`
`Paper ____
`
`Filed on behalf of: Software Rights Archive, LLC
`
`
`
`
`By: Martin M. Zoltick, Lead Counsel
`
`Nancy J. Linck, Backup Counsel
`
`Rothwell, Figg, Ernst & Manbeck, P.C.
`607 14th St., N.W., Suite 800
`
`
`Washington, DC 20005
`
`Phone: 202-783-6040
`
`Facsimile: 202-783-6031
`
`E-mail: mzoltick@rfem.com
`
`
`
`nlinck@rfem.com
`
`
`UNITED STATES PATENT AND TRADEMARK OFFICE
`
`_______________
`
`
`BEFORE THE PATENT TRIAL AND APPEAL BOARD
`
`_______________
`
`
`FACEBOOK, INC., LINKEDIN CORP., AND TWITTER, INC.,
`Petitioners
`v.
`
`SOFTWARE RIGHTS ARCHIVE, LLC
`Patent Owner
`_______________
`
`
`Case IPR2013-00480
`Patent 5,832,494
`_______________
`
`
`PATENT OWNER’S RESPONSE
`
`
`
`TABLE OF CONTENTS
`TABLE OF AUTHORITIES ................................................................................... iv
`I.
`INTRODUCTION ........................................................................................... 1
`II.
`OVERVIEW OF U.S. PATENT NO. 5,832,494 ............................................ 3
`A. General Background .............................................................................. 3
`B.
`Claims of the ‘494 Patent ...................................................................... 7
`III. THE BOARD’S DECISION INSTITUTING INTER PARTES
`REVIEW ......................................................................................................... 10
`IV. PERSON OF ORDINARY SKILL IN THE ART ........................................ 12
`V.
`CLAIM CONSTRUCTION .......................................................................... 12
`A. Applicable Law ................................................................................... 12
`B.
`Board Claim Construction ................................................................... 15
`VI. FOX SMART DOES NOT ANTICIPATE CLAIMS 1 AND 5
`(GROUND 2) ................................................................................................. 16
`A.
`Summary of Fox SMART ................................................................... 16
`B.
`Fox Smart does not disclose “an analysis of one or more indirect
`relationships in the database,” as recited in claim 1 ........................... 21
`1.
`The Board’s construction appears to contain a technical
`error ........................................................................................... 21
`Fox SMART fails to disclose or suggest “indirect
`relationships in the database” as properly construed ................ 23
`Even if an Indirect Relationship Did Not Require a Chain of
`Citations to be in a Database, there are Still No disclosed
`Indirect Relationships in the Database. ..................................... 29
`Fox Smart does not disclose “generating candidate cluster links
`for the selected node,” as recited in claim 1 ........................................ 31
`Fox SMART does not disclose “deriving actual cluster links from
`the candidate cluster links,” as recited in claim 1 ............................... 33
`Fox SMART does not disclose or suggest “selecting a node for
`analysis,” as recited in claim 1 ............................................................ 39
`Fox Smart does not disclose or suggest “displaying the identity of
`one or more nodes using the actual cluster links,” as recited in
`claim 1 ................................................................................................. 41
`ii
`
`D.
`
`2.
`
`3.
`
`C.
`
`E.
`
`F.
`
`
`
`
`
`
`
`
`
`G.
`
`H.
`
`B.
`
`C.
`
`Fox SMART Does not Disclose the Features of Claim 1 As
`Arranged in the Claim ......................................................................... 42
`Fox Smart does not disclose the additional features “eliminating
`candidate cluster links, wherein the number of candidate cluster
`links are limited, and the closest candidate cluster links are
`chosen,” as recited in claim 5. ............................................................. 44
`VII. FOX THESIS DOES NOT ANTICIPATE CLAIMS 14-16
`(GROUND 1) ................................................................................................. 46
`A.
`Fox Thesis does not disclose initializing a set of candidate cluster
`links ..................................................................................................... 47
`Fox Thesis does not disclose or suggest “selecting the destination
`node of a path as the selected node,” as recited by claim 14 .............. 47
`Fox Thesis does not disclose or suggest “retrieving the set of
`direct links from the selected node to any other node in the
`database,” as recited by claim 14 ........................................................ 49
`Fox Thesis does not disclose or suggest “determining the weight
`of the path using the retrieved direct links,” as recited by claim 14 ... 51
`Fox Thesis does not disclose or suggest “storing the determined
`weights as candidate cluster links,” as recited by claim 14 ................ 53
`Fox Thesis does not disclose or suggest “deriving the actual
`cluster links,” as recited by claim 15 .................................................. 54
`Fox Thesis does not disclose or suggest “choosing the top rated
`cluster links,” as recited by claim 16 .................................................. 55
`VIII. CONCLUSION .............................................................................................. 58
`
`
`
`
`D.
`
`E.
`
`F.
`
`G.
`
`
`
`iii
`
`
`
`Cases
`
`TABLE OF AUTHORITIES
`
`Ex parte Akihiko Toyoshima,
`2012 WL 4718521 (BPAI 2012) .......................................................................... 14
`Ex Parte Papst-Motoren,
`1 U.S.P.Q.2d 1655 (B.P.A.I. 1986) ............................................................... 13, 23
`Ex parte Quickie,
`2012 WL 2316828 (BPAI 2012) .......................................................................... 13
`Ex parte Suitco Surface, Inc.,
`2011 WL 3876561 (BPAI 2012) .......................................................................... 14
`Ex parte Tractus Medical, Inc.,
`2012 WL 759848 (BPAI 2012) ............................................................................ 14
`Phillips v. AWH Corp.,
`415 F.3d 1303, 75 USPQ2d 1324 (Fed. Cir. 2005) ......................... 12, 13, 14, 23
`Statutes
`
`35 U.S.C. § 316(d)(1)(A) ........................................................................................... 1
`Other Authorities
`
`MPEP § 2258 ........................................................................................................... 13
`Regulations
`
`37 C.F.R. § 42.121 ..................................................................................................... 1
`
`
`
`
`
`iv
`
`
`
`
`I.
`
`
`
`
`INTRODUCTION
`
`The claims of the ‘494 patent are directed to the analysis and searching of a
`
`computerized database of textual objects containing citations to other objects
`
`stored in the database. The ‘494 patent specification discloses computerized tools
`
`and methods for extracting keywords and direct relationships from the content of
`
`database objects and generating representations of direct and indirect relationships
`
`existing among the database objects.
`
`The institution of the present trial on claims 1, 5, and 14-16 of the ‘494
`
`patent1 is entirely based on prior art describing experiments conducted by
`
`Petitioners’ expert Dr. Fox. The experiments of Dr. Fox are not directed to the
`
`analysis of a computer database of objects, but rather are directed towards limited
`
`experimentation with relationships existing among printed documents. This prior
`
`art describes limited experiments to determine whether direct and indirect
`
`relationships existing among paper documents are useful for clustering and
`
`searching.
`
`
`1 Patent Owner is filing concurrently herewith, in accordance with the Board’s
`
`Order (Paper 28 at 2) and pursuant to 35 U.S.C. § 316(d)(1)(A) and 37 C.F.R.
`
`§ 42.121, a motion to amend the ’494 patent by cancelling challenged claims 8, 10,
`
`11, 35, and 40.
`
`
`
`1
`
`
`
`
`
`
`
`None of the prior art describes any tools or methods of analysis of objects in
`
`a computer database with citations. Indeed, the information analyzed was either
`
`supplied to Dr. Fox prior to the experiments by third parties or gleaned from
`
`manual reading of paper documents. These experiments certainly do not teach or
`
`suggest a method of analyzing a database by generating “cluster links,” a unique
`
`and novel method of representing indirect relationships contained within a
`
`computer database, or for using cluster links in connection with providing search
`
`results.
`
`The Petitioners generally rely on a “clustering” process described in Dr.
`
`Fox’s papers for teaching the “cluster link” features of the claims of the ‘494
`
`patent. However, the Petitioner’s sole reliance on this “clustering” process appears
`
`to be based on the fact that “clustering” sounds like “cluster link,” and therefore,
`
`Dr. Fox’s papers must anticipate the claims of the ‘494 patent. The Petitioner’s
`
`citations to disparate unrelated portions of Dr. Fox’s papers in the Petition
`
`evidence the severe deficiencies of Dr. Fox’s papers. The steps recited in the
`
`challenged claims of the ‘494 patent are arranged in a specific manner to achieve
`
`the analysis and searching of a computerized database. However, nowhere can
`
`Petitioners show that Dr. Fox teaches “selecting a node for analysis,” “generating
`
`candidate cluster links for the selected node,” and “deriving actual clusters links”
`
`from these generated “candidate cluster links,” as required by claim 1, and
`
`
`
`2
`
`
`
`
`“selecting the destination node,” retrieving the set of direct links from the selected
`
`
`
`node,” “determining the weight of the path of the retrieved links,” and “storing the
`
`determined weights as candidate cluster links,” as required by claim 14. At best,
`
`the Petitioners merely slap together various disparate quotes from unrelated
`
`portions of Dr. Fox’s papers, which when applied to the claimed features as
`
`arranged, do not anticipate the claims of the ‘494 patent.
`
`II. OVERVIEW OF U.S. PATENT NO. 5,832,494
`A. General Background
`U.S. Patent No. 5,832,494 (“the ‘494 patent”) (Ex. 1001) is directed towards
`
`“[a] computer research tool for indexing, searching and displaying data …” ’494
`
`patent at Abstract. “Textual objects and other data in a database or network [are]
`
`indexed by creating a numerical representation of the data.” Id. The ‘494 patent
`
`specification describes a system that analyzes the content of an existing computer
`
`database to determine the relationships and patterns that exist between the objects
`
`contained within the database:
`
`This invention is a system for computerized searching of data.
`Specifically, the present invention significantly aids a researcher in
`performing computerized research on a database. … The invention
`can be used with an existing database by indexing the data and
`creating a numerical representation of the data. This indexing
`technique called proximity indexing generates a quick-reference of the
`relations, patterns, and similarity found among the data in the
`
`
`
`3
`
`
`
`
`
`
`
`database. Using this proximity index, an efficient search for pools of
`data having a particular relation, pattern or characteristic can be
`effectuated.
`
`‘494 patent at 3:19-34. To this end, the claimed system employs specialized
`
`computer tools for extracting relationships from the textual content of database
`
`objects for purposes of conducting an analysis of both semantical and non-
`
`semantical relationships that exist among the database objects. A “Proximity
`
`Indexing Application Program indexes the database into a more useful format to
`
`enable the Computer Search Program for Data Represented by Matrices (CSPDM)
`
`to efficiently search the database.” ‘494 patent at 3:61-63. This program includes
`
`an “extractor” which analyzes the text of database objects to extract citations (i.e.,
`
`any reference to and from database objects to other database objects) and keywords
`
`for purposes of creating a proximity index. ‘494 patent at 16:47-17:32; FIGS 3B,
`
`3D. This program also includes a “patterner” and “cluster link generator” to
`
`automatically generate cluster links of indirect relationships from the extracted
`
`content so that a large computer database can be analyzed and searched. ‘494
`
`patent at 17:33-18:58; FIGS 3B-3H; 21:30-24:16.
`
`Similarly, the ‘494 patent includes search and display routines that display
`
`content from the “referenced” electronic database objects:
`
`Another feature of the preferred embodiment is the "show usage"
`command. FIG. 8 is a screen display 38 depicting the use of the "show
`
`
`
`4
`
`
`
`
`
`
`
`usage" command. The preferred embodiment includes this command
`to allow the user to see a portion of an object in the database 54 which
`uses cites or refers to the node 2008 from which the show usage
`command is requested.
`
`‘494 patent at 47:1-7 (other content display routines including Cases-In, Cases-Out
`
`and Similar Cases: ‘494 patent at 25:24-31:47; FIGS 5A-D). Critical to the
`
`functioning of the above systems and methods is a database that contains objects
`
`that cite to other objects in the database. This is a non-trivial limitation. Without
`
`such a database, the “extractor” cannot analyze the content of database objects to
`
`extract their direct relationships (and keywords) and the “patterner” and “cluster
`
`link generator” cannot automatically generate representations of indirect
`
`relationships that exist in the database. ‘494 patent at 16:47-17:32; 17:33-18:58;
`
`FIGS 3B-3D. Similarly, the search and display routines cannot display the content
`
`of objects that refer to each other without such content being stored in an electronic
`
`database. See, e.g., ‘494 patent at 25:24-31:47; FIGS 5A-D.
`
`Further, the inventive method may generate “candidate cluster links” and
`
`derive “actual cluster links” to determine the strongest cluster link relationships in
`
`the database to efficiently analyze a database. ‘494 patent at 21:45 – Col. 24:15.
`
`Only a subset of the candidate cluster links, the “actual cluster links,” need be used
`
`in the proximity index and may be selected based on certain criteria. See ‘494
`
`patent at 22:1-4. A cluster link generator is a specific type of a patterner, and, as
`
`
`
`5
`
`
`
`
`such, its output is used to support the pool and single node search routines of the
`
`
`
`CSPDM. ‘494 patent at 21:30-33; see, e.g., ‘494 patent (Step 460 of Fig. 4H).
`
`Ranks and other values for the search nodes (i.e., nodes that are analyzed for
`
`searching) of the pool based and single node search routines are generated using
`
`cluster links. See, e.g., ‘494 patent (Step 464 and 460 of Fig. 4H). Cluster links
`
`are used in searching (i.e., stored so that they may be used to search for objects or
`
`generate values used to search for objects). ‘494 patent at 21:30-33. The claims of
`
`the ‘494 patent are directed to several embodiments of the inventive computer
`
`research tools for indexing, searching, and displaying data detailed in the
`
`specification. Limitations such as “generating candidate cluster links,” “deriving
`
`actual cluster links” and “determining the weight of the path using the retrieved
`
`direct links” are all related to the analysis and/or display of indirect relationships
`
`(i.e., inherent relationships2 of objects related through a chain of referential
`
`
`2“Direct relationship” and “citations” are inherent referential relationships between
`database objects. An inherent relationship is a referential relationship that exists
`within the database object to another database object and is not merely imputed or
`inserted as part of the subsequent proximity analysis of the database. For example,
`a hyperlink citation reflecting a reference placed by an author on the web is an
`inherent link, whereas a hyperlink subsequently generated to reflect a semantical
`word match that is imputed as part of analyzing the database is not an inherent
`relationship or non-semantical relationship.
`
`
`
`6
`
`
`
`
`citations) in a database. The analysis indirect relationships is an important aspect
`
`
`
`of the invention.
`
`B. Claims of the ‘494 Patent
`Claims 1, 5, and 14-16 of the ‘494 patent recite methods of analyzing a
`
`computerized database using cluster links to analyze indirect relationships in the
`
`database.
`
`Claim 1 recites:
`
`A method of analyzing a database with indirect relationships, using
`links and nodes, comprising the steps of:
`
`selecting a node for analysis;
`
`generating candidate cluster links for the selected node, wherein
`the step of generating comprises an analysis of one or more indirect
`relationships in the database;
`
`deriving actual cluster links from the candidate cluster links;
`
`identifying one or more nodes for display; and
`
`displaying the identity of one or more nodes using the actual
`cluster links.
`‘494 patent at 51:38-49. The features of claim 1 are tied to a database. In
`
`particular, claim 1 recites “a method of analyzing a database” and includes the
`
`limitation “an analysis of one or more indirect relationships in the database.”
`
`Moreover, the plain language of claim 1 requires that nodes and links represent
`
`
`
`
`
`
`7
`
`
`
`
`information in the database, where the nodes may represent textual objects and the
`
`
`
`links may represent direct and indirect “non-semantic” citation relationships.
`
`Jacobs Dec. at ¶ 24.
`
`Claim 5, which depends from claim 1, recites:
`
`The method of claim 1 wherein the step of generating the candidate
`cluster links comprises the step of:
`eliminating candidate cluster links, wherein the number of candidate
`cluster links are limited and the closest candidate cluster links are
`chosen over the remaining links.
`
`‘494 patent at 51:66 –52: 4 (emphasis added). As illustrated above, claim 5 adds
`
`three distinct limitations to the “generating” step of claim 1: (1) eliminating
`
`candidate cluster links; (2) the number of candidate cluster links are limited; and
`
`(3) the closest candidate cluster links are chosen over the remaining links.
`
`Furthermore, the requirement of claim 5 that “the number candidate clusters links
`
`are limited” is a distinct claim limitation that cannot be met simply by eliminating
`
`or failing to generate links without a specific numeral constraint. The ‘494 patent
`
`specification teaches that “[a]fter each iteration, the candidate cluster link set Ci
`
`may be pruned so that it contains only the top candidate cluster links (for example,
`
`the top 200).” ‘494 patent at 24:12-14; Jacobs Dec. at ¶ 28. Based on this teaching
`
`in the ‘494 patent specification, which is the only teaching in the ‘494 patent
`
`specification regarding limiting the number of generated candidate cluster links,
`
`
`
`8
`
`
`
`
`one of ordinary skill in the art that would understand that the claim language
`
`
`
`requiring the number of candidate cluster links to be limited is referring to a
`
`specific numeric limit. Jacobs Dec. at ¶ 28.
`
`The recited feature of “eliminating candidate cluster links” requires actually
`
`paring down generated candidate cluster links in which the eliminated links have to
`
`be generated prior to the elimination. Jacobs Dec. at ¶ 29. Moreover, since claim
`
`5 further limits the generating step, the claim 5 limitation “the closest … are
`
`chosen” must also correspond to links that are generated and meet the limitations
`
`of the generating step recited in claim 1 (e.g., including the “for the selected node”
`
`requirement). Jacobs Dec. at ¶ 30. Additionally, because the limitations of claim 5
`
`apply to the “generating” step and not the “deriving” step of claim 1, there must be
`
`a separate step of “deriving actual cluster links” that is different from the
`
`eliminating candidate cluster links step recited in claim 5. Jacobs Dec. at ¶ 31.
`
`Independent claim 14 recites:
`
`A method for representing the relationship between nodes using
`stored direct links, paths, and candidate cluster links, comprising the
`steps of:
`
`a) initializing a set of candidate cluster links;
`b) selecting the destination node of a path as the selected node to
`analyze;
`c) retrieving the set of direct links from the selected node to any other
`node in the database;
`
`
`
`9
`
`
`
`
`
`
`
`d) determining the weight of the path using the retrieved direct links;
`e) repeating steps b through d for each path; and
`f) storing the determined weights as candidate cluster links.
`‘494 patent at 52:51-64. As illustrated above, claim 14 is directed to the process of
`
`generating candidate cluster links by examining and weighting links and paths.
`
`One of ordinary skill in the art would understand that:
`
`(1) the “direct links” in this claim represent direct references (i.e.,
`citations) in the database; (2) a “path” is a chain of direct links, thus
`representing a chain of relationships made up of references or
`citations in the database; and (3) a weight is determined for each path
`using direct links between nodes (e.g., using weights of direct links or
`using a chain of direct links to assign a path weight). See discussion
`infra, ¶ 20 (claim construction). The person skilled in this art would
`further have understood that the “determined weight of the path”
`would be for a path that uses “direct links” and therefore would be an
`analysis of indirect relationships. This series of steps is repeated for
`each path and the result is stored as a candidate cluster link.
`
`Jacobs Dec. at ¶ 34.
`
`III. THE BOARD’S DECISION INSTITUTING INTER PARTES
`REVIEW
`
`The Institution Decision of February 3, 2014 instituted review on the
`
`following grounds:3
`
`
`3Institution Decision, paper 17, at 24 (February 3, 2014).
`
`
`
`10
`
`
`
`
`
`Ground 1 – anticipation of claims 14-16 based on Edward A. Fox,
`
`Extending the Boolean and Vector Space Models of Information Retrieval
`
`with P-Norm Queries and Multiple Concept Types, (Aug.1983) (Ph.D.
`
`dissertation, Cornell Univ. Dept. of Comp. Sci.) (“Fox Thesis”) (Ex. 1008).
`
`Ground 2 – anticipation of claims 1 and 5 based on Edward A. Fox, Some
`
`Considerations for Implementing the SMART Information Retrieval System
`
`under UNIX, (Sept. 1983) (Ph.D. dissertation, Cornell Univ. Dept. of Comp.
`
`Sci.) (“Fox SMART”) (Ex. 1005).
`
`Ground 3 – obviousness of claims 8, 10, 11, 35, and 40 based on Fox
`
`Thesis or Fox Smart in view of Gerard Salton, “Associative Document
`
`Retrieval Techniques Using Bibliographic Information,” JACM 10(4) (Oct.
`
`1963) (Ex. 1012) and Edward A. Fox, “Some Considerations for
`
`Implementing the SMART Information Retrieval System Under UNIX,”
`
`Department of Computer Science, Cornell University (Sept. 1983) (Ex.
`
`1006).
`
`Ground 4 – obviousness of claim40 based on Fox Thesis, in view of
`
`Edward A. Fox, et. al., Users, User Interfaces, and Objects: Envision, a
`
`Digital Library, 44 J. Am. Soc. Inf. Sci., no. 8 at 480-91 (Sept. 1993) (Ex.
`
`1006), and Gerard Salton, “Associative Document Retrieval Techniques
`
`Using Bibliographic Information,” JACM 10(4) (Oct. 1963)(Ex. 1012).
`
`11
`
`
`
`
`
`
`
`
`
`
`
`The cancellation of claims 8, 10, 11, 35, and 40 obviate Grounds 3 and 4.
`
`IV. PERSON OF ORDINARY SKILL IN THE ART
`The relevant field to the invention of the ‘494 patent is computerized search
`
`and information retrieval. Jacobs Dec. at ¶ 46. A person of ordinary skill in the
`
`relevant field as of May 17, 1996, would have had familiarity with information
`
`retrieval and at least a bachelor’s degree in one of computer science or electrical
`
`and computer engineering, or a comparable amount of combined education and
`
`equivalent industry experience in computer and information retrieval. Jacobs Dec
`
`at ¶ 47.
`
`V. CLAIM CONSTRUCTION
`A. Applicable Law
`The ‘494 patent expired on June 14, 2013. Thus, instead of applying the
`
`broadest reasonable interpretation to the claims, the Board explained that the claim
`
`interpretation standard applied in this inter partes review proceeding “is similar to
`
`that of a district court’s review” and that “[w]e are, therefore, guided by the
`
`principle that the words of a claim ‘are generally given their ordinary and
`
`customary meaning’ as understood by a person of ordinary skill in the art in
`
`question at the time of the invention.” Institution Decision at 10 (citations
`
`omitted). The Board elaborated further that, “[i]n determining the meaning of the
`
`disputed claim limitation, we look principally to the intrinsic evidence of record,
`
`
`
`12
`
`
`
`
`examining the claim language itself, the written description, and the prosecution
`
`
`
`history, if in evidence.” Id. (citations omitted).
`
`In applying the Phillips standard to the expired ‘494 patent claims, the Board
`
`must also construe the claims so as to sustain their validity, if possible. See
`
`Ex Parte Katz, Appeal 2008-005127, Reexamination Control Nos. 90/006,978 and
`
`90/007,074 (merged) (Mar. 15, 2010), following Ex Parte Papst-Motoren,
`
`1 U.S.P.Q.2d 1655 (B.P.A.I. 1986) as clarified by the subsequent decision in
`
`Phillips v. AWH Corp., 415 F.3d 1303, 1316, 75 USPQ2d 1324, 1329 (Fed. Cir.
`
`2005). As the Board held in Ex Parte Papst-Motoren, when it has the
`
`interpretation of claims of an expired patent before it:
`
`a policy of liberal construction may properly and should be applied.
`Such a policy favors a construction of a patent claim that will render
`it valid, i.e., a narrow construction, over a broad construction that
`would render it invalid.
`
`Ex parte Papst-Motoren, 1 USPQ2d 1655, 1986 WL 83328 (BPAI 1986)
`
`(emphasis added). This standard is also reflected in the MPEP. See MPEP § 2258.
`
`See also Phillips, 415 F.3d at 1327 (holding that “the maxim that claims should be
`
`construed to preserve their validity” applies where the claim is “ambiguous,” such
`
`that it is susceptible to more than “one reasonable construction”).
`
`Under this policy of liberal claim construction for expired patent claims, the
`
`Board has repeatedly overruled Examiners’ rejections of claims. See, e.g., Ex parte
`
`
`
`13
`
`
`
`
`Quickie, 2012 WL 2316828 at *4 (BPAI 2012) (internal citations omitted) (Board
`
`
`
`overruled the Examiner’s rejection of the claim, holding that, “[w]ith respect to
`
`construction, we first observe that the ‘160 patent has expired so as to preclude any
`
`amendments of the claim. As such, we apply a narrower construction to the claims
`
`instead of the broadest reasonable construction. Correspondingly, we are of the
`
`view that the preamble as well as the functional recitations in the body of the claim
`
`referencing ‘suture’ distinguishes the claimed invention over the device of
`
`Emery.”); Ex parte Tractus Medical, Inc., 2012 WL 759848 at *4 (BPAI 2012)
`
`(Board overruled the Examiner’s rejection of the Patent Owner’s claim and,
`
`relying on the fact that the patent was expired and a policy of liberal construction
`
`should be applied, found that the Examiner should have construed the patent to
`
`sustain it’s validity.); Ex parte Akihiko Toyoshima, 2012 WL 4718521 (BPAI
`
`2012) (reversing Examiner’s rejection of the claim after relying on Phillips and
`
`looking to the specification and dictionary meanings of the claim’s terms); Ex
`
`parte Suitco Surface, Inc., 2011 WL 3876561 (BPAI 2012) (reversing Examiner’s
`
`rejection of patent’s claims where in view of the policy considerations of favoring
`
`a construction of a claim to render it valid the Board adopted the claim
`
`construction advocated by the Patent Owner.).
`
`As explained in further detail below, by applying the Phillips claim
`
`construction standard to the disputed language of the challenged claims of the ‘494
`
`
`
`14
`
`
`
`
`patent and construing the disputed language to preserve the validity of the
`
`
`
`challenged claims, the Board should use the following interpretations.
`
`Board Claim Construction
`
`B.
`The Board’s claim construction set forth on pages 9-13 of the Institution
`
`Decision construed the terms “cluster links,” “candidate cluster links,” “actual
`
`cluster links,” “wherein the step of generating comprises an analysis of one or
`
`more indirect relationships in the database,” “selecting a node of analysis,” and
`
`“displaying” In this proceeding, for the purposes of responding to the Petition and
`
`establishing that the challenged claims of the ‘494 patent are patentable over the
`
`relied-upon references, Patent Owner is using the Board’s constructions. These
`
`constructions are followed for the purpose of evaluating the patentability of the
`
`claims of the ‘494 patent.4 Jacobs Dec. at ¶ 40.
`
`
`4 Patent owner disagrees with the board’s construction of “cluster link.” A Cluster
`link is: a relationship between two nodes based on a statistical analysis of multiple
`relationships between two nodes in a database. For example, two nodes, both
`directly linked to the same intermediate nodes, may be indirectly linked through
`many paths and therefore have a Cluster link between them. See District Court’s
`Claim Construction Order in Google et al at 16 (Exhibit 2021). Patent owner’s use
`of the board’s construction in this document is not acquiescence to this
`construction. Patent Owner expressly reserves the right to construe, and is not
`estopped from construing, the claim language in accordance with above. Cluster
`links as properly construed are arranged in the steps of the claim.
`
`
`
`15
`
`
`
`
`VI. FOX SMART DOES NOT ANTICIPATE CLAIMS 1 AND 5
`(GROUND 2)
`
`
`
`Claims 1 and 5 are not taught, and would not have been suggested by Fox
`
`SMART. Particularly, the Board’s construction of the terms “an analysis of one or
`
`more indirect relationships in the database” appears to include a technical error,
`
`which when properly construed, is not disclosed or suggested in Fox SMART. See
`
`discussion infra at Section VI(B)(1)-(3). Moreover, Fox SMART does not disclose
`
`or suggest the claimed features of “generating candidate cluster links for the
`
`selected node,” and “deriving actual cluster links from the generated cluster links.”
`
`For example, the clustering process of Fox SMART relied upon by the Petitioners
`
`does not derive any candidate cluster links from which a subset are derived as
`
`actual cluster links. See discussion infra at Section VI(C)-(D).
`
`Additionally, claim 1 requires a specific arrangement of “selecting a node
`
`for analysis,” “generating candidate cluster links for the selected node,” “deriving
`
`actual cluster links from the candidate cluster links,” and “displaying the identity
`
`of one or more nodes using the actual cluster links.” The Petitioners rely on
`
`disparate, unrelated portions of Fox SMART which do not anticipate claim 1 as
`
`arranged. See discussion infra at Section VI (G).
`
`Summary of Fox SMART
`
`A.
`A better understanding of the deficiencies of Fox SMART starts with an
`
`overview of Fox SMART.
`
`
`
`16
`
`
`
`
`
`
`
`Fox SMART describes a set of tools used to build and test information
`
`retrieval systems and organizing documents in a tree comprising a hierarchy of
`
`clusters. Fox SMART at 12-15 and 44. The SMART project was a multi-decade
`
`compendium of research programs conducted by Gerard Salton and his students
`
`and associates. Fox SMART at 3-4; Fox Declaration (Ex. 1009) at ¶¶ 39-42 and
`
`57. Documents in SMART are indexed using the vector space model, in which
`
`each document is represented by a weighted vector of “concepts.” Fox SMART at
`
`27-29. Fox SMART describes including bibliographic “concept types” (or
`
`“ctypes”) in an extended vector, a “vector with multiple subvectors,” which
`
`comprises the SMART index for each document. Id. at 29.
`
`Fox SMART teaches that bibliographic data “such as that from processing
`
`co-citations or other bibliographic information” is “combined with the result of
`
`previous indexing steps, and the end product, a vector with multiple subvectors, is
`
`stored in the appropriate internal form.” Id. at 27. One of ordinary skill would
`
`have understood that this “internal form,” which includes potentially all of the
`
`relevant subvectors, is what is used for clustering, searching, etc. Jacobs Dec. at ¶¶
`
`63-65.
`
`The initial placement of a document in a tree includes a “depth first,” search
`
`starting with the root node, and descends the tree to a “twig,” i.e., a cluster node
`
`