Ex Parte Ding et al

Patent Trial and Appeal Board

Jun 27, 2014

11467183 (P.T.A.B. Jun. 27, 2014)

UNITED STATES PATENT AND TRADEMARK OFFICE
____________

BEFORE THE PATENT TRIAL AND APPEAL BOARD
____________

Ex parte MIN DING, CHI QI, and XIAO BIN YANG
____________

Appeal 2012-002484
Application 11/467,183
Technology Center 2600
____________

Before JOHN A. JEFFERY, STEVEN D. A. MCCARTHY, and JEREMY J.
CURCURI, Administrative Patent Judges.

JEFFERY, Administrative Patent Judge.

DECISION ON APPEAL
Appellants appeal under 35 U.S.C. § 134(a) from the Examiner’s
decision to reject claims 1–12. We have jurisdiction under 35 U.S.C. § 6(b).
We reverse.

STATEMENT OF THE CASE
Appellants’ invention enables test scripts generated by automated test
tools to be played back correctly in locales of different test languages by
using a synonymy dictionary. See generally Abstract; Spec. 1, 6–9; Figs. 1,
4-5. Claim 1 is illustrative:
1. An automated computer-implemented test method,
comprising:
storing, in a synonymy dictionary of a computer-
implemented test system, a group of synonyms including a
Appeal 2012-002484
Application 11/686,914

2
plurality of synonyms corresponding to a plurality of different
languages for a property value of a property of an object of a
globalized software product to be tested in different test locales;
performing a pre-recorded test script on the software
product;
identifying the object in the software product, including
the property value;
comparing the property value of the property of the
object to a corresponding property value in the test script and
the group of synonyms in the synonymy dictionary
corresponding to the property value; and
determining if the object passes the test based on whether
the property value matches either the property value in the test
script or one of the synonyms in the synonymy dictionary
corresponding to the property.

THE REJECTION
The Examiner rejected claims 1–12 under 35 U.S.C. § 103(a) as
obvious over Raman (US 5,579,223; issued Nov. 26, 1996) and Yang (US
6,530,039 B1; issued Mar. 4, 2003). Ans. 5–10.1

CONTENTIONS
Regarding claim 1, the Examiner finds that Raman (1) stores a group
of synonyms corresponding to plural different languages for an object
property value in a synonymy dictionary in connection with, and (2)
identifies an object in a software product, including the property value. Ans.
5–6, 11–16. The Examiner acknowledges that Raman lacks (1) a globalized
software product to be tested in different test locales; (2) performing a pre-

1 Throughout this opinion, we refer to (1) the Appeal Brief filed May 18,
2011 (“App. Br.”); (2) the Examiner’s Answer mailed August 5, 2011
(“Ans.”); and (3) the Reply Brief filed October 5, 2011 (“Reply Br.”).
Appeal 2012-002484
Application 11/686,914

3
recorded test script on the product; and (3) the recited comparing and
determining steps, but cites Yang as teaching these features in concluding
that the claim would have been obvious. Ans. 6–8, 16–33.
Appellants argue that Raman’s block comparison does not teach or
suggest the recited string-based synonymy dictionary, let alone identify an
associated object in a globalized software product, including the object’s
property value. App. Br. 8–9; Reply Br. 3.2 Appellants also contend that the
Examiner’s reliance on Yang’s translating test strings in test script
commands is likewise flawed. According to Appellants, Yang does not
compare an object’s property value to a corresponding value in a test script
and a group of synonyms in the synonymy dictionary, let alone determine if
an object passes the test based on whether a match occurs from the recited
comparison. App. Br. 9–10; Reply Br. 3–4. Appellants add that not only is
there no motivation to combine Raman and Yang as proposed, modifying
Raman as proposed would impermissibly change Raman’s principle of
operation. App. Br. 10–11; Reply Br. 1–2.

ISSUE
Under § 103, has the Examiner erred by finding that Raman and Yang
collectively would have taught or suggested:
(1) storing, in a synonymy dictionary, a group of synonyms including
plural synonyms corresponding to plural different languages for a property
value of a property of an object of a globalized software product to be tested

2 Although the Reply Brief is not paginated (unlike the Appeal Brief), we
nonetheless refer to the Reply Brief’s pages in the order that they appear in
the record.
Appeal 2012-002484
Application 11/686,914

4
in different test locales; (2) identifying the object in the product, including
the property value; (3) comparing the property value to a corresponding
property value in the test script and the group of synonyms in the dictionary
corresponding to the property value; and (4) determining if the object passes
the test based on whether the property value matches either the property
value in the test script or one of the synonyms in the dictionary as recited in
claim 1?

ANALYSIS
We do not sustain the Examiner’s rejection of representative claim 1
principally on the Examiner’s reliance on Yang to cure Raman’s
acknowledged deficiencies.
That said, we see no error in the Examiner’s reliance on Raman’s
language translation for at least suggesting some sort of synonymy
dictionary to facilitate not only the translation,3 but also determining whether
corresponding event-based responses are equivalent. See Ans. 6, 11–14
(citing Raman, col. 6, ll. 50–63 (noting that a text event such as “good” in
English may be equivalent to a text event such as “guten” in German); Figs.
1A–B); see also Raman, Fig. 5D (step 546). We reach this conclusion even
assuming, without deciding, that the recited synonymy dictionary is limited
to the cited description in the Specification, namely that strings with the
same meaning are associated and stored together within one group, one
record, or “in any other manner.” App. Br. 9 (citing Spec. 6, ¶ 3).

3 Although Raman’s preferred embodiment does not automatically translate
text from one language to another, but rather marks changes made to that
text for subsequent human translation, Raman nonetheless teaches that
automatic translators can be used. Raman, col. 3, ll. 43–48.
Appeal 2012-002484
Application 11/686,914

5
Appellants’ contention that Raman does not compare strings despite
comparing objects (App. Br. 9) is unavailing. In short, some sort of string-
based comparison would be used for both Raman’s language translation and
the merge utility’s corresponding response equivalency determination to
produce the modified translated file—both of which involve objects and
their associated string-based property values as the Examiner indicates. See
Ans. 17 (“[A] ‘property value’ is merely a string within an object”); Ans. 14
(finding that Raman compares objects to find matches for textual
translations as well as attribute translations within a script). Appellants’
contention that Raman’s English-to-Spanish conversion is a single
translation and, therefore, does not teach plural synonyms corresponding to
different languages (Reply Br. 3) is likewise unavailing. Notably, Raman
teaches that a first natural language can be translated into one or more such
languages (Raman, col. 2, ll. 9–11): a passage at least suggesting that
synonyms associated with these plural different languages would be used for
these translations.
Nevertheless, the Examiner’s reliance on Yang is problematic.
Yang’s system translates a test script at run time using dictionaries to allow a
testing program to test an application program according to its particular
language. Yang, Abstract. In the embodiment of Figure 4, Yang’s system
interrupts a test script, and attempts to match the command string to the
original string in step 420 if the appropriate option is selected. Yang, col. 6,
ll. 18–25; col. 8, ll. 8–12; Figs. 4, 7 (“Using Original Strings” option). If
this option is not selected, then the system tries to find a match in internal
and external dictionaries. Yang, col. 8, ll. 8–12; Fig. 4 (steps 425, 430). If
Appeal 2012-002484
Application 11/686,914

6
any of these attempts find a match, the command string is translated before
resuming the test script. Yang, col. 8, ll. 15–29; Fig. 4 (steps 445, 450).
Based on this functionality, we disagree with the Examiner that Yang
teaches the recited property value comparison. The Examiner equates an
object’s property value to a string within an object, such as the exemplary
“OK” function described in Yang. Ans. 17; see also Yang, col. 6, ll. 19–25.
Even assuming, without deciding, that this string-based property value is
compared to a corresponding value in the test script at step 420, and can also
be compared to corresponding synonyms in a dictionary in step 425 as the
Examiner seemingly suggests, this property-value comparison is with
respect to objects in the test script—not objects in the software product to be
tested as claimed.
As Appellants indicate, the object being compared in claim 1 is an
object of a globalized software product to be tested and, notably, is
distinguished from the test script in the claim. App. Br. 9–10.4 The
Examiner’s reliance on Yang’s comparisons that are limited to test script
object property values therefore ignores the fundamental aspect of the
claimed invention that compares property values between the test script and
the software product to be tested, and determines whether the object passes
the test based on this comparison. Nor does Raman cure that deficiency. So

4 Although Appellants argue that the objects of Gutta are different from
those in Pyo (App. Br. 10; emphases added), these references (along with
Guttan) were not cited in the rejection and are, therefore, inapposite. Accord
Ans. 17 (noting this inconsistency). Nevertheless, we presume that this
inconsistency is a harmless typographical error, and that the argument was
intended to be directed to the combination of references actually cited in the
rejection.
Appeal 2012-002484
Application 11/686,914

7
even if the references were combinable as the Examiner proposes, they still
do not teach or suggest the recited comparison.
Therefore, we are persuaded that the Examiner erred in rejecting (1)
independent claim 1; (2) independent claims 5 and 10 that recite
commensurate limitations; and (3) the dependent claims for similar reasons.
Because this issue is dispositive regarding our reversing the rejection of
these claims, we need not address Appellants’ other arguments.

CONCLUSION
The Examiner erred in rejecting claims 1–12 under § 103.

DECISION
The Examiner’s decision rejecting claims 1–12 is reversed.

REVERSED

mls