do those things. And the board, I think, properly found that selection was not going on in good hand. And on the basis of that alternative holding, I think you can affirm the board's decision without even reaching all of the claim construction arguments. In good hand, what is the triggering mechanism for the analysis that occurs? Do you initially trigger something with good hand and then enter the material in the address file? Or do you first enter the material in the address field and then move the cursor or do whatever you do, and that triggers it? Thank you, Judge Bryson. That goes directly to Mr. Sonnenstein's point that the analysis follows directly after the single entry of the execute command. In good hand, the analysis is triggered by clicking on the send button or clicking on the check names button, or in the alternative, moving the cursor from one field to another. And this is disclosed in the good hand patent at column 20, lines 18 to 21, I believe. And it's what the board found, too, based on this. So nothing occurs in good hand before the user types in whatever the user types in to the address field. Is that correct? When I say nothing occurs, nothing that is a triggering event for the operation of the good hand mechanism. Yes, that's correct, Your Honor. That's correct. The board, of course, we would submit, was also correct in its analysis of the claim construction, however. And let me focus on the prosecution history disclaimer argument, because that's the argument that Arendi makes most vehemently. There are three principal problems with that prosecution history disclaimer argument. And the first is that Arendi characterizes the prosecution history as sort of an agreement between the examiner and the applicant at that time. If that was an agreement between the applicant and the examiner, Arendi broke it, because they didn't do what the examiner was asking them to do. In the October 17, 2000 interview summary, the examiner says, we discussed the prior art with the applicant's representative. The applicant's representative suggests the so prior art requires selection. The present invention does not require selection. The applicant can submit an after final amendment to raise this distinction. This is appendix page 342. Now in response to that, I think a reasonable examiner would expect an amendment that negatively excludes user selection of text. But we have the experience, as Mr. Sunstein keeps stressing, of what a reasonable examiner did in 2001. And I've been just wondering, now that the board, now that there are three, three heads are better than one, but nonetheless, they are bound by their broadest reasonable construction. We accept that. Is it reasonable to adopt a position that was overcome, countered, and resolved 15 years ago? Your Honor, I think it is reasonable for the board to do what it did, because the file history is not clear why the examiner did what she did. And it's not clear for two reasons. It's not clear first because when Arendi entered its amendment, it did not enter an amendment that was relevant to selection. Facially, just based on the language of the amendment, it positively recited analyzing the claims. But it satisfied the examiner. We can say it's not clear, but it apparently was sufficiently clear at that time. That's correct, Your Honor. But the existence of the disclaimer and the scope of the disclaimer has to be decided under the clear and unambiguous standard. And of course, this court is deciding that de novo, but applying the same standard that the board would have applied, which is the clear and unambiguous standard. I think it is not clear, as you say, because the examiner's actions could be interpreted as accepting the amendment that Arendi gave them facially, positively reciting analyzing, and then going back to the prior art and checking whether the prior art teaches analyzing, not checking whether the positive teaching of selection causes the prior art to be excluded from the scope of the claim. That's correct, because the examiner said, and I'll read you the quote from his reasons for allowance, the present invention does not require the user to select the text string to be processed since it functions automatically upon a single click. So the examiner, in his reasons for allowance, clearly and unambiguously stated his belief that this is what the claims require, and he put that expressly in his reasons for allowance. So I don't understand your statement about how there isn't clear and unmistakable evidence in this record about the disclaimer. The examiner seems to have acknowledged the disclaimer expressly in his reasons for allowance, said the claims are being allowed for that reason. The statement that it's not clear and ambiguous, in fact, results in part from exactly that notice of allowance, because what the examiner did not say there is the present invention excludes selection. What the examiner actually said was the present invention does not require selection. Those are two very different things. In fact, if you go back up in that paragraph, this appendix page 349, just above there, the examiner describes the invention as requiring a single click, which triggers analyzing. So it seems to me that the examiner is characterizing the invention as positively reciting, analyzing, but not requiring selection. And that happens to also be consistent with the literal language of the claims, which does not require or does not say anything about selection at all, and the specification itself, as the board expressly found on pages 6 and 7, that seems to at least imply that selection can be a part of the invention or cannot be a part of the invention at column 10, lines 5 to 10. The third problem I think that this prosecution history argument has is the problem of scope. And that was raised, I think, by Mr. Sunnestein in his opening argument. What Mr. Sunnestein said was the prosecution history disclaimer disclaims all user actions that might be of assistance. And that is expressly not what the prosecution history says. In the parts that Mr. Sunnestein is quoting and Arendi is quoting in its briefing about the prosecution history, the prosecution history is only talking about selection. Once we try to get outside of that concept of selection and move to something else that's more broad, all user selection of text that could assist the analysis process, I would argue that even if there is a tenable disclaimer argument about the word selection, there is not a tenable disclaimer argument where the prosecution history doesn't contain any language. For example, all user actions that might somehow assist the analyzing. And there are a couple of reasons for that. First is we're operating under the clear and unambiguous standard, and as Judge Newman said, under the broadest reasonable interpretation of the claims. Both of those, I think, would counsel toward narrowing the scope of the disclaimer as much as possible and keeping the scope of the claims as reasonably broad as possible. The second problem with that argument is in the 853 patent itself, we know very little about the analysis that's actually happening. It's broadly claimed in the claims. In the specification, it is described only in this section in column four, principally at lines 29 to 37. And in that portion of the specification, the specification simply points out several features in the document that the analysis could take advantage of. For example, paragraph and line formatting, portions in the text that say Mr. or Mrs. or provide a zip code, from which the analysis could find a foothold and then infer that there is some information that it would want to use later in the process. But all of those things are things that the user types into the document. In other words, all of these features that the analysis can take advantage of are things that the user gives to the analysis system. So the user is, even in the 853 patent, in the only description of the analysis there, providing assistance to the analyzing step of the claims. And so if one tries to go beyond the selection idea in the scope of the disclaimer and broaden that out to all user actions, as Mr. Sunstein was suggesting, you eventually capture the essence of the analysis in the 853 patent. And that's what I think the board was driving at. What Goodhand does, in terms of textual analysis, is very difficult, conceptually, to separate from what the core analysis description is in the 853 patent. And if that disclaimer becomes too broad, if it goes beyond the idea of selection, you eventually capture that core of the 853 patent, and you shouldn't do that. As Judge Newman raised, the patent owner here did have the opportunity to file an amendment, a motion to amend, and there is some discussion about how plausible amendments are in inter-parties review. The Patent Office actually did a study on this in April. It's not a record, but you can certainly look it up. But the fact of the matter is that if Arendi's disclaimer is allowed to stand in this case, then the scope of the claims will not match the scope of the patent monopoly. I don't know why Arendi did not submit its amendment, but there are a couple of plausible reasons why patent owners might not want to do that. One is to avoid the effect of 35 U.S.C. 252, the effect on past damages, potential intervening rights. And the other, which I think is pertinent here, is to potentially do an end run around the scrutiny under the written description section of 35 U.S.C. 112, which might occur any time one submits an amendment. And I think it's not at all clear here that a negative limitation excluding selection would find written description support in a specification that expressly says selection can be part of the invention. Maybe it's questionable, maybe it's not. But the issue didn't come up because of the disclaimer argument, whereas it would have come up if Arendi had filed a motion to amend. You're talking when you refer to the express reference in the specification to the paragraph on column 10? That's exactly right. Lines 5 to 10. And we would submit that that is not a course of action that this court should encourage. Under the broadest reasonable interpretation of the claims, where the patent owner has the opportunity to amend the claims, the patent owner should be cleaning up this mismatch between the literal scope of the claims and the scope of the patent monopoly. And for that reason, we would suggest that under the broadest reasonable scope of the claims, a patent owner simply should not be allowed to assert even prior disclaimers if they have the opportunity to amend. But all of this argument is probably not even necessary, given the board's alternative holding, given the board's fact findings on how close the analysis in Goodhand is to the 853 patent, given the board's express fact finding based on substantial evidence that the Goodhand patent does not do user selection. I think you can affirm on that basis alone and should. Thank you very much. Okay. Thank you, Mr. Smith. Mr. Sunstein. Thank you, Your Honor. I want to first talk about the kind of analyzing that's said to go on in Goodhand, because assuming the board agrees with our views on disclaimer, we believe that given a proper claim construction, including the disclaimer, there's no evidence of analyzing without user selection in Goodhand. Goodhand necessarily requires the user to identify first information to the system by putting it in Goodhand's selection box. By identifying where the first information to be searched is, the selection box of Goodhand serves the same purpose in terms of the patent claim as does the selection of text, and so. The execute command asserted by Google is the cursors leaving the selection box or clicking on an enter command when the cursor is in the selection box. The execute command in the Goodhand system cannot trigger analyzing the document for contact information. The document is not analyzed to determine if there is any such first information. You say it's not analyzing the document. It's analyzing everything that's within the address field, right? Right, but the Goodhand system knows. It's not analyzing the text of an email, for example. Right. I can put the text, I can put whatever I want, including a recipe for a great breakfast, in the body of the email, in that template, and Goodhand sees nothing of that. It requires the user to put address information in that selection box. Without that, it analyzes nothing. And, of course, it is assumed what the user puts in there includes first information. So the execute command in Goodhand can't trigger analyzing the document for contact information. The document is not analyzed to determine if there is any such first information since that's what the user put into the selection box. And then the claim, there's a searching requirement in the claim as well. And that claim, searching for first information, takes place in Goodhand only with respect to information entered by the user into the selection box. Now, the board referred to other kinds of analyzing. One of them was, and it was cited by the petitioner, checking to see if there is text in the selection box. That's one of the theories for analyzing, just whether there's something put there. But that really, if you think about it, it means checking to determine if the user has placed text into the selection box. And that checking is not the same as the claim limitation of analyzing the document to determine if the first information is contained therein. If the user has typed nothing in the selection box, then there's nothing to analyze. It is the user who determines if first information is present by putting it into the selection box. And in fact, there's no evidence even. It's just a conjecture on the part of their expert that that's something that Goodhand does. And then if you look at figure one of our patent, the 853 patent, this limitation corresponds to the step in box four of figure one, analyze what the user has typed in the document. And per column four lines 22 to 27, the user has already typed text into the document, possibly including first information such as an address. As a result, analyzing the document to determine if first information is contained therein, as required by the claim, the URENDI system is analyzing what the user has typed, meaning that there must be text in order for the text to be subject to analyzing. And then in box six afterwards, which occurs in the logical flow after box four, therefore after the user has typed something into the document in referring to finding nothing, that doesn't mean that the document contains no text, but rather that the analyzing did not find any first information in the document. So the presence of text to analyze is a precondition for the claim limitation of analyzing the document to determine if the first information is contained therein. And so that doesn't constitute the analyzing that's required by the claim, just looking to see if text is there. And there are other kinds of analyzing that go on, parsing to find out what's an email address, what's a name, and what email addresses correspond to the name. But that's all after the user has said this is first information by sticking it in the selection box. So there's no analyzing in the record that supports the board's determination. We think if you properly construe the claim, then taking into account the surrender, and then looking at the claim limitation of analyzing which is required to occur automatically because it's part of this computer process, that means if it's automatic, you can't cheat by, okay, I'll stick it here. I'll put it in this selection box. It means that the analyzing has to occur automatically. When you click on the one button in the Aurendi technology, it figures out where the contact information is. It finds the first information, automatically does the searching. And the comment by my brother that the user is assisting because of the techniques used by the Aurendi system to find stuff, those are techniques used by the Aurendi system to find it. They're built into the software. What the user has done before that first information has been found is simply enter text, just as in box four of the figure. So we think there's a clear prosecution history disclaimer and that there's no evidentiary support for the board's conclusion that the good hand system meets the claim limitation. Okay. Thank you, Mr. Sensenbrenner. Thank you so much. Thank you. --- Source: Frix Law Library, https://www.frixlaw.com/law-library/documents/argument%3A26723. Public record. Not legal advice.