`
`1 matches because either you ' re using all the peaks or
`
`2
`
`]
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`1]
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`a subset of the peaks as indicated in Column 12
`
`somewhere it says - -
`
`le t me find it . Yeah--
`
`lines 6 and 7 , you have the option of unmarking
`
`peaks , which is disclosed in Iwamura .
`
`So as soon as you evaluate only a subset
`
`of the number o f locations , you get sublinear time
`
`search , because all it t akes is - - if the length of
`
`my string is , say , N, and the number of peaks or t he
`
`number of posit i ons tha t I ' m evaluating is sublinear
`
`in N,
`
`I get a sublinear search .
`
`Q
`
`You said if it ' s sublinear in N, but it ' s
`
`not , sir .
`
`A
`
`Q
`
`It is .
`
`Would you agree that as we increase the
`
`size of the database
`
`A
`
`Q
`
`Right .
`
`-- the dataset we ' re searching , that the
`
`amount of search time will be li near?
`
`A
`
`It ' s linear only in the size
`
`in the
`
`number of musical works . But , again , another
`
`dimension , as we have said , is the length of each
`
`"
`
`23 musical work .
`o But lengthening the work doesn ' t reduce
`the number of peaks .
`25
`NETWORK- J EXHIBIT 2006
`L.., ____________________ ~Google Inc . v. Network-I Technologies, Inc.
`JPR20 I 5·00345
`
`TSG Reporting WorldwidE
`Pa~ 289 of 384
`
`
`
`1 matches because either you ' re using all the peaks or
`
`Page 289
`
`2
`
`]
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`1]
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`a subset of the peaks as indicated in Column 12
`
`somewhere it says -- le t me find it . Yeah--
`
`lines 6 and 7 , you have the option of unmarking
`
`peaks , which is disclosed in Iwamura .
`
`So as soon as you evaluate only a subset
`
`of the number o f locations , you get sublinear time
`
`search , because all it t akes is - - if the length of
`
`my string is , say , N, and the number of peaks or the
`
`number of positions tha t I ' m evaluating is sublinear
`
`in N,
`
`I get a sublinear search .
`
`Q
`
`You said if it ' s sublinear in N, but it ' s
`
`not, sir .
`
`A
`
`It is .
`
`Q Would you agree that as we increase the
`
`size of the database
`
`A
`
`Q
`
`Right .
`
`-- the dataset we ' re searching , that the
`
`amount of search time will be linear?
`
`A
`
`It ' s linear only in the size
`
`in the
`
`number of musical works . But , again, another
`
`dimension , as we have said , is the length of each
`
`23 musical work.
`o But lengthening the work doesn ' t reduce
`the number of peaks .
`
`"
`
`25
`
`TSG Reporting - Worldwide - 877-702-9580
`Pa~ 289 of 384
`
`
`
`Page 290
`
`1
`
`2
`
`]
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`A Well , you can unmark them .
`
`So if I' m using 20 percent or if I ' m using
`
`5 percent . I ' m -- I ' m reducing my search speed
`
`accordingly.
`
`Let ' s take it one step at a time .
`
`Yes .
`
`Case 1 . We lengthen the number of musical
`
`Q
`
`A
`
`Q
`
`works .
`
`Would you agree that Iwamura is not
`
`sublinear in that sense?
`
`A
`
`We increase the number of linear -- yes , I
`
`agree .
`
`Q
`
`Now , we increase the size of the musical
`
`works . We don ' t unmark any peaks ; we just increase
`
`the size of the musical works .
`
`Would you agree that lwamura is not
`
`sublinear to inc r easing the size of the dataset
`
`then?
`
`A
`
`It is sublinear in the size of the
`
`dataset. All i t
`
`takes is to use a fraction of the
`
`data that is $ublinear , which is what everyone will
`
`do.
`
`Q
`
`Okay . Then I ' m talking about what Iwamura
`
`teaches .
`
`I ' m not talking about modifying it by
`
`taking a fraction of the data .
`
`TSG Reporting - Worldwide - 877-702-9580
`Pa~ 290 of 384
`
`
`
`Page 291
`
`1
`
`2
`
`]
`
`/I
`
`5
`
`6
`
`7
`
`8
`
`A
`
`This is what everyone does .
`
`So when you do string matching , you are --
`
`this technique is known as subsampling ; right?
`
`I t ' s
`
`very common . You -- you try to evaluate matches,
`
`and you only eva l uate a certain number of positions .
`
`If you have more and more data , you can
`
`get away with subsampling even more , meaning you
`
`look at an even smaller fraction of possible
`
`9 matches . That 's always h ow you get sublinear time .
`
`10
`
`11
`
`12
`
`Q Does Iwamura t each that as we increase the
`
`size of our dataset or the size of the song , that we
`
`are going to then change the number of samples that
`
`1] we're going to look at?
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`A
`
`It says it ' s an option that the user
`
`can -- can select .
`
`I mean , this is --
`
`Q Where?
`
`A Well , again ,
`
`i f you look at Column 12 --
`
`Q
`
`Okay . Does Column 12 say anything about
`
`if we increase the size of the dataset , we ' re going
`
`to then r educe t he number of pea ks that we look at?
`
`A
`
`The user defines -- it ' s very clear .
`
`Okay .
`
`It says you can select. So you
`
`select -- you unmark peaks; therefor e , you select a
`
`subset .
`
`'rhis is up to t he user .
`
`So all the user
`
`has to do -- of course , the user could choose not to
`
`TSG Reporting - Worldwide - 877 - 702-9580
`PaS'-' 291 ofJ84
`
`
`
`Page 292
`
`1
`
`2
`
`]
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`do that or the user could do that in a way that
`
`depends on the length o f the musical work .
`
`Q
`
`I didn ' t ask you what Pierre Moulin , as
`
`the user , could do
`
`A
`
`Q
`
`Right .
`
`wit h all of his knowledge , in 2015 ,
`
`sitting here in this deposition .
`
`I ' m asking you
`
`about what ' s taught here .
`
`A
`
`Q
`
`A
`
`Q
`
`Yeah .
`
`So le t me ask you a specific question .
`
`Right .
`
`Does this column -- first of all , you ' re
`
`pointing to Column 12 , lines 5 through 9 ; is that
`
`right?
`
`A
`
`Q
`
`Yes .
`
`Does that -- in Iwamura , Column 12 ,
`
`lines 5 through 9 -- state that the algorithms
`
`should be run as one opt ion by reducing the number
`
`of peaks if the size of the database increases?
`
`A
`
`It does not say what you just said .
`
`I t,
`
`however , discloses that you can select how many
`
`peaks you use for -- for matching . And it ' s not
`
`Pierre Moulin in 2015 who is saying this ; this was a
`
`technique that was used in the ' 80s already .
`
`It ' s a
`
`very old technique .
`
`TSG Re p orting - Worldwide - 877-702-9580
`PaS'-' 292 of 384
`
`
`
`Page 293
`
`1
`
`2
`
`]
`
`Q
`
`A
`
`Q
`
`And you ' re saying that right now . Okay?
`
`Yes .
`
`Did you point , in your Declaration , to any
`
`4 written work that discloses that technique? Yes or
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`no?
`
`A
`
`I don ' t
`
`remember if I -- if I did .
`
`Again , I want to supplement my opinion if
`
`I did not write it down .
`
`It ' s a well-known fact in
`
`the field of searching that you can use this kind of
`
`technique .
`
`It ' s very well known .
`
`Q
`
`By supplement your opinion you mean put
`
`something in a new Declaration that ' s not in this
`
`one?
`
`A
`
`No.
`
`It ' s just complementing -- just
`
`complementing , giving more details about what I have
`
`16 written . The fact that peaks can be subsampled is
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`"
`
`25
`
`not a new opinion .
`
`It is there already .
`
`I ' m
`
`explaining--
`
`Q Well , if it ' s there already, then open up
`
`your Declaration and point to the portion where you
`
`cite to any prior art that talks about decreasing
`
`the number of samples we ' re going to use as our
`
`dataset increases .
`
`MR . ELA~UUA : Objection .
`
`THE WITNESS :
`
`I don ' t recall I did that . As I
`
`TSG Reporting - Worldwide - 877-702-9580
`PaS'-' 293 ofJ84
`
`
`
`Page 294
`
`1
`
`2
`
`]
`
`•
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`said , this addit ional e xplanation is supplementing
`
`my written opinion in the Declaration.
`
`BY MR. DOVEL :
`
`Q Well , I' ll get a chance to take your
`
`deposition when I see your supplemental Declaration .
`
`That ' s a separa t e deposition .
`
`I ' m talking about
`
`this one .
`
`A
`
`Again , let me clarify .
`
`We're talking about sublinearity . The
`
`Board made a construction of "sublinearity" which is
`
`somewhat different from the definition I had used in
`
`my Declaration .
`
`Q
`
`Somewhat different? It ' s not materially
`
`different though , is it .
`
`A Well , it is somewhat different . 50--
`
`Q
`
`Is it materially different? Does it
`
`say anything
`
`does it mean anything different?
`
`MR . ELACQUA : Objection .
`
`THE WITNESS : That ' s your interpretation --
`
`BY MR. DOVEL :
`
`Q
`
`A
`
`Yours .
`
`for me, any time I see something that
`
`is different from , you know , my assumption , my
`
`construction , 1 reevaluate everything --
`
`Q
`
`Now .
`
`TSG Reporting - Worldwide - 877-702-9580
`PaS'-' 294 ofJ84
`
`
`
`Page 295
`
`1
`
`2
`
`]
`
`4
`
`5
`
`6
`
`7
`
`8
`
`A
`
`Q
`
`-- I ' m being careful.
`
`I need an answer to this question .
`
`Is the Board ' s definition -- definition o f
`
`" sublinear time search " -- does it mean something
`
`different than t he definition -- definition that you
`
`set forth in your Decla rat ions?
`
`A
`
`Q
`
`No . They ' re essentially the same .
`
`Now , let ' s go back to your - - your
`
`9 Declarations .
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`211
`
`Anywhere in your declarations do you
`
`identify any prior art that disclosed reducing the
`
`amount of sampl ing when our database -- our dataset
`
`increases?
`
`A
`
`I don ' t recall I did .
`
`It ' s a well-known
`
`fact. Again , in light o f the Board ' s const r uction ,
`
`it made me thin k of additional , s upplemental way t o
`
`e xplain this .
`
`I t ' s a well-known fact .
`
`I t ' s not a
`
`revelation .
`
`Q
`
`Does
`
`in your Declaration , did you point
`
`to any part o f Iwamura t hat teaches reducing the
`
`amount of samples or reducing the number of peaks as
`
`our dataset size increase s?
`
`A
`
`I believe I referred to that passage o f
`
`lwamura .
`
`I could chec k where in my Declaration i t
`
`25 might be . Okay?
`
`It ' s an option , unmarking peaks .
`
`TSG Re porting - Worldwide - 877 - 702-9580
`PaS'-' 29S of 384
`
`
`
`Page 296
`
`1
`
`2
`
`]
`
`•
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`Q Well , in Iwamu r a .
`
`I thought we just talked
`
`about that . This Section 12 , li ne s 5 through 9 , is
`
`that what you ' re ta lking about?
`
`A
`
`Q
`
`Ye s .
`
`Yes .
`
`Column 12. lines 5 through 9 , does it
`
`disclose reducing the number of peaks that we ' re
`
`going to search as the size o f the database
`
`increases?
`
`A
`
`It ma kes it clear to t he user that they
`
`can select the f r action of peaks that they unmar k.
`
`It -~ it ' s very clear .
`
`Q
`
`I didn ' t as k whether it says you can
`
`select the frac t ion .
`
`A
`
`Q
`
`Right.
`
`Withdrawn .
`
`You would agree , sir , that what Iwamura
`
`teaches is that you could sel ect the fract i on o f the
`
`peaks -- of the peaks that you -- withdrawn .
`
`You would agree that Iwarnura says that we
`
`can disrega r d a r epeated section of music . That ' s
`
`what it says . That's one example ; right?
`
`A
`
`Q
`
`Yes .
`
`I t also says t he user setting up the
`
`database can choose to disregard the unimportant
`
`sections of musi c ; right ?
`
`TSG Reporting - Worldwide - 877 - 702-9580
`PaS'-' 296 of 384
`
`
`
`Page 297
`
`1
`
`2
`
`]
`
`•
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`A
`
`Q
`
`Yes .
`
`Does Iwamura give any other reason for
`
`disregarding a portion of the music?
`
`A
`
`Yes .
`
`I n the interest of accelerating the
`
`search -- now this is li ne 9, right -- to accelerate
`
`the search , you can unmark peaks . 50 - -
`
`Q Why -- why does it say to unmark peaks?
`
`It says to avoid searching unnecessary
`
`portions ; right?
`
`A Well , it says by unmarking peaks , you can
`
`certainly . yes , select portions that shouldn ' t be
`
`searched . And then, in addition , this , as you said ,
`
`avoids searching unnecessary portions but also
`
`accelerates search speed .
`
`So any practitioner seeing this is going
`
`to say , " Well ,
`
`I can choose my -- my fraction of
`
`peaks that I want to work with . And there ' s a
`
`tradeoff .
`
`If I made that fr3ction sm3ll , I
`
`accelerate my search , but my matching is not going
`
`to be as good ."
`
`50 the practitioner , if he ' s faced with
`
`22 musical works that are twice as long , is going to
`
`23
`
`24
`
`experiment with that parameter . You will conclude
`
`the number of peaks should not double .
`
`It should be
`
`25 multiplied by , say , 1 . 5 , and you obtain ,
`
`then ,
`
`T5G Reporting - Worldwide - 877-702-9580
`PaS'-' 297 of 384
`
`
`
`Page 298
`
`1
`
`2
`
`]
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`similarity .
`
`Q
`
`Does Iwamura teach that , or is that
`
`something you ' re saying t hat one of ordinary skill
`
`in the art would know to a dd?
`
`A
`
`One o f o r dinary ski ll would abso l utely
`
`understand that the choice of this fraction is a
`
`tradeoff between search speed and matching
`
`performance .
`
`Q
`
`Does I wamura teach reducing the number of
`
`peaks that you check based upon an inc r ease in the
`
`size of the musical wor k in the database?
`
`MR. £LACQUA : Objection .
`
`THE WITNESS : As I said , he says the user can
`
`select that -- t hat fraction .
`
`BY MR . DOVEL :
`
`Q
`
`The -- he says the user can select the --
`
`can unmar k pea ks; right?
`
`A
`
`Q
`
`That ' s right . Yes .
`
`Does he say that the user should do it
`
`based upon the size o f
`
`t he dataset that he ' s working
`
`21 with?
`
`22
`
`A
`
`He does not say it because it ' s a
`
`23 well-understood fact in -- in that field .
`
`"
`
`25
`
`Q Would you agree that lwamura does not
`
`e xpressly teach reducing the number of peaks that
`
`TSG Re porting - Worldwide - 877 - 702-9580
`Pa~ 298 of 384
`
`
`
`Page 299
`
`1
`
`2
`
`]
`
`4
`
`are checked based upon the size of the dataset?
`
`A
`
`He does not e x plicitly say that .
`
`It ' s
`
`simply a known fact in -- in this field that if the
`
`length of your string increases , this is a really
`
`5 well-known technique to trade off speeds against
`
`6 matching performance .
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`Q
`
`A
`
`Q
`
`You ' re not answering my question .
`
`Yes , I am .
`
`I said he is not saying this .
`
`Okay . Then that would be the answer . But
`
`then you went on and added a bunch of other stuf f
`
`that ' s not responsive .
`
`MR . £LACQUA : Objection .
`
`BY MR . DOVEL :
`
`Q
`
`I think it ' s important for you to address
`
`my question .
`
`I f you ' ve got other things that you
`
`16 want to say , you will have plenty of time to do
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`that.
`
`A
`
`Q
`
`answered .
`
`A
`
`Q
`
`Okay --
`
`I promise you .
`
`I need to get my questions
`
`Okay .
`
`So please restate it .
`
`Do you a gree that Iwamura does not
`
`e xpressly teach reducing the number of peaks that
`
`are searched based upon the size of the dataset
`
`that ' s being sea r ched?
`
`TSG Re p orting - Worldwide - 877 - 702-9580
`PaS'-' 299 of 384
`
`
`
`1
`
`2
`
`]
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`MR . ELACQUA : Object ion .
`
`THE WITNESS :
`
`I agree in this paragraph it
`
`does not say anything about the size of t he musica l
`
`Page ]00
`
`work .
`
`BY MR . DOVEL :
`
`Q
`
`A
`
`That doesn 't answer my question .
`
`I said I agree .
`
`Q Well , you said you agree that it doesn ' t
`
`teach , and then you answered a different question .
`
`So answe r my question --
`
`A Well , you quot ed the paragraph -- you
`
`quoted the paragraph and you asked if he teaches
`
`that .
`
`So I -- I said I agree with what you said .
`
`In that paragraph , it does not t e ach that .
`
`Q
`
`Sir , do you agree that Iwamura does not
`
`teach altering the number of pea ks that are searched
`
`based upon the size of the dataset?
`
`It does not e xp licitly say it .
`
`I agree .
`
`Do you agree that it ' s not inherent in
`
`A
`
`Q
`
`Iwamura?
`
`A
`
`To me , it is inherent , as in that field ,
`
`everyone unders t ands that this is a tradeoff between
`
`search speed and matching perfo r mance .
`
`It is
`
`inherent . When you deal with a large database ,
`
`it ' s -- it ' s inherent .
`
`TSG Reporting - Worldwide - 877 - 702-9580
`PaS'-' 100 of 184
`
`
`
`Page 301
`
`1
`
`2
`
`]
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`"
`
`25
`
`Q
`
`DO you know what the word ~ inherent " means
`
`in the context of patents?
`
`A Well , I ' m not a -- an attorney . Okay? So
`
`I -- my understanding of inherent is that it is
`
`implied .
`
`So you ' re making some assumptions here t he
`
`size of the musical wor k would increase .
`
`Q
`
`Let me give you a definition of i nherent
`
`I ' d li ke you to apply .
`
`A
`
`Q
`
`Okay .
`
`I want you to assume inherent means tha t
`
`something is unstated in a reference , but it ' s
`
`necessarily present.
`
`It ' s the only way it could be
`
`done . There ' s no other possibility .
`
`A
`
`Okay . By
`
`MR . ELACQUA : Objection .
`
`Wait for the question .
`
`BY MR. DOVEL :
`
`Q
`
`A
`
`Q
`
`You understand that definition?
`
`Yes .
`
`Does
`
`is i t
`
`inherent in what Iwamura
`
`teaches that one would reduce the -- the number of
`
`peaks that are searched b ased upon database size?
`
`MR . ELACQUA : Objection .
`
`'l'HE WITNESS : There ' s no other reasonable way
`
`to do it . There ' s always a way to do it in a way
`
`TSG Reporting - Worldwide - 877-702-9580
`PaS'-' 10 I o D84
`
`
`
`Page 302
`
`1
`
`2
`
`3
`
`•
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`that does not reduce number of pe aks .
`
`It ' s always
`
`an option . He mentions that option .
`
`BY MR. DOVEL :
`
`Q
`
`So you would agree that it ' s not
`
`necessarily the case ; it ' s not inhe r ent in Iwamura
`
`that the only way to do it would be to reduce the
`
`number of peaks based u pon database size ; cor r ect?
`
`A
`
`It would be a bad way , okay , no reasonable
`
`person would do that .
`
`Q Correct . Yes or no?
`
`MR . ELACQUA : Object ion .
`
`THE WITNESS : My u nderstanding is yes , it
`
`would be bad engineering .
`
`BY MR . DOVEL :
`
`Q
`
`Okay . You ' re not responding to my
`
`question .
`
`I didn ' t ask you whether it was a bad way ;
`
`I didn ' t ask you whether it was a way that one o f
`
`in your field would consider to be a way that you
`
`shouldn ' t do it. It ' s about inherency and
`
`necessary .
`
`A
`
`Q
`
`All right.
`
`Do you agree , si r, that it ' s not inheren t
`
`in what lwamura teaches t o do a search in which the
`
`peaks are reduced based upon increasing the size of
`
`TSG Re porting - Worldwide - 877 - 702-9580
`Pa~ 102 of 184
`
`
`
`Page 303
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`"
`
`25
`
`the database?
`
`A
`
`I think it is necessary to obtain good
`
`performance with a reasonable search speed .
`
`I think
`
`it's necessary . That ' s my opinion .
`
`Q
`
`I didn ' t ask you whether it ' s necessary to
`
`obtain good speed .
`
`Is it the case that the only way you could
`
`perform the Iwamura search is by reducing the number
`
`of peaks that are searched based upon database size?
`
`A
`
`Q
`
`The -- say it again.
`
`Is it the case that it ' s necessary and
`
`only
`
`withdrawn .
`
`IS it - - with respect to the Iwamura peak
`
`search , is it necessarily the case that the only way
`
`it could be done was by reducing the number of peaks
`
`that are searched when we increase the size of the
`
`database?
`
`A
`
`So that would be the only way to obtain
`
`sublinearity, okay , by using this technique of
`
`unmarking peaks . There ' s a variety of techniques
`
`Q
`
`I didn ' t ask about the only way to obtain
`
`sublinearity . You are now consciously avoiding my
`
`question .
`
`I need an answer --
`
`A
`
`Q
`
`No .
`
`~lease restate . Okay?
`
`Is it the case that in Iwamura ,
`
`i t would
`
`T5G Reporting - Worldwide - 877-702-9580
`PaS'-' lQ3 ofJ84
`
`
`
`Page 304
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`be inheren t and necessa r ily the case that t he only
`
`way to do the search in Iwamura is by reducing the
`
`number of peaks as we inc r ease the size of the
`
`database?
`
`A
`
`Q
`
`No ,
`
`i t' s not i nherent .
`
`Would you agree , sir , that if it ' s not
`
`e xpress and it ' s not inherent, that I wamura then
`
`does not teach increasing the number of peaks based
`
`upon the size of the database?
`
`A
`
`It -- it teaches it by stating that the
`
`user has the opt ion of selecting a fraction of the
`
`peaks which is understood to mean it depends -- the
`
`way you do it depends on t he parameters .
`
`So it is taught in my view , in my opinion .
`
`Q
`
`Now , let ' s ta ke a
`
`l ook -- in your
`
`In your declarations , the four
`
`declarations , you , at the beginning o f your
`
`18 Declaration , say that , " I understand that subject
`
`19 matter can be anticipated if each limitation is
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`f ound e xpressly or inherently in a single prior ar t
`
`reference ." You make that general comment .
`
`In your analysis ,
`
`I did not see any place
`
`where you e xpress the conclusion that a limitation
`
`or an element was inherent in any of the prior art
`
`references .
`
`T5G Reporting - Worldwide - 877 - 702-9580
`PaS'-' 104 o fJ84
`
`
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`1
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`Do you recall any place in your
`
`declarations
`
`A
`
`Let me -- the language -- so the part that
`
`you quoted from in my Declaration , which page is it?
`
`Page 305
`
`Q
`
`A
`
`Q
`
`A
`
`Paragraph 26.
`
`Okay.
`
`Page 10.
`
`Yeah .
`
`Okay .
`
`I ' ve read it .
`
`Q
`
`Now , as I look through your Declaration , I
`
`notice that you identified -- you made various
`
`statements about the various elements that were
`
`taught by lwamura , taught by Ghias and so on .
`
`A
`
`Q
`
`Yes.
`
`And then you identified portions of Ghias
`
`or Iwamura that you contended disclosed those
`
`portions?
`
`A
`
`Q
`
`Yes.
`
`I did not see any opinions where you said,
`
`"This element is not expressly taught ; however, it
`
`is inherent , and here ' s why . n
`
`Do you recall expressing any opinions to
`
`the effect that a element was not expressly taught
`
`but was instead inherent?
`
`A
`
`I don ' t recall making that statement , nO .
`
`TSG Reporting - Worldwide - 877-702-9580
`P3ge )05 of 384
`
`
`
`MR . DOVEL : Let ' s go ahead and take a break
`
`Page 306
`
`for the evening .
`
`MR . ELACQUA : Sure .
`
`THE VIDEOGRAPHER : This will conclude today ' s
`
`proceedings in t he deposition of Pierre Moulin . The
`
`total number of videotapes used today was four . And
`
`we ' re off the record at 5 : 39 PM .
`
`(The deposition was concluded at 5 : 39 PM)
`
`1
`
`2
`
`]
`
`•
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`TSG Reporting - Worldwide - 877 - 702-9580
`Pag<.' 106 of 384
`
`
`
`1
`
`2
`
`DECLARATION
`
`Page 307
`
`3
`
`•
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`I hereby declare that I am the deponent
`
`in the within matter; that I have read the foregoing
`
`proceedings and know the contents thereof , and I
`
`declare that the same is true of my knowledge except
`
`as to the matters which are therein stated upon my
`
`information or belief , and as to those matters , I
`
`believe it to be true .
`
`I declare , under the penalties ot
`
`perjury of the state of California , that the
`
`foregoing is true and correct.
`
`Executed on the
`
`day of
`
`at
`
`17 California .
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`PIERRE MOULIN , PhD
`
`TSG Reporting - Worldwide - 877-702-9580
`PaS'-' 107 of 384
`
`
`
`STATE OF CALIFORNIA
`
`COUNTY OF LOS ANGELES
`
`Page 308
`
`I , Rich A10ssi , California Certified
`
`Shorthand Reporter Number 13497 , do hereby certi f y :
`
`That prior to being examined , the witness
`
`named in the foregoing proceedings was duly sworn ;
`
`That said proceedings were taken before me
`
`at the time and place therein set forth and were
`
`taken by me in stenographic shorthand and t hereaft er
`
`transcribed into typewritten form under my direction
`
`and supervision ;
`
`That the dismantling of this transcript
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14 will void the Reporter ' s certificate ;
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`Before completion of the deposition , review
`
`of the transcri pt was (XX) was not [
`
`requested .
`
`If requested , any changes made by the
`
`deponent and provided t o the Reporter during the
`
`period allowed are appended hereto .
`
`IN WITNESS WHER80F ,
`
`I hereunto subscribe my
`
`name this 31st day of August. 2015 .
`
`RI CH ALOSSI , RPR. CCRR , CSR No . 13497
`
`TSG Reporting - Worldwide - 877-702-9580
`PaS'-' l OS o f l 84
`
`
`
`1
`
`2
`
`3
`
`•
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`Page 30'
`
`NAME OF CASE :
`
`DATE OF DEPOSITION :
`
`NAME OF WITNESS :
`
`Reason Codes :
`
`l. To clarify the record .
`
`2 . To con f orm to the facts.
`
`3 . To correct transcription errors .
`
`Page
`
`From
`
`Page
`
`From
`
`Page
`
`From
`
`Page
`
`From
`
`Page
`
`From
`
`Page
`
`From
`
`P.'IOg!"
`
`From
`
`Page
`
`From
`
`Line
`
`Reason
`
`to
`
`Line
`
`Reason
`
`Line
`
`Line
`
`to
`
`Reason
`
`to
`
`Reason
`
`co
`
`Line
`
`Reason
`
`Line
`
`Lin!"
`
`to
`
`Reason
`
`to
`
`Rf'!.'IO!!on
`
`to
`
`Line
`
`Reason
`
`to
`
`TSG Reporting - Worldwide - 877 - 702-9580
`PaS'-' 109 of 184
`
`
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`15
`
`16
`
`17
`
`IS
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`UNITED STATES PATENT AND TRADEMARK O~~ICE
`
`BE~ORE THE PATENT TRI AL AND APPEAL BOARD
`
`Page 310
`
`GOGGLE , INC ., and YOUTUBE , LLC ,
`
`Peti tioner ,
`
`"' .
`NETWORK- 1 TECHNOLOGIES ,
`
`INC .,
`
`Pa t ent. Owner .
`
`Case No . IPR20 15- 00347
`
`VIDEOTAPED DE POSITION O ~ PI ERRE MOULIN , PhD, VOLUME II
`
`Sa nt.a Monica , Californi a
`
`Thursday, August 20 , 2015
`
`REPORTED BY: RI CH ALOSS I, RPR, CCRR , CSR NO. 13497
`
`Job No : 96810
`
`TSG Repor t ing - worldwide - 877 - 702 - 9580
`PaS'-' 110 ofJ84
`
`
`
`UNITED STATES PATENT AND TRADEMARK O~~ICE
`
`BE~ORE THE PATENT TRIAL AND APPEAL BOARD
`
`Page 311
`
`GOGGLE ,
`
`INC ., and YOUTUBE , LLC ,
`
`Petitioner ,
`
`"' .
`
`NETWORK-1 TECHNOLOGIES ,
`
`INC .,
`
`Pa t ent. Owner .
`
`Case No . IPR2015 - 00347
`
`VIDEOTAPED DEPOSITION O~ PIERRE MOULIN , PhD,
`
`VOLUME II ,
`
`t.aken on behalf of the Patent Owner , a t 201 Santa
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`15
`
`16
`
`17
`
`18 Monica Bouleva r d , Si x th ~loor , Santa Monica , Cali f ornia , on
`
`Thursday , Augus t 20 , 20 15 ,
`
`from 9 ; 05 AM to 11 ; 14 AM, before
`
`RIcn ALO!>!>I , RPR , CCRR, CGR NO . 13497 .
`
`~ * *
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`TSG Re p orting - worldwide - 877 - 702-9580
`PaS'-' 11 1 ofJ84
`
`
`
`Page 312
`
`APPEARANCES :
`
`For the Plaintiff :
`
`SKADDEN ARPS SLATE MEAGHER & FLOM
`BY :
`JAMES ELACQUA , Attorney at Law
`IAN CHEN , Attorney at Law
`525 University Avenue
`Palo Alto , CA 94301
`
`For the Patent Owner Network- 1 Technologies :
`
`, LUNER
`DOVEL
`BY : GREGORY DOVEL, Attorney at Law
`201 Santa Monica Boulevard
`Santa Monica , CA 90401
`
`Also Present ;
`
`SCOTT MCNAIR , Videographer
`RICH SONNENTAG , Litigation Counsel , Google , Inc .
`
`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
`
`TSG Reporting - Worldwide - 877-702-9580
`PaS'-' 112 oD84
`
`
`
`Page 31 3
`
`PAGE
`
`313
`
`376
`
`1
`
`2 WITNESS
`
`I N D E X
`
`PIERRE MOULIN , PhD, VOLUME I I
`
`BY MR . DOVEL
`
`BY MR . ELACQUA
`
`E X HIBITS
`
`(None . )
`
`]
`
`•
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`TSG Repor t ing - Worldwide - 877 - 702 - 9580
`PaS'-' 113 ofJ84
`
`
`
`SANTA MONICA , CALIFORNIA; THURSDAY , AUGUST 20 , 2015
`
`9 : 05 AM - 11 ; 14 AM
`
`Page 314
`
`THE VIDEOGRAPHER : Good morning . We are bacK
`
`on the record for Day 2 o f the continuing deposition
`
`of Pierre Moulin . Today ' s date is August 20th ,
`
`2015 . The time i s 9 : 05 AM . And the witness has
`
`already been sworn .
`
`-
`
`-
`
`-
`
`PIERRE MOULIN , PhD ,
`
`having been previously duly sworn by
`
`the court reporter , was examined
`
`and testi f ied as follows :
`
`EXAMINATION
`
`MR . DOVEL : Can I have the e xhibits . Thanks .
`
`THE WITNESS : ThanK you .
`
`MR . DOVEL :
`
`I ' ve placed in front of t he
`
`1
`
`2
`
`]
`
`•
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19 witness Exhibit 1012 ,
`
`the Iwamura prior art
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`reference .
`
`BY MR . DOVEL :
`
`Q Why don ' t you turn to Column 9 .
`
`At lines 44 to 45 , does Iwamu r a teach that
`
`a peak that is -- is in an unimportant section can
`
`be skipped?
`
`TSG Reporting - Worldwide - 877 - 702-9580
`Pag<.' 11 4 oD84
`
`
`
`Page ]15
`
`1
`
`2
`
`]
`
`A
`
`Q
`
`Yes .
`
`If you ' ll look a little further down, does
`
`Iwamura teach that certain portions of the song are
`
`4 well recognized and remembered by the user?
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`"
`
`25
`
`A
`
`Q
`
`A
`
`Q
`
`Which lines would that be?
`
`46 -- or 47 and 48 .
`
`These portions , yes, I see that .
`
`Does Iwamura then teach right after that
`
`that the user identifies such important portions as
`
`a keyword or key melody .
`
`Do you see that?
`
`Yes .
`
`And by " keyword or key melody , " that ' s
`
`A
`
`Q
`
`going to be the melody that we ' re using as our - - as
`
`our query in the Iwamura search?
`
`A
`
`It ' s certainly part of the query .
`
`Q Well , the query is going to be based upon
`
`the -- what ' s entered by the user ; is that right?
`
`A
`
`Q
`
`Yes .
`
`And so when you say " part ," is there some
`
`other part of the query that ' s not entered by the
`
`user?
`
`A
`
`I ' m just reading this again .
`
`It appears that that is what will be input
`
`by the user , yes .
`
`TSG Reporting - Worldwide - 877-702-9580
`PaS'-' 11 S oD84
`
`
`
`Page ]16
`
`1
`
`2
`
`]
`
`4
`
`5
`
`6
`
`7
`
`Q
`
`A
`
`When you say " t hat. ~ what do you mean?
`
`When you asked me if those -- keywords is
`
`what will be input by t he user ; correct? And I
`
`agree .
`
`Q
`
`And what the -- withdrawn .
`
`Does Iwamura t each here that the user
`
`identifies the important parts as the keyword or key
`
`8 melody that is used as t he query?
`
`9
`
`10
`
`A
`
`Q
`
`Yes .
`
`Is it the case that if the keyword or key
`
`11 melody consists of important parts . and unimportant
`
`12
`
`1]
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`parts are omitted from the reference database , that
`
`the Iwamura search would not exclude a -- a
`
`potential reference as a match based upon a failure
`
`to search the unimportant parts?
`
`A Well , this sentence says , "The user " -- so
`
`we ' re talking about the query here .
`
`" The user
`
`identifies important portions ." That sentence says
`
`nothing about the database itself .
`
`Q
`
`A
`
`Q
`
`I understand t hat .
`
`Okay .
`
`I want you to assume that we ' ve got a
`
`database where we ' re going to use this method that ' s
`
`identified in Column 9, line 44 , that a peak that is
`
`in an unimportant section can be skipped .
`
`TSG Reporting - Worldwide - 877-702-9580
`PaS'-' 116 o D84
`
`
`
`Page 317
`
`1
`
`2
`
`3
`
`4
`
`A
`
`Q
`
`Okay .
`
`All right .
`
`So let ' s assume we ' ve got our
`
`dat abase up . We ' ve ident ified the unimport ant
`
`sections , and we ' re no t going to assess those when
`
`5 we're doing our Iwamura search .
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`19
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`Does t hat make sense?
`
`Yes .
`
`In that case , failing to test a melody
`
`A
`
`Q
`
`against an unimportant part will not result in us
`
`ignoring a match . Would you agree?
`
`A
`
`That is correct , assuming it ' s truly an
`
`unimportant part , yes .
`
`Q Would you agree -- withdrawn .
`
`Is it the case that if we do the Iwamura
`
`search using the peaks as our basis , and we set up
`
`our database such that the unimportant peaks are
`
`skipped , that we ' re still going to be identifying
`
`the closest match when we produce our resu lts?
`
`A
`
`That would be assuming that no peaks have
`
`been dropped and everything we discussed yesterday .
`
`Dropping an unimportant part is not going to affect
`
`the ability to find the best matCh .
`
`Q
`
`A
`
`Why is that?
`
`We ll , because as we assume t hese are
`
`unimportant portions , and so we do not need to
`
`TSG Reporting - Worldwide - 977-702-9580
`Page 3 17 o f 384
`
`
`
`Page 318
`
`consider them in order to find the best match .
`o Let ' s take a look just above that .
`There ' s another teature of Iwamura that
`
`says , in one variation, melodies have repeated
`
`patterns, and we can avoid having to search a
`
`repeated pattern more than once .
`
`Do you see that?
`
`This would be line 36 , 31?
`
`A
`o
`A
`o
`is it the case that by skipping a repeated pattern ,
`
`Yes .
`
`Yes .
`
`If we implement tha t feature of Iwamura ,
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`1
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`it ' s not going to stop us from producing the best
`
`14 match?
`
`15
`
`16
`