Ex Parte Chengalvarayan et alDownload PDFPatent Trial and Appeal BoardJun 19, 201311336081 (P.T.A.B. Jun. 19, 2013) Copy Citation UNITED STATES PATENT AND TRADEMARKOFFICE UNITED STATES DEPARTMENT OF COMMERCE United States Patent and Trademark Office Address: COMMISSIONER FOR PATENTS P.O. Box 1450 Alexandria, Virginia 22313-1450 www.uspto.gov APPLICATION NO. FILING DATE FIRST NAMED INVENTOR ATTORNEY DOCKET NO. CONFIRMATION NO. 11/336,081 01/20/2006 Rathinavelu Chengalvarayan GP-307550 (2760/274) 4597 60770 7590 06/20/2013 General Motors Corporation c/o REISING ETHINGTON P.C. P.O. BOX 4390 TROY, MI 48099-4390 EXAMINER CHAWAN, VIJAY B ART UNIT PAPER NUMBER 2658 MAIL DATE DELIVERY MODE 06/20/2013 PAPER Please find below and/or attached an Office communication concerning this application or proceeding. The time period for reply, if any, is set in the attached communication. PTOL-90A (Rev. 04/07) UNITED STATES PATENT AND TRADEMARK OFFICE ____________ BEFORE THE PATENT TRIAL AND APPEAL BOARD ____________ Ex parte RATHINAVELU CHENGALVARAYAN and JOHN J. CORREIA ____________ Appeal 2011-002359 Application 11/336,081 Technology Center 2600 ____________ Before JEFFREY S. SMITH, KALYAN K. DESHPANDE, and MATTHEW R. CLEMENTS, Administrative Patent Judges. CLEMENTS, Administrative Patent Judge. Dissenting Opinion filed by SMITH, Administrative Patent Judge. DECISION ON APPEAL Appellants appeal under 35 U.S.C. § 134(a) from the Examiner’s final rejection of claims 1-8 and 19-23. Claims 9-18 are cancelled. We have jurisdiction under 35 U.S.C. § 6(b). We REVERSE. Appeal 2011-002359 Application 11/336,081 2 STATEMENT OF THE CASE Appellants’ invention generally relates to a speech recognition system using dynamic nametag scoring. See generally Abstract. Claim 19 is illustrative: 19. A method for determining a plurality of confusion tables comprising: receiving a plurality of test utterances in a speech recognizer including acoustic models; generating via the speech recognizer, a plurality of component sequences based on the test utterances and at least one circumstantial condition; determining via the speech recognizer, a confusion matrix based on the component sequences and the test utterances; and generating a confusion table based on the confusion matrix but sorted by phoneme misrecognition, and stored in at least one of a mobile vehicle telematics unit or a remote server. REFERENCES The prior art relied upon by the Examiner in rejecting the claims on appeal is: Goldberg US 6,223,158 B1 Apr. 24, 2001 Bushey US 7,110,949 B2 Sept. 19, 2006 (filed Sept. 13, 2004) Endo US 7,292,978 B2 Nov. 6, 2007 (filed Dec. 4, 2003) Ng et al., Recent Advances In Cantonese Speech Recognition, Proceedings, 12th European Signal Processing Conference, Sept. 2004, pp. 341-344. Appeal 2011-002359 Application 11/336,081 3 THE REJECTIONS 1. Claim 19 stands rejected under 35 U.S.C. § 102(b) as anticipated by Goldberg. Ans. 4-5.1 2. Claims 1-3, 6-8, and 20-23 stand rejected under 35 U.S.C. § 103(a) as unpatentable over Goldberg, Ng, and Endo. 2 Ans. 5-10. 3. Claims 4-5 stand rejected under 35 U.S.C. § 103(a) as unpatentable over Goldberg, Ng, Endo, and Bushey. Ans. 10-11. THE ANTICIPATION REJECTION OVER GOLDBERG The Examiner finds that Goldberg discloses every recited element of representative claim 19. Ans. 4-5. Appellants argue, inter alia, that Goldberg’s confusion table is sorted alphanumerically and is therefore not “sorted by phoneme misrecognition.” App. Br. 7-8. ISSUE Under § 102, has the Examiner erred in rejecting claim 19 by finding that Goldberg discloses “a confusion table based on the confusion matrix but sorted by phoneme misrecognition?” 1 Throughout this opinion, we refer to (1) the Appeal Brief filed June 22, 2010 (“App. Br.”); (2) the Examiner’s Answer mailed Sept. 10, 2010 (“Ans.”); and (3) the Reply Brief filed Nov. 10, 2010 (“Reply Br.”). 2 Appellants identify claims 1-3, 6-10, and 20-23. App. Br. 6, 8. However, claims 9 and 10 are cancelled. App. Br. 16; accord Ans. 2. Appeal 2011-002359 Application 11/336,081 4 ANALYSIS We agree with Appellants’ argument that the Examiner’s construction of “sorted by phoneme misrecognition” to include sorting by alphabetical order is unreasonably broad. While the Examiner has not provided a clear construction of “sorted by phoneme misrecognition,” the Examiner found that Goldberg describes a “matrix [that] is sorted so that the axes are in alphabetical order, wherein the populated fields represent the probability of a match, as well as a mismatch, between characters and spoken characters.” Ans. 12. As such, the Examiner appears to construe “a confusion table…sorted by phoneme misrecognition” to include arranging the table such that the rows and columns, i.e. the axes, are in alphabetical order. However, we find such a construction to be neither reasonable nor consistent with the Specification. As described in the Specification, “[a] confusion table is sorted” based on the numerical values within the cells of the table such that those values are in order of increasing frequency of misrecognition or of decreasing frequency of misrecognition. Reply Br. 3; Spec. ¶ 0025. Although this description is not to be read in to the claims, the claims must be read in light of this description. Furthermore, the plain and ordinary meaning of sorting a table based on a value, as understood by a person with ordinary skill in the art, includes sorting based on the values within the table rather than the column or row headings or labels. Accordingly, we do not agree with the Examiner that Goldberg describes “a confusion table based on the confusion matrix but sorted by phoneme misrecognition” and therefore we do not sustain the Examiner’s anticipation rejection of claim 19. Appeal 2011-002359 Application 11/336,081 5 We have decided the appeal of the Examiner’s anticipation rejection before us. However, we leave it to the Examiner to determine whether it would have been obvious to a person with ordinary skill in the art to modify Goldberg’s confusion table, which includes phoneme misrecognition values, by sorting the table based on the phoneme misrecognition in the table and whether such a modification would have yielded anything more than predictable results. THE OBVIOUSNESS REJECTION OVER GOLDBERG, NG, AND ENDO We are persuaded by Appellants’ argument that Goldberg does not teach a confusion table “sorted by phoneme misrecognition” (App. Br. 8-9; Reply Br. 1-4) for the reasons discussed above with respect to claim 19. We therefore do not sustain the Examiner’s rejection of (1) independent claim 1; (2) independent claim 20 which recites a commensurate limitation; and (3) dependent claims 2-8 and 21-23 not separately argued with particularity. Because this issue is dispositive, we need not reach Appellants’ other arguments. THE OBVIOUSNESS REJECTION OVER GOLDBERG, NG, ENDO, AND BUSHEY Since the Examiner has not shown that Bushey cures the deficiencies noted above, we will not sustain the Examiner’s obviousness rejection of dependent claims 4 and 5. Ans. 10-11. Appeal 2011-002359 Application 11/336,081 6 CONCLUSION The Examiner erred in rejecting claim 19 under § 102 and claims 1-8 and 20-23 under § 103. ORDER The Examiner’s decision rejecting claims 1-8 and 19-23 is reversed. REVERSED rwk Appeal 2011-002359 Application 11/336,081 1 SMITH, Administrative Patent Judge, dissenting. Principal Issue Does the confusion matrix sorted by the alphabetical order of misrecognized phonemes shown in Figure 2 of Goldberg describe a confusion table sorted by phoneme misrecognition as recited in claim 1? Standard for Claim Interpretation The claims measure the invention.3 During prosecution before the USPTO, claims are to be given their broadest reasonable interpretation, and the scope of a claim cannot be narrowed by reading disclosed limitations into the claim.4 “Giving claims their broadest reasonable construction ‘serves the public interest by reducing the possibility that claims, finally allowed, will be given broader scope than is justified.’”5 Our reviewing court has repeatedly warned against confining the claims to specific embodiments described in the specification.6 Though understanding the claim language may be aided by explanations contained in the written description, it is important not to import into a claim limitations that are not part of the claim. For example, a particular embodiment appearing in the written description may not be read into a claim when the claim language is broader than the embodiment.7 3 See SRI Int’l v. Matsushita Elec. Corp., 775 F.2d 1107, 1121 (Fed. Cir. 1985) (en banc). 4 See In re Morris, 127 F.3d 1048, 1054 (Fed. Cir. 1997). 5 In re Amer. Acad. of Sci. Tech Ctr., 367 F.3d 1359, 1364 (Fed. Cir. 2004). 6 Phillips v. AWH Corp., 415 F.3d 1303, 1323 (Fed. Cir. 2005) (en banc). 7 Superguide Corp. v. DirecTV Enterprises, Inc., 358 F.3d 870, 875 (Fed. Cir. 2004). Appeal 2011-002359 Application 11/336,081 2 Interpretation of confusion table sorted by phoneme misrecognition Appellants’ Specification does not explicitly define “phoneme misrecognition.” In fact, Appellants’ Specification does not mention the term “phoneme misrecognition” at all. Paragraph 25 of Appellants’ Specification discloses an example of phonemes as spoken letters such as m and n, where n is a misrecognized phoneme, or “phoneme misrecognition,” of n. Thus, the term “phoneme misrecognition,” when read in light of Appellants’ Specification, encompasses misrecognized phonemes. Appellants’ Specification also does not explicitly define “sorted.” Appellants’ Specification merely distinguishes between a confusion matrix as unsorted and a confusion table as sorted. Spec. ¶ 25. Appellants’ Specification gives an example of sorting a matrix by the frequencies of “phoneme misrecognition.” However, the unsorted matrix may be alphabetically unsorted. For instance, the unsorted matrix disclosed by paragraph 25 places the letter m “a positional distance from” the letter n, rather than in alphabetical order next to the letter n.8 When the alphabetically unsorted matrix is sorted by the alphabetical order of “phoneme misrecognition,” the result is a “confusion table … sorted by phoneme misrecognition” within the meaning of claim 1. Thus, the scope of “confusion table … sorted by phoneme misrecognition,” when read in light of Appellants’ Specification, encompasses a confusion matrix sorted by the alphabetical order of misrecognized phonemes. 8 “A confusion table is unsorted … ‘m’ is a positional distance from ‘n’ in the confusion matrix”. Appeal 2011-002359 Application 11/336,081 3 Analysis of Appellants’ arguments Appellants contend that the characters of Goldberg are not sorted by phoneme misrecognition. According to Appellants, if Goldberg’s Figure 2 taught “a confusion table … sorted by phoneme misrecognition,” the characters would be sorted in ascending or descending order of recognition probabilities, rather than in alphabetical order. App. Br. 7; Reply Br. 3. The Examiner finds that claim 1 does not limit the scope of “sorted by phoneme misrecognition” to a table sorted in ascending or descending order of recognition probabilities. Ans. 12. I agree with the Examiner. Appellants’ argument is based on the premise that the meaning of “phoneme misrecognition” is limited to recognition probabilities. However, Appellants have not provided persuasive evidence to show that the meaning of “phoneme misrecognition” is limited to, or even includes, recognition probabilities. Rather, Appellants’ Specification discloses examples of phonemes as spoken letters such as m and n, where m is a misrecognized phoneme, or “phoneme misrecognition,” of n. The scope of “phoneme misrecognition,” when read in light of Appellants’ Specification, encompasses misrecognized phonemes described by Goldberg. That is, misrecognized letters shown in Figure 2 of Goldberg describe “phoneme misrecognition.” Thus, the alphabetically sorted table of Goldberg is “sorted by phoneme misrecognition” within the meaning of claim 1. Appellants contend that paragraph 25 of their Specification uses the term “sorted” consistent with the ordinary meaning. Reply Br. 3. Paragraph 25 gives an example of an unsorted table sorted based on the frequency of misrecognized phonemes. However, this non-limiting example from Appellants’ Specification does not provide a definition of “sorted” that Appeal 2011-002359 Application 11/336,081 4 excludes the ordinary meaning of “sorted” by alphabetical order of misrecognized phonemes. “Absent an express definition in their specification, the fact that appellants can point to definitions or usages that conform to their interpretation does not make the PTO’s definition unreasonable when the PTO can point to other sources that support its interpretation.” In re Morris, 127 F.3d at 1056. Conclusion Appellants’ Specification only gives a non-limiting example of “sorted,” does not explicitly define “phoneme misrecognition,” and does not exclude the interpretation that the scope of “a confusion table … sorted by phoneme misrecognition” encompasses the table sorted by alphabetical order of misrecognized phonemes shown in Figure 2 of Goldberg. Therefore, I would affirm the Examiner’s rejection. Copy with citationCopy as parenthetical citation