`
`
`
`
`
`
` Exhibit 15
`
`
`
`
`
`Case 1:14-cv-02396-PGG-SN Document 234-17 Filed 11/11/20 Page 2 of 69
`
`Paper No. 1
`Date Filed: December 3, 2014
`
`
`Filed on behalf of: Google Inc.
`
`By:
`James J. Elacqua
`james.elacqua@skadden.com
`(650) 470-4510
`
`
`
`UNITED STATES PATENT AND TRADEMARK OFFICE
`________________
`
`BEFORE THE PATENT TRIAL AND APPEAL BOARD
`________________
`
`Google Inc.
`Petitioner,
`v.
`Network-1 Technologies, Inc.
`Patent Owner.
`________________
`
`Case IPR2015-00347
`U.S. Patent 8,010,988
`________________
`
`
`
`PETITION FOR INTER PARTES REVIEW OF
`U.S. PATENT NO. 8,010,988 UNDER 35 U.S.C. §§ 311-319
`
`
`
`
`
`
`
`Case 1:14-cv-02396-PGG-SN Document 234-17 Filed 11/11/20 Page 3 of 69
`
`TABLE OF CONTENTS
`
`Page
`I. MANDATORY NOTICES UNDER 37 C.F.R. § 42.8(a)(1) ........................ 1
`II.
`PAYMENT OF FEES UNDER 37 C.F.R. § 42.103 .................................... 1
`III. GROUNDS FOR STANDING UNDER 37 C.F.R. § 42.104(a) ................... 2
`IV. SUMMARY OF CHALLENGE AND RELIEF REQUESTED ................... 2
`V.
`SUMMARY OF THE '988 PATENT ......................................................... 3
`A. Algorithms for Automatic Content Recognition ................................ 3
`B.
`Summary of the '988 Claims ............................................................ 3
`C.
`Person of Ordinary Skill in the Art of the '988 Patent ........................ 5
`D.
`Claim Construction Under 37 C.F.R. § 42.104(b)(3) ......................... 5
`VI. ANALYSIS OF GROUNDS FOR TRIAL.................................................. 8
`A. Ground 1: Claims 1-3, 7-9, 13-17, 21-23, 28, 30-31, 40, 49, and
`51 of the '988 Patent Are Anticipated By Ghias Under U.S.C.
`§ 102(b) .......................................................................................... 8
`1.
`Summary of Ghias ................................................................. 8
`2.
`Ghias Discloses Every Element of Claims 1-3, 7-9, 13-
`17, 21-23, 28, 30-31, 40, 49, and 51 of the '988 Patent ............. 9
`B. Ground 2: Claims 1-2, 7-9, 13-16, 21-23, 27-33, 38, 40-42, 47,
`49, and 51 Are Anticipated By Wood Under U.S.C. § 102(e)........... 16
`1.
`Summary of Wood ............................................................... 16
`2. Wood Discloses Every Element of Claims 1-2, 7-9, 13-
`16, 21-23, 27-33, 38, 40-42, 47, 49, and 51 ........................... 16
`C. Ground 3: Claims 1-3, 7-17, 21-34, 37-38, 40-43, 46-47, and
`49-52 Patent Are Obvious Under 35 U.S.C. § 103 Over Levy in
`view of Arya ................................................................................. 23
`1.
`Summary of Levy ................................................................ 24
`2.
`Summary of Arya................................................................. 25
`3.
`The Combination of Levy and Arya Discloses Every
`Limitation of Claims 1-3, 7-17, 21-34, 37-38, 40-43, 46-
`47, and 49-52 ....................................................................... 25
`A Person Having Ordinary Skill in the Art Would Have
`Been Motivated to Combine Levy with Arya......................... 27
`D. Ground 4: Claims 15-17, 21-22, 24-25, 27-28, 31-32, 37, 40-41,
`and 46 of the '988 Patent Are Obvious Under 35 U.S.C. § 103
`Over Iggulden in view of Böhm ..................................................... 36
`1.
`Summary of Iggulden ........................................................... 36
`2.
`Summary of Böhm ............................................................... 38
`
`4.
`
`
`
`-i-
`
`
`
`Case 1:14-cv-02396-PGG-SN Document 234-17 Filed 11/11/20 Page 4 of 69
`
`3.
`
`4.
`
`The Combination of Iggulden and Böhm Discloses Every
`Limitation of Claims 15-17, 21-22, 24-25, 27-28, 31-32,
`37, 40-41, and 46 ................................................................. 39
`A Person Having Ordinary Skill in the Art Would Have
`Been Motivated to Combine Iggulden with Böhm ................. 41
`Ground 5: Claims 1-3, 7, 9, 13-17, 21, 23, 27-33, 38, 40-42, 47,
`49, and 51 of the '988 Patent Are Anticipated By Iwamura .............. 46
`1.
`Summary of Iwamura ........................................................... 46
`2.
`Iwamura Discloses Every Element of Claims 1-3, 7, 9,
`13-17, 21, 23, 27-33, 38, 40-42, 47, 49, and 51 ...................... 46
`Ground 6: Claims 8, 10-12, 22, 24-26, 50, and 52 of the '988
`Patent Are Obvious Under 35 U.S.C. § 103 Over Ghias .................. 54
`G. Ground 7: Claims 8, 10-12, 22, 24-26, 50, and 52 of the '988
`Patent Are Obvious Under 35 U.S.C. § 103 Over Wood .................. 57
`VII. CONCLUSION ....................................................................................... 60
`
`
`E.
`
`F.
`
`
`
`-ii-
`
`
`
`Case 1:14-cv-02396-PGG-SN Document 234-17 Filed 11/11/20 Page 5 of 69
`
`EXHIBIT LIST
`
`U.S. Patent No. 8,010,988 to Cox
`
`Prosecution History of U.S. Patent No. 8,010,988
`
`Visual Summary of Petition Grounds
`
`December 3, 2014 Declaration of Dr. Pierre Moulin
`
`Curriculum Vitae of Dr. Pierre Moulin
`
`Sunil Arya, et al., "An Optimal Algorithm for Approximate Nearest
`Neighbor Searching in Fixed Dimensions" ("Arya")
`
`Christian Böhm, et al., "Efficient Similarity Search in Digital
`Libraries" ("Böhm")
`
`U.S. Patent No. 7,444,353 ("Chen")
`
`U.S. Patent No. 6,970,886 ("Conwell")
`
`U.S. Patent No. 5,874,686 ("Ghias")
`
`U.S. Patent No. 6,597,405 ("Iggulden")
`
`U.S. Patent No. 6,188,010 ("Iwamura")
`
`U.S. Patent No. 6,505,160 ("Levy")
`
`U.S. Patent No. 6,098,106 ("Philyaw")
`
`U.S. Patent No. 7,743,092 ("Wood")
`
`Prosecution History of U.S. Patent App. No. 09/438,469
`
`Timo Raita, "Tuning the Boyer–Moore–Horspool String Searching
`Algorithm"
`
`Aristides Gionis, et al., "Similarity Search in High Dimensions via
`Hashing"
`
`Plaintiff Network-1 Technologies, Inc.'s Responses to Defendants
`Google, Inc. and YouTube, LLC's First Set of Interrogatories (Nos.
`
`-iii-
`
`1001
`
`1002
`
`1003
`
`1004
`
`1005
`
`1006
`
`1007
`
`1008
`
`1009
`
`1010
`
`1011
`
`1012
`
`1013
`
`1014
`
`1015
`
`1016
`
`1017
`
`1018
`
`1019
`
`
`
`
`
`Case 1:14-cv-02396-PGG-SN Document 234-17 Filed 11/11/20 Page 6 of 69
`
`1-4) in Network-1 Technologies, Inc. v. Google, Inc., et al.
`(S.D.N.Y. Case No. 14-cv-2396-PGG)
`
`
`
`
`
`
`
`
`
`-iv-
`
`
`
`Case 1:14-cv-02396-PGG-SN Document 234-17 Filed 11/11/20 Page 7 of 69
`
`Google Inc. ("Petitioner"), petitions for inter partes review ("IPR") under 35
`
`U.S.C. §§ 311-319 and 37 C.F.R. § 42 of claims 1-3, 7-17, 21-34, 37-38, 40-43, 46-
`
`47, and 49-52 ("the '988 Claims") of U.S. Patent No. 8,010,988 ("the '988 patent")
`
`(Ex. 1001).
`
`I. MANDATORY NOTICES UNDER 37 C.F.R. § 42.8(A)(1)
`
`Petitioner and YouTube, LLC, a subsidiary of Petitioner, are real parties-in-
`
`interest with respect to the instant petition.
`
`The '988 patent is asserted in an action captioned Network-1 Technologies,
`
`Inc. v. Google Inc., et al., Case No. 1:14-cv-2396 (S.D.N.Y.), filed April 4, 2014.
`
`Pursuant to 37 C.F.R. § 42.10(b), a Power of Attorney accompanies this
`
`petition. Service of any documents via hand-delivery may be made at the postal
`
`mailing address of the respective lead or back-up counsel designated below:
`
`Lead Counsel
`James J. Elacqua (Reg. # 28,412)
`SKADDEN, ARPS, SLATE, MEAGHER &
`FLOM LLP
`525 University Avenue - Suite 1400
`Palo Alto, California 94301
`Telephone:
`(650) 470-4510
`James.Elacqua@skadden.com
`
`Back-Up Counsel
`Douglas R. Nemec (Reg. # 41,219)
`SKADDEN, ARPS, SLATE, MEAGHER &
`FLOM LLP
`Four Times Square
`New York, New York 10036-6522
`Telephone:
`(212) 735-2419
`Douglas.Nemec@skadden.com
`
`II.
`
`PAYMENT OF FEES UNDER 37 C.F.R. § 42.103
`
`Petitioner authorizes the Patent and Trademark Office to charge Deposit
`
`Account No. 19-2385 for the fee set in 37 C.F.R. § 42.15(a) for this Petition and
`
`authorize payment of any additional fees to be charged to this Deposit Account.
`
`
`
`--
`
`
`
`Case 1:14-cv-02396-PGG-SN Document 234-17 Filed 11/11/20 Page 8 of 69
`
`III. GROUNDS FOR STANDING UNDER 37 C.F.R. § 42.104(a)
`
`Petitioner certifies that the '988 patent is available for IPR and that: (1)
`
`Petitioner does not own the '988 patent; (2) prior to the date this Petition was filed,
`
`neither Petitioner nor any real party-in-interest filed a civil action challenging the
`
`validity of a claim in the '988 patent; (3) this Petition has been filed less than one
`
`year after April 11, 2014, which was the date on which Petitioner was served with a
`
`complaint alleging infringement of the '988 patent; and (4) neither Petitioner, any
`
`real parties-in-interest, nor any privies of Petitioner, are estopped from challenging
`
`the claims on the grounds identified in this Petition.
`
`IV. SUMMARY OF CHALLENGE AND RELIEF REQUESTED
`
`Petitioner asserts the following specific grounds of rejection under 35 U.S.C.
`
`§§ 102 and 103. These grounds are visually summarized in Exhibit 1003.
`
`Ground
`1
`
`Grounds for Trial
`'988 Patent Claims
`1-3, 7-9, 13-17, 21-23, 28, 30-31, 40, 49, 51 Anticipated by Ghias
`
`2
`
`3
`
`4
`
`5
`
`6
`
`Anticipated by Wood
`
`1-2, 7-9, 13-16, 21-23, 27-33, 38, 40-42,
`
`47, 49, 51
`1-3, 7-17, 21-34, 37-38, 40-43, 46-47, 49-52 Obvious Over Levy in
`
`View of Arya
`Obvious Over Iggulden
`
`in View of Böhm
`Anticipated by Iwamura
`
`15-17, 21-22, 24-25, 27-28, 31-32, 37,
`
`40-41, 46
`1-3, 7, 9, 13-17, 21, 23, 27-33, 38, 40-42,
`
`47, 49, 51
`8, 10-12, 22, 24-26, 50, 52
`
`Obvious over Ghias
`
`
`
`2
`
`
`
`Case 1:14-cv-02396-PGG-SN Document 234-17 Filed 11/11/20 Page 9 of 69
`
`7
`
`8, 10-12, 22, 24-26, 50, and 52
`
`Obvious over Wood
`
`V.
`
`SUMMARY OF THE '988 PATENT
`
`A. Algorithms for Automatic Content Recognition
`
`In the 1990s, numerous individuals concurrently developed systems for
`
`computer-automated recognition of audio and video content. Ex. 1004 at ¶ 9. Such
`
`systems universally relied on two widely known technologies: feature extraction
`
`and neighbor searching. Ex. 1004 at ¶ 10. Feature extraction refers to quantifying a
`
`media work in a form that—unlike a raw video feed—is easily parsed by a
`
`computer. Id. at ¶ 11. Neighbor searching refers to algorithms for comparing a first
`
`set of extracted features with one or more additional sets of extracted features to
`
`locate a close, but not necessarily exact, match. Id. at 12. Because neighbor
`
`searching is computationally intensive, content recognition schemes typically
`
`employed search algorithms that increased efficiency by intelligently searching
`
`only a subset of potential matches (i.e., "non-exhaustive" algorithms). Id.
`
`B.
`
`Summary of the '988 Claims
`
`The '988 Patent issued from U.S. Patent App. No. 11/445,928, filed on June
`
`2, 2006, and claims priority to Provisional App. No. 60/232,618, filed on
`
`September 14, 2000. Ex. 1001. The '988 patent has never previously been
`
`challenged in post-grant proceedings or in Court.
`
`The '988 Claims relate to identifying an unknown media work and, based on
`
`its determined identity, performing an action. Ex. 1001 at Abstract, Claims 1, 15.
`
`
`
`3
`
`
`
`Case 1:14-cv-02396-PGG-SN Document 234-17 Filed 11/11/20 Page 10 of 69
`
`The independent '988 Claims are directed to four elements for identifying an
`
`electronic work and performing a corresponding action: (1) "electronically
`
`extracting features from the electronic work"; (2) "electronically determining an
`
`identification of the electronic work based on the extracted features" by
`
`performing "a non-exhaustive search identifying a neighbor"; (3) "electronically
`
`determining an action based on the identification of the electronic work"; and
`
`(4) "electronically performing the action." Id. at Claims 1, 15; Ex. 1004 at ¶ 15.
`
`Although the specification highlights the use of the claimed features to
`
`identify a television commercial (e.g., Ex. 1001 at 10:5-39), the claims are not
`
`limited to television commercials and the specification explicitly envisions the
`
`identification of "content" generally (e.g., id. at 1:23-27 ("e.g., content or an
`
`advertisement")). Ex. 1004 at ¶ 23. Further, while the specification highlights the
`
`use of certain exemplary algorithms for extracting features from an electronic work
`
`(e.g., Ex. 1001 at 20:11-45,7:14-8:2) and determining an identification of an
`
`electronic work by performing a non-exhaustive search identifying a neighbor (e.g.,
`
`id. at 9:29-39, 11:15-25, 21:23-39, 22:1-37), the broadest reasonable construction
`
`of the claims is not restricted to any particular algorithm. Ex. 1004 at ¶ 23.
`
`The claimed steps are illustrated in several figures of the '988 patent,
`
`including Figure 1, which demonstrates (1) "feature (vector) extraction
`
`operation(s)" (140), which "extract features from [an electronic] work" (Ex. 1001 at
`
`
`
`4
`
`
`
`Case 1:14-cv-02396-PGG-SN Document 234-17 Filed 11/11/20 Page 11 of 69
`
`6:61-65); (2) "feature (vector) lookup operation(s)" (150), which identify a work by
`
`"search[ing] for a matching feature vector" (id. at 6:66-7:3); (3) "work-associated
`
`information lookup operation(s)" (160), which "retrieve associated information,
`
`such as an action" (id. at 7:4-7); and (4) "action initiation operation(s)" (170),
`
`which "perform some action based on the associated information (id. at 7:8-9). Ex.
`
`1004 at ¶ 24. Figures 2-4 clarify that, in the absence of claim language to the
`
`contrary, the foregoing claimed steps can be performed on any combination of a
`
`local device and a remote server. Id.
`
`C.
`
`Person of Ordinary Skill in the Art of the '988 Patent
`
`In 2000, a person of ordinary skill in the art of the '988 patent would have
`
`been highly skilled, and typically would have possessed at least an M.S. in
`
`computer science, electrical engineering, or mathematics; knowledge of video and
`
`audio processing techniques; and 1-2 years of experience in audio, video, or image
`
`processing. Ex. 1004 at ¶ 7.
`
`D. Claim Construction Under 37 C.F.R. § 42.104(b)(3)
`
`The claim terms should be given their "broadest reasonable construction in
`
`light of the specification." 37 C.F.R. § 42.100(b). Although Petitioner reserves the
`
`right to present different constructions in related litigation—which applies a
`
`different standard—proposed broadest reasonable constructions are set forth below.
`
`Variations of the term "extracting features" appear in all independent '988
`
`
`
`5
`
`
`
`Case 1:14-cv-02396-PGG-SN Document 234-17 Filed 11/11/20 Page 12 of 69
`
`Claims, and should be construed to mean "generating a quantitative representation
`
`of the features of a media work, or a subset of the features of a media work." Ex.
`
`1004 at ¶ 39. In 2000, persons skilled in the art consistently employed this
`
`definition. Id. at ¶ 41. Moreover, the patent specification recognizes that "[t]he
`
`recognition literature contains many different" extraction algorithms, and
`
`enumerates a broad range of algorithms capable of "feature extraction," including
`
`"pseudo-random sample[s] of pixels," "the average intensities of nxn blocks of
`
`pixels," "frequency-based decomposition of the signal, such as produced by the
`
`Fourier, wavelet and or discrete cosine transforms" "principal component analysis,"
`
`"a temporal sequence of feature vectors," or "a combination of these." E.g., Ex.
`
`1001 at 7:14-43. The unifying characteristic of these methods is that they generate
`
`a quantitative representation of the features of a media work. Ex. 1004 at ¶ 40.
`
`The term "non-exhaustive search" appears in all independent '988 Claims,
`
`and should be construed—under the "broadest reasonable construction" standard—
`
`to mean "a search that locates a match without conducting a brute force comparison
`
`of all possible matches, and all data within all possible matches." Ex. 1004 at ¶ 43.
`
`The specification never uses the term "non-exhaustive" or "exhaustive." Ex. 1001.
`
`However, in 2000, persons skilled in the art employed multiple constructions, the
`
`broadest of which was consistent with the above definition. Ex. 1004 at ¶ 43. Any
`
`search algorithm that is not guaranteed to find an existing match is necessarily non-
`
`
`
`6
`
`
`
`Case 1:14-cv-02396-PGG-SN Document 234-17 Filed 11/11/20 Page 13 of 69
`
`exhaustive under this definition. Id. at ¶ 44.
`
`The term "identifying a neighbor" appears in all independent '988 Claims,
`
`and should be construed to mean "identifying a close, but not necessarily exact,
`
`match." Ex. 1004 at ¶ 48. In 2000, persons skilled in the art consistently employed
`
`this definition. Id. Consistent with this definition, the specification notes that one
`
`type of neighbor search is a "fixed radius search" in which "if the database contains
`
`a vector that is within X of the query, then there is a match." Ex. 1001 at 22:1-23;
`
`Ex. 1004 at ¶ 52. Consistent with the above, the term "neighbor" should be
`
`construed to mean "a close, but not necessarily exact, match." Ex. 1004 at ¶ 50.
`
`The term "sublinear" search appears in dependent claims 3 and 17, and
`
`should be construed to mean "a search whose execution time has a sublinear
`
`relationship to database size." Ex. 1004 at ¶ 53. For instance, a linear search of a
`
`200-item database would take twice as long as a linear search of a 100-item
`
`database. Id. By contrast, a sublinear search of a 200-item database would take less
`
`than twice as long as a sublinear search of a 100-item database, perhaps, for
`
`instance, 1.5 times as long. Id. If search execution time were plotted as a function
`
`of database size, a linear search would yield a diagonal line of constant slope, while
`
`a sublinear search would yield a line with a slope approaching zero. Id at ¶¶ 53-54.
`
`The specification explains that at least the following algorithms are sublinear: (1) a
`
`search with execution time proportional to the logarithm of the size of the data set
`
`
`
`7
`
`
`
`Case 1:14-cv-02396-PGG-SN Document 234-17 Filed 11/11/20 Page 14 of 69
`
`(e.g., execution time proportional to A log (BN), where A and B are constants, and
`
`N is the number of entries in the database) (Ex. 1001 at 9:19-28 (contrasting binary
`
`search with linear search, noting that, "[i]f binary search was possible, then a
`
`database containing N vectors would require at most log(N) comparisons.")); (2)
`
`"kd-trees and vantage point trees" (id. at 22:5-9); (3) and "excluded middle vantage
`
`point forest" (id. at 22:10-23). Ex. 1004 at ¶¶ 56-58. In 2000, persons skilled in the
`
`art consistently employed this definition. Ex. 1004 at ¶ 55.
`
`For all other claim terms, the plain and ordinary meaning should apply.
`
`VI. ANALYSIS OF GROUNDS FOR TRIAL
`
`A. Ground 1: Claims 1-3, 7-9, 13-17, 21-23, 28, 30-31, 40, 49, and 51
`of the '988 Patent Are Anticipated By Ghias Under U.S.C.
`§ 102(b)
`
`1.
`
`Summary of Ghias
`
`Ghias was filed on October 31, 1996, and issued on February 23, 1999. Ex.
`
`1010. Accordingly, it is prior art under 35 U.S.C. § 102(b). Ghias was not before
`
`the examiner during prosecution of the '988 patent.
`
`Ghias discloses a computer system for identifying a hummed melody. As
`
`Figure 1 illustrates, a "tune 12 is hummed by a person 18 into a microphone 20."
`
`Ex. 1010 at 2:41-42. The data from the microphone is fed into "a pitch tracking
`
`module 22 in computer 16" which extracts "a contour representation" of the
`
`melody. Id. at 2:46-49. The computer then employs a "query engine 24" which
`
`"searches the melody database 14" to locate a matching melody. Id. at 2:50-59.
`
`
`
`8
`
`
`
`Case 1:14-cv-02396-PGG-SN Document 234-17 Filed 11/11/20 Page 15 of 69
`
`Then, the computer takes the action of "output[ting] a ranked list of approximately
`
`matching melodies." Id. at 2:51-52, 6:60-63, 7:4-5, 8:26-28, 8:61-63.
`
`2. Ghias Discloses Every Element of Claims 1-3, 7-9, 13-17, 21-
`23, 28, 30-31, 40, 49, and 51 of the '988 Patent
`
`Ghias discloses all four elements of the independent '988 Claims. Ex. 1004 at
`
`¶¶ 67-72. First, the '988 Claims require "electronically extracting features from the
`
`electronic work." Ex. 1001 at Claims 1, 15. Ghias discloses extracting "a contour
`
`representation" of the melody (Ex. 1010 at 2:46-49) wherein humming "is
`
`converted into a sequence of relative pitch transitions . . . To accomplish this . . . a
`
`sequence of pitches in the melody must be isolated and tracked" (id. at 3:8-22).
`
`Second, the '988 Claims require "electronically determining an identification
`
`of the electronic work based on the extracted features by performing a non-
`
`exhaustive search identifying a neighbor." Ex. 1001 at Claims 1, 15. Ghias
`
`discloses "search[ing] the melody database 14" to locate a matching melody. Ex.
`
`1010 at 2:50-59; Abstract ("A melody database is searched for at least one
`
`sequence of digitized representations of relative pitch differences between
`
`successive notes which at least approximately matches . . . the melody."). Ghias
`
`further discloses that this search may be non-exhaustive, through the use of "an
`
`efficient approximate pattern matching algorithm" rather than an algorithm that is
`
`guaranteed to yield a match. Id. at 6:7-11. Moreover, Ghias teaches that "Several
`
`Algorithms have been developed that address [this] problem" ranging from "brute
`
`
`
`9
`
`
`
`Case 1:14-cv-02396-PGG-SN Document 234-17 Filed 11/11/20 Page 16 of 69
`
`force" to substantially faster algorithms. Id. at 23-35 ("Several Algorithms have
`
`been . . . Running times have ranged from 0(mn) for the brute force algorithm to
`
`O(kn) or O(nlog(m) . . ."). Id. at 6:23-35.
`
`Ghias further discloses that this search locates a neighbor by determining "a
`
`ranked list of approximately matching melodies, as illustrated at 26" or "the single
`
`most approximate matching melody." Ex. 1010 at 2:50-59, 6:60-63.
`
`Third, the '988 Claims require "electronically determining an action based on
`
`the identification of the electronic work." Ghias discloses using the results of a
`
`search to determine the potential matches (Ex. 1010 at 2:50-52, 6:60-63, 7:4-5,
`
`8:26-28, 8:61-63), and how many potential matches to display (id. at 6:60-63 ("The
`
`number of matches that the database 14 should retrieve depends upon the error-
`
`tolerance."), 2:53-55 ("The query engine 24 may of course alternatively be
`
`programmed to output the single most approximate matching melody . . . .")).
`
`Ghias further discloses that "[i]f the list [of potential matches] is too large" the
`
`system may perform the additional action of allowing the user to perform a "new
`
`query on a restricted search list consisting of songs just retrieved." Id at 7:5-7.
`
`Fourth, the '988 Claims require "electronically performing the action." Ex.
`
`1001 at Claims 1, 15. Ghias discloses that, after generating the list of matches, the
`
`system performs the action of "output[ting] the ranked list of approximately
`
`matching melodies." Ex. 1010 at 2:51-52, 6:60-63, 8:26-28, 8:61-63.
`
`
`
`10
`
`
`
`Case 1:14-cv-02396-PGG-SN Document 234-17 Filed 11/11/20 Page 17 of 69
`
`Claims 3 and 17 further require that the search is "sublinear." Ex. 1001 at
`
`Claims 3, 17. Ghias discloses search algorithms that are substantially faster than
`
`"brute force" searches. Ex. 1010 at 6:23-35. In particular Ghias discloses searches
`
`whose execution times are proportional to the logarithm of the size of the data set
`
`(id. at 6:24-28 ("O(kn) or O(nlog(m)")), which, as explained above in Section
`
`V(D), are sublinear (Ex. 1001 at 9:19-28). Ex. 1004 at ¶ 74.
`
`For the foregoing reasons, and as shown in detail in the claim chart below,
`
`claims 1-3, 7-9, 13-17, 21-23, 28, 30-31, 40, 49, and 51 of the '988 Patent are
`
`taught by Ghias, and are therefore unpatentable. Ex. 1004 at ¶ 75.
`
`'988 Patent
`1. A method for associating an
`electronic work with an action, the
`electronic work comprising at least
`one of audio and video, the
`method comprising:
`
`a) electronically extracting within
`a portable client device features
`from the electronic work;
`
`b) transmitting the extracted
`features from the portable client
`device to one or more servers;
`
`Ghias (Ex. 1010)
`
`Non-limiting preamble.
`
`Ghias discloses extracting features when
`humming "is converted into a sequence of
`relative pitch transitions, as follows . . ." 3:8-
`22, 2:42-50, 4:31-5:58, 8:29-40, 8:64-10:4.
`These electronic extraction steps are
`performed by a "computer," which is a
`portable client device. E.g., Fig. 1 at 16.
`
`Ghias discloses a computer, which is a
`portable client device (Ex. 1004 at ¶ 73), that
`transmits the extracted features to the server,
`i.e., "the database 14 [which] may be located
`apart from the computer 16 and suitably
`connected thereto for communicating
`between the computer and database." 2:34-
`40.
`
`
`
`11
`
`
`
`Case 1:14-cv-02396-PGG-SN Document 234-17 Filed 11/11/20 Page 18 of 69
`
`'988 Patent
`
`Ghias (Ex. 1010)
`
`c) receiving at the portable client
`device from the one or more
`servers an identification of the
`electronic work based on the
`extracted features, wherein the
`identification is based on a non-
`exhaustive search identifying a
`neighbor;
`
`d) electronically determining an
`action based on the identification
`of the electronic work; and
`
`e) electronically performing the
`action on the portable client
`device.
`
`Ghias receives and outputs at the computer,
`which is a portable client device (Ex. 1004 at
`¶ 73), a list of identifications of electronic
`works. 2:50-52, 6:60-63, 7:4-5, 8:26-28,
`8:61-63. Such identifications are determined
`by "searching the melody database 14" to
`locate a matching melody. 2:50-59, 6:60-63,
`7:4-5, Abstract, 8:26-28, 8:61-63. This
`search may employ a non-exhaustive
`"approximate pattern matching algorithm" or
`another algorithm that operates faster than a
`brute force search. 6:7-11, 6:23-35. This
`non-exhaustive search identifies a neighbor,
`i.e., "a ranked list of approximately matching
`melodies." 2:50-59, 6:60-63.
`
`Ghias uses the results of a search to
`determine the potential matches (2:50-52,
`6:60-63, 7:4-5, 8:26-28, 8:61-63), and how
`many potential matches to display to a user
`(6:60-65 ("The number of matches that the
`database 14 should retrieve depends upon the
`error-tolerance"). "If the list [of potential
`matches] is too large" the system may
`determine that it should perform the
`additional action of allowing the user to
`perform "a new query on a restricted search
`list consisting of songs just retrieved." Id at
`7:5-7.
`
`After generating the list of matches, the
`system performs the action of "output[ting]
`the ranked list of approximately matching
`melodies." 2:50-52, 6:60-63, 7:4-5, 8:26-28,
`8:61-63. This occurs on the computer, i.e.,
`the portable client (Ex. 1004 at ¶ 73). 8:26-
`28 ("outputting from said computer a ranked
`list"), 8:61-63.
`
`
`
`12
`
`
`
`Case 1:14-cv-02396-PGG-SN Document 234-17 Filed 11/11/20 Page 19 of 69
`
`'988 Patent
`
`Ghias (Ex. 1010)
`
`2. A method of claim 1, wherein
`the identification is based on a
`non-exhaustive search identifying
`a neighbor within a fixed radius.
`
`3. The method of claim 1, wherein
`the non-exhaustive search is
`sublinear.
`
`7. The method of claim 1, wherein
`the electronic work is an audio
`work.
`
`8. The method of claim 7, wherein
`the audio work is obtained from at
`least one of a broadcast and an
`audio file format.
`
`Ghias discloses that the non-exhaustive
`neighbor search can be limited to results
`within a fixed radius of an exact match by
`applying a "preselected error tolerance to the
`search." 2:52-53, 6:63-7:3.
`
`Ghias discloses sublinear searches whose
`execution times are proportional to the
`logarithm of the data set. 6:24-28 ("O(kn) or
`O(nlog(m)"); Ex. 1004 at ¶ 74.
`
`Ghias discloses an electronic audio work that
`is "hummed by a person 18 into a
`microphone 20" and "suitably digitized."
`E.g., 2:41-43, Fig. 1 at 21.
`
`Ghias discloses that "[h]ummed queries may
`be recorded in a variety of formats"
`including audio file formats such as "16-bit,
`44 Khz WAV format on a Pentium system,
`or a 8-bit 8 Khz AU format on a Sun
`Sparcstation." 2:67-3:4.
`
`9. The method of claim 7, wherein
`the identification includes at least
`one of a song title, an album title,
`and a performer name.
`
`Ghias discloses "return[ing] a list of songs."
`6:60-63, 2:51-52, 7:4-5, 8:26-28, 8:61-63.
`"From the results of the query the user can
`identify the song of interest." 7:4-8.
`
`13. The method of claim 1,
`wherein the step of electronically
`determining the action includes
`receiving at the portable client
`device an action based on the
`identification of the electronic
`work from the one or more
`servers.
`
`Ghias discloses, at the computer, which is a
`portable client device, receiving and
`"output[ting] the ranked list of
`approximately matching melodies" based on
`the identity of the electronic work. 2:50-52,
`6:60-63, 7:4-5, 8:26-28, 8:61-63. This action
`is received from "the database 14 [which]
`may be located apart from the computer 16
`and suitably connected thereto for
`communicating." 2:33-40.
`
`14. The method of claim 1,
`wherein the step of electronically
`extracting the features is
`performed by at least one of a
`
`Ghias discloses that extracting " may be
`performed in, for example, Matlab software,"
`which runs on a microprocessor on a
`computer, which is a portable client device.
`
`
`
`13
`
`
`
`Case 1:14-cv-02396-PGG-SN Document 234-17 Filed 11/11/20 Page 20 of 69
`
`'988 Patent
`
`Ghias (Ex. 1010)
`
`microprocessor of the portable
`client device and a digital signal
`processor of the portable client
`device.
`15. A method for associating an
`electronic work with an action, the
`electronic work comprising at least
`one of audio and video, the
`method comprising:
`
`2:63-67. Similarly, Ghias further discloses
`operating on a "Pentium system" or "a Sun
`Sparcstation." 3:2-4.
`
`Non-limiting preamble.
`
`a) electronically extracting
`features from the electronic work;
`
`Petitioner incorporates the above discussion
`of Ghias regarding Claim 1a.
`
`b) electronically determining an
`identification of the electronic
`work based on the extracted
`features, wherein the identification
`is based on a non-exhaustive
`search identifying a neighbor;
`c) electronically determining an
`action based on the identification
`of the electronic work; and
`d) electronically performing the
`action.
`16. A method of claim 15, wherein
`the identification is based on a
`non-exhaustive search identifying
`a neighbor within a fixed radius.
`17. The method of claim 15,
`wherein the non-exhaustive search
`is sublinear.
`21. The method of claim 15,
`wherein the electronic work is an
`audio work.
`22. The method of claim 21,
`wherein the audio work is
`obtained from at least one of a
`broadcast and an audio file format.
`23. The method of claim 21,
`
`Petitioner incorporates the above discussion
`of Ghias regarding Claim 1c.
`
`Petitioner incorporates the above discussion
`of Ghias regarding Claim 1d.
`
`Petitioner incorporates the above discussion
`of Ghias regarding Claim 1e.
`
`Petitioner incorporates the above discussion
`of Ghias regarding Claim 2.
`
`Petitioner incorporates the above discussion
`of Ghias regarding Claim 3.
`
`Petitioner incorporates the above discussion
`of Ghias regarding Claim 7.
`
`Petitioner incorporates the above discussion
`of Ghias regarding Claim 8.
`
`Petitioner incorporates the above discussion
`
`
`
`14
`
`
`
`Case 1:14-cv-02396-PGG-SN Document 234-17 Filed 11/11/20 Page 21 of 69
`
`'988 Patent
`
`Ghias (Ex. 1010)
`
`wherein the identification includes
`at least one of a song title, an
`album title, and a performer name.
`
`of Ghias regarding Claim 9.
`
`Ghias discloses that "a user can" interact
`with the system by "perform[ing] a new
`query on a restricted search list consisting of
`songs just retrieved. This allows the user to
`identify sets of songs that contain similar
`melodies." Id at 7:5-7.
`
`Petitioner incorporates the above discussion
`of Ghias regarding Claim 28.
`
`Ghias discloses providing and displaying "a
`list of songs." 6:60-63, 2:51-52, 7:4-5, 8:26-
`28, 8:61-63. This list includes additional
`information such as a ranking of the strength
`of the match. 6:60-63 ("the database 14
`returns a list of songs ranked by how well
`they matched the query"), 2:50-52, 8:26-28,
`8:61-63.
`
`Petitioner incorporates the above discussion
`of Ghias regarding Claim 31.
`
`Petitioner incorporates the above discussion
`of Ghias regarding Claims 7 and 9.
`
`Petitioner incorporates the above discussion
`of Ghias regarding Claims 7 and 9.
`
`28. The method of claim 15,
`wherein the action promotes
`interaction.
`
`30. The method of claim 1,
`wherein the action promotes
`interaction.
`
`31. The method of claim 15,
`wherein the action comprises
`providing a