Ex Parte Hakkani-Tur et al

Patent Trial and Appeal Board

May 8, 2018

14846486 (P.T.A.B. May. 8, 2018)

UNITED STA TES p A TENT AND TRADEMARK OFFICE
APPLICATION NO. FILING DATE
14/846,486 09/04/2015
69316 7590 05/10/2018
MICROSOFT CORPORATION
ONE MICROSOFT WAY
REDMOND, WA 98052
FIRST NAMED INVENTOR
Dilek Hakkani-Tur
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
ATTORNEY DOCKET NO. CONFIRMATION NO.
357730.01 4696
EXAMINER
OPSASNICK, MICHAEL N
ART UNIT PAPER NUMBER
2658
NOTIFICATION DATE DELIVERY MODE
05/10/2018 ELECTRONIC
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.
Notice of the Office communication was sent electronically on above-indicated "Notification Date" to the
following e-mail address(es):
usdocket@microsoft.com
chriochs@microsoft.com
PTOL-90A (Rev. 04/07)
UNITED STATES PATENT AND TRADEMARK OFFICE
BEFORE THE PATENT TRIAL AND APPEAL BOARD
Ex parte DILEK HAKKANI-TUR, YUN-CHENG JU,
GEOFFREY G. ZWEIG, AND GOKHAN TUR1
Appeal2017-010766
Application 14/846,486
Technology Center 2600
Before ALLEN R. MacDONALD, ROBERT E. NAPPI, and JAMES W.
DEJMEK, Administrative Patent Judges.
NAPPI, Administrative Patent Judge.
DECISION ON APPEAL
This is a decision on appeal under 35 U.S.C. § 134(a) from the
Examiner's final rejection of claims 1 through 20. We have jurisdiction
under 35 U.S.C. § 6(b ).
We REVERSE.
1 According to Appellants, the real party in interest is Microsoft Technology
Licensing, LLC. App. Br. 3.
Appeal2017-010766
Application 14/846,486
INVENTION
Appellants' disclosed invention is directed to a system that trains a
spoken language understanding (SLU) classifier. The system makes use of a
corpus of user utterances, semantically parses utterances and produces a
parse graph representing all user utterances. The user utterances graph is
clustered into intent-wise homogeneous groups of user utterances which are
then used to train the SLU classifier. Abstract. Claim 1 is representative of
the invention and reproduced below.
1. A system for training a spoken language understanding
(SLU) classifier, comprising:
one or more computing devices, said computing devices
being in communication with each other via a computer
network whenever there is a plurality of
computing devices; and
a computer program having program modules executable
by the one or more computing devices, the one or more
computing devices being directed by the program modules of
the computer program to,
receive a corpus of user utterances,
for each of the user utterances in the corpus,
semantically parse the user utterance, and
represent the result of said semantic parsing as a rooted
semantic parse graph,
combine the parse graphs representing all of the user
utterances in the corpus into a single corpus graph that
represents the semantic parses of the entire corpus and
comprises a root node that is common to the parse graph
representing each of the user utterances in the corpus,
cluster the user utterances in the corpus into intent-wise
homogeneous groups of user utterances, said clustering
comprising finding subgraphs in the corpus graph that represent
different groups of user utterances, each of said different groups
having a similar user intent, each of the subgraphs being more
specific than the root node alone and more general than the full
semantic parses of the individual user utterances,
2
Appeal2017-010766
Application 14/846,486
use the intent-wise homogeneous groups of user
utterances to train the SLU classifier, and
output the trained SLU classifier.
REJECTION AT ISSUE2
The Examiner has rejected claims 1 through 20 under 35 U.S.C. § 101
for being directed to patent-ineligible subject matter. Answer 2--4.
PRINCIPLES OF LAW
Patent-eligible subject matter is defined in 35 U.S.C. § 101 of the
Patent Act, which recites:
Whoever invents or discovers any new and useful process,
machine, manufacture, or composition of matter, or any new
and useful improvement thereof, may obtain a patent therefor,
subject to the conditions and requirements of this title.
There are, however, three judicially created exceptions to the broad
categories of patent-eligible subject matter in 35 U.S.C. § 101: laws of
nature, natural phenomena, and abstract ideas. Alice Corp. Pty. Ltd. v. CLS
Bank Int'!, 134 S. Ct. 2347, 2354 (2014); Mayo Collaborative Servs. v.
Prometheus Labs., Inc., 566 U.S. 66, 70 (2012). Although an abstract idea
itself is patent ineligible, an application of the abstract idea may be patent
eligible. Alice, 134 S. Ct. at 2355. Thus, we must consider "the elements of
each claim both individually and 'as an ordered combination' to determine
whether the additional elements 'transform the nature of the claim' into a
patent-eligible application." Id. (citing Mayo, 566 U.S. at 78-80). The
2 Throughout this Decision we refer to the Appeal Brief filed May 19, 201 7,
Reply Brief filed August 14, 2017, Final Office Action mailed January 5,
2017, and the Examiner's Answer mailed July 28, 2017.
3
Appeal2017-010766
Application 14/846,486
claim must contain elements or a combination of elements that are
"'sufficient to ensure that the patent in practice amounts to significantly
more than a patent upon the [ abstract idea] itself."' Id. ( citing Mayo, 566
U.S. at 72-73).
The Supreme Court sets forth a two-part "framework for
distinguishing patents that claim laws of nature, natural phenomena, and
abstract ideas from those that claim patent-eligible applications of those
concepts." Id. at 2355.First, we determine whether the claims at issue are
directed to one of those patent-ineligible concepts. Mayo, 566 U.S. at 76-
77. If so, we then ask, "[ w ]hat else is there in the claims before us?" Id. at
77-78. To answer that question, we consider the elements of each claim
both individually and "as an ordered combination" to determine whether the
additional elements "transform the nature of the claim" into a patent-eligible
application. Id.. The court has described step two of this analysis as a
search for an "'inventive concept"'-i.e., an element or combination of
elements that is "sufficient to ensure that the patent in practice amounts to
significantly more than a patent upon the [ineligible concept] itself." Id. at
71-73.
Claims directed to "limited rules, in a process specifically designed to
achieve an improved technical result in conventional industry practice" are
not directed to an abstract idea. McRO, Inc. v. Bandai Namco Games Am.
Inc., 837 F.3d 1299, 1316 (Fed. Cir. 2016).
4
Appeal2017-010766
Application 14/846,486
ANALYSIS
We have reviewed Appellants' arguments in the Briefs, the
Examiner's rejections, and the Examiner's response to Appellants'
arguments. Appellants' arguments have persuaded us of error in the
Examiner's conclusion that the claims are directed to patent-ineligible
subject matter.
Appellants argue on pages 10 through 17 of the Appeal Brief and
pages 2 through 7 of the Reply Brief that the precepts of the McRO analysis
demonstrate that independent claim 1 is directed to patent-eligible subject
matter. Specifically, Appellants argue that claim 1 includes an utterance
intent clustering limitation that amounts to significantly more than an
abstract idea. 3 Appellants argue:
In the McRO decision the court explained that" ... We
have previously cautioned that courts 'must be careful to avoid
oversimplifying the claims' by looking at them generally and
failing to account for the specific requirements of the claims ...
In addition, in the McRO decision the court further explained
3 Appellants refer to the limitations "for each of the user utterances in ... [a]
corpus [of user utterances], semantically parse the user utterance, and
represent the result of said semantic parsing as a rooted semantic parse
graph, combine the parse graphs representing all of the user utterances in the
corpus into a single corpus graph that represents the semantic parses of the
entire corpus and comprises a root node that is common to the parse graph
representing each of the user utterances in the corpus, cluster the user
utterances in the corpus into intent-wise homogeneous groups of user
utterances, said clustering comprising finding subgraphs in the corpus graph
that represent different groups of user utterances, each of said different
groups having a similar user intent, each of the subgraphs being more
specific than the root node alone and more general than the full semantic
parses of the individual user utterances" as the utterance intent clustering
limitation. App. Br. 10-11. For the sake of brevity, we adopt this
shorthand.
5
Appeal2017-010766
Application 14/846,486
that" ... Whether at step one or step two of the Alice test, in
determining the patentability of a method, a court must look to
the claims as an ordered combination, without ignoring the
requirements of the individual steps ... " (emphasis added). In
the McRO decision the court further clarified these issues by
explaining that" ... The preemption concern arises when the
claims are not directed to a specific invention and instead
improperly monopolize 'the basic tools of scientific and
technological work"', "[ w ]e therefore look to whether the
claims ... focus on a specific means or method that improves
the relevant technology ... " (emphasis added).
App. Br. 11 ( emphases omitted). Appellants assert that similar to McRO,
appealed claim 1 focuses on a specific improvement to a spoken language
understanding subsystem that is not used by other spoken language
understanding systems. App. Br. 12. Further, Appellants argue that the
Specification describes how the utterance-clustering feature of claim 1 is
used to determine user intent in a manner not found using conventional
spoken dialog computing systems. App. Br. 15.
In response to Appellants' arguments, the Examiner states:
The fact pattern established in the independent claim 1, does
not mirror the McRO v Bandai et al, wherein a "set of rules"
previously performed by human animators (non-calculation in
nature) were now automated via a computer (McRO, pp 24--25,
and in contradistinction, Flook, Bilski, and Alice performing
computer automated processes in the same way as previous
methods, as noted in McRO).
Answer 4 ( emphasis omitted). Further, the Examiner finds that the claimed
clustering is a mathematical calculation, where the improvement is to the
calculation itself and not an improvement to the computer or other
technology. Answer 5. The Examiner responds to Appellants' assertion that
the claims are an improvement to the SLU classifiers by finding that SLU
6
Appeal2017-010766
Application 14/846,486
classifiers are themselves mathematical classifiers, and that the limitations
directed to training a SLU classifier are merely an intended use and do not
amount to significantly more than the abstract idea. Answer 3, 5 (citing
paragraphs 21, 27 through 29, and Figure 4 as showing that SLUs are
mathematical calculations).
We disagree with the Examiner. Initially, we note that paragraphs 21,
27 through 29 and Figure 4, cited by the Examiner to show that the SLU is a
mathematical algorithm are not discussing the SLU, but rather the method of
developing the parsing graph used to train the SLU. Thus, although the cited
paragraphs may describe mathematical algorithms ( set theory algorithms),
they do not support the Examiner's finding that SLUs are mathematical
algorithms. Additionally, the limitation that the intent-wise homogeneous
groups are used to train the SLU classifier and the limitation directed to
outputting a trained SLU classifier, are more than just field of use
limitations.
Further, we disagree with the Examiner's statement that claim 1, does
not mirror the reasoning applied in McRO. In McRO, the court found that a
system of automated facial animation through the use of rules, rather than
artists setting weights, to automate tasks that humans perform is directed to
patent-eligible subject matter. McRO, 837 F.3d at 1313. The court found
that the process performed by human animators is not the same as that as the
rules-based process recited in the claimed automation, as the human process
is driven by subjective determinations. McRO, 837 F.3d at 1314. Thus, the
court in McRO held that it was the rules which improved the technological
process, and, that by incorporating the specific rules, the claim is limited to a
specific process for animation that is not an abstract idea. McRO, 837 F.3d
7
Appeal2017-010766
Application 14/846,486
at 1315-16. We consider the instant claim limitation to be similar to the
claim limitations at issue in McRO. Although the utterance intent clustering
limitation involves mathematical operations (see, e.g., Specification
paragraphs 29 and 38), similar to McRO, the Examiner has not shown that
the utterance intent clustering limitation is the same as conventionally
performed by person (paragraphs 14 through 16 of Appellants' Specification
describe the manual process as being different from the claimed intent
clustering limitation). Further, similar to McRO, the claims do not merely
organize information into a new form. Rather, the utterance intent clustering
limitation recites a specific order of steps (parsing utterances, combining all
utterances into one graphs with a common root node, and then clustering
into intent-wise homogeneous groups) that renders the information in a
specific format used to create the desired results. McRO, 837 F.3d at 1315.
The utterance intent clustering limitation, which produces intent-wise
homogeneous groups that are then used to train the SLU classifier, is
improving a technological process, as it is improving a specific process by
which an SLU classifier is trained. Accordingly, we do not sustain the
Examiner's rejection of independent claim 1 and dependent claims 2 through
12 as being directed to patent ineligible subject matter.
Appellants' arguments directed to independent claims 13 and 20
present similar arguments as those discussed with respect to claim 1. App.
Br. 17-30. The Examiner relies upon the same response discussed above to
support the rejection of independent claims 13 and 20. Answer 5---6. Each
of independent claims 13 and 20 recite utterance intent clustering limitations
similar to those discussed above in claim 1. Accordingly, we do not sustain
8
Appeal2017-010766
Application 14/846,486
the Examiner's rejection of claims 13 through 20 for the same reasons as
discussed above with respect to claim 1.
DECISION
We reverse the Examiner's rejections of claims 1 through 20 under
35 U.S.C. § 101.
REVERSED
9