MIYAZAKI, Hiroaki

Patent Trials and Appeals Board

Apr 1, 2020

15280022 - (D) (P.T.A.B. Apr. 1, 2020)

UNITED STATES PATENT AND TRADEMARK OFFICE
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.
15/280,022 09/29/2016 Hiroaki MIYAZAKI 1033413-000255 5734
21839 7590 04/01/2020
BUCHANAN, INGERSOLL & ROONEY PC
POST OFFICE BOX 1404
ALEXANDRIA, VA 22313-1404
EXAMINER
CHEN, ALAN S
ART UNIT PAPER NUMBER
2125
NOTIFICATION DATE DELIVERY MODE
04/01/2020 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):
ADIPDOC1@BIPC.com
PTOL-90A (Rev. 04/07)

UNITED STATES PATENT AND TRADEMARK OFFICE
____________

BEFORE THE PATENT TRIAL AND APPEAL BOARD
____________

Ex parte HIROAKI MIYAZAKI
____________

Appeal 2018-007636
Application 15/280,022
Technology Center 2100
____________

Before JAMES R. HUGHES, STEVEN M. AMUNDSON, and
MICHAEL T. CYGAN, Administrative Patent Judges.

AMUNDSON, Administrative Patent Judge.

DECISION ON APPEAL
Appellant1 seeks our review under 35 U.S.C. § 134(a) from a final
rejection of claims 1, 3–16, and 18–31, i.e., all pending claims. We have
jurisdiction under 35 U.S.C. § 6(b).
We affirm.

1 We use the word “Appellant” to refer to “applicant” as defined in
37 C.F.R. § 1.42 (2017).
Appeal 2018-007636
Application 15/280,022

2
STATEMENT OF THE CASE
The Invention
According to the Specification, the invention “relate[s] generally to
artificial intelligence apparatus and robots which autonomously process
problems and improve knowledge, and a software to execute those
functions.” Spec. ¶ 1.2 The Specification describes an “artificial
intelligence apparatus” including the following: (1) “an input processor
configured to convert input information to patterns”; (2) “an analyzer
configured to analyze the input information”; (3) “a recorder configured to
record the information”; (4) “a controller configured to perform at least one
of” (a) “a development of a process according to a type of a sentences and
an intention,” (b) “a search for information and a logic development to solve
a problem,” (c) “an execution of a process and activating a program,”
(d) “a generalization of information and a procedure,” (e) “an update to
a better knowledge and a logic,” (f) “a search and an arrangement of
information about an interesting field and an item,” (g) “recording and
updating of information, connective relations and relationship,” and
(h) “a transition control between information to a goal”; and (5) “an output
processor configured to convert the patterns to information or control
signals.” Id. at Abstract.

2 This decision uses the following abbreviations: “Spec.” for the
Specification, filed September 29, 2016; “Final Act.” for the Final Office
Action, mailed June 14, 2017; “Appeal Br.” for the Appeal Brief,
filed January 12, 2018; “Ans.” for the Examiner’s Answer, mailed
May 17, 2018; and “Reply Br.” for the Reply Brief, filed July 17, 2018.
Appeal 2018-007636
Application 15/280,022

3
Exemplary Claims
Independent claims 1 and 16 exemplify the claims at issue and read
as follows:
1. An artificial intelligence apparatus comprising:
an input processor configured to convert input
information to patterns, the input information including at least
one of: a word, language, a sentence, knowledge, a numerical
expression, a symbol, an image and sound;
an analyzer configured to analyze, of the patterns, at least
one of: words, meanings, parts of speech, fields, reliability,
newness, validity, comparison with recorded information, types
of sentences, intention and relations between sentences, and the
parts of speech including at least one of: subject, predicate,
modifiers of the subject, modifiers of the predicate and relations
between modifiers;
a recorder configured to record the patterns, resulting
from analysis by the analyzer, and at least one of: a type of a
problem, a method to solve the problem, connective relations
and a relationship between the patterns, an activation history of
the patterns, and an activation history of a program;
a controller configured to perform, autonomously using
the patterns, at least one of: a development of a process
according to the result of analysis, a search for information and
a logic development to solve a problem based on the result of
analysis, the type of the problem and the method to solve the
problem, an activation of a program according to the activation
history of the program, recording and updating of connective
relations and relationship between the patterns, and a transition
control between the patterns to a goal based on the result of
analysis; and
an output processor configured to convert a result of the
performance by the controller to information or control signals.
Appeal 2018-007636
Application 15/280,022

4
16. A computer implemented method for realizing an
artificial intelligence comprising:
converting input information to patterns, the input
information including at least one of: a word, language, a
sentence, knowledge, a numerical expression, a symbol, an
image and sound;
analyzing, of the patterns, at least one of: words,
meanings, parts of speech, fields, reliability, newness, validity,
comparison with recorded patterns, types of sentences, and
intention and relations between sentences of the input
information, and the parts of speech including at least one of:
subject, predicate, modifiers of the subject, modifiers of the
predicate and relations between modifiers;
recording the patterns, resulting from analysis by the
analyzer, and at least one of: a type of a problem, a method to
solve the problem, connective relations and a relationship
between the patterns, an activation history of the patterns,
and an activation history of a program;
performing, autonomously with the patterns, at least one
of:
a development of a process according to the result
of analysis,
a search for information and a logic development
to solve a problem with the result of analysis, the type of
the problem and the method to solve the problem,
an activating a program according to the activation
history of the program,
recording and updating of connective relations and
relationship between the patterns;
a transition control between the patterns to a goal based
on the result of analysis; and
converting a result of the performance by the controller
to information or control signals.
Appeal Br. Claims App. 1, 3–4.
Appeal 2018-007636
Application 15/280,022

5
The Prior Art Supporting the Rejections on Appeal
As evidence of unpatentability under 35 U.S.C. §§ 102 and 103, the
Examiner relies on U.S. Patent Application Publication 2016/0239751 A1
to Mosterman et al., titled “Multimodal Input Processing,” filed on
February 17, 2015, and published on August 18, 2016 (“Mosterman”).
The Rejections on Appeal
Claims 1, 3–16, and 18–31 stand rejected under 35 U.S.C. § 112(b) as
indefinite for failing to particularly point out and distinctly claim the subject
matter regarded as the invention. Final Act. 2–3; Ans. 8.
Claims 16 and 18–30 stand rejected under 35 U.S.C. § 102 as
anticipated by Mosterman. Final Act. 3–11; Ans. 3–7.
Claims 1, 3–15, and 31 stand rejected under 35 U.S.C. § 103 as
unpatentable over Mosterman. Final Act. 11–12; Ans. 7–8.
ANALYSIS
We have reviewed the rejections in light of Appellant’s arguments
that the Examiner erred. For the reasons explained below, we concur with
the Examiner’s conclusion concerning indefiniteness under § 112(b) and
largely agree with the Examiner’s determinations regarding anticipation
under § 102 and obviousness under § 103. But we disagree that Mosterman
anticipates dependent claims 28–30. We adopt the Examiner’s findings and
reasoning in the Final Office Action and Answer except for dependent
claims 28–30. See Final Act. 2–9, 11–12; Ans. 3–40. We add the following
to address and emphasize specific findings and arguments.
The § 112(b) Rejection of Claims 1, 3–16, and 18–31
Appellant states that the § 112(b) rejection “is not being appealed at
this time.” Appeal Br. 4. Because Appellant does not contest the § 112(b)
Appeal 2018-007636
Application 15/280,022

6
rejection, we summarily sustain that rejection. See In re Berger, 279 F.3d
975, 980, 984–85 (Fed. Cir. 2002) (holding that the Board did not err in
sustaining an indefiniteness rejection when the appellant failed to contest
that rejection on appeal); see also Manual of Patent Examining Procedure
(MPEP) § 1205.02 (9th ed. rev. 08.2017 Jan. 2018) (explaining that “[a]n
appellant’s brief must present arguments responsive to every ground of
rejection stated by the examiner in the Office action from which the appeal
has been taken (as modified by any advisory action and/or pre-appeal brief
conference decision)”).
The § 102 Rejection of Claims 16 and 18–30
ANTICIPATION UNDER § 102
“[T]o demonstrate anticipation, the proponent must show ‘that the
four corners of a single, prior art document describe every element of the
claimed invention.’” Net MoneyIN, Inc. v. VeriSign, Inc., 545 F.3d 1359,
1369 (Fed. Cir. 2008) (quoting Xerox Corp. v. 3Com Corp., 458 F.3d 1310,
1322 (Fed. Cir. 2006)). The reference “must not only disclose all elements
of the claim within the four corners of the document, but must also disclose
those elements ‘arranged as in the claim.’” Id. (quoting Connell v. Sears,
Roebuck & Co., 722 F.2d 1542, 1548 (Fed. Cir. 1983)). The “dispositive
question regarding anticipation [i]s whether one skilled in the art would
reasonably understand or infer from the [reference’s] teaching” that it
discloses every claim element arranged as in the claim. In re Baxter
Travenol Labs., 952 F.2d 388, 390 (Fed. Cir. 1991); see Dayco Prods.,
Inc. v. Total Containment, Inc., 329 F.3d 1358, 1368 (Fed. Cir. 2003).
Appeal 2018-007636
Application 15/280,022

7
CLAIM 16: ARTIFICIAL INTELLIGENCE
As noted above, the § 102 rejection of claim 16 rests on Mosterman.
See Final Act. 10–12; Ans. 3. Appellant attempts to distinguish claim 16
from Mosterman by contending that Mosterman “is not directed to artificial
intelligence” and instead “describes a system for recognizing and classifying
input information into one of a plurality of known formalism types.” Appeal
Br. 4–5. Appellant asserts that “the claimed subject matter is directed to a
method for realizing artificial intelligence, i.e., a self-learning mechanism.”
Id. at 7; see id. at 8.
Appellant’s arguments do not persuade us of Examiner error for two
reasons. First, claim 16 does not recite a self-learning mechanism. Appeal
Br. Claims App. 3–4. Appellant cannot rely on unclaimed features to
establish patentability. See In re Self, 671 F.2d 1344, 1348, 1350 (CCPA
1982).
Although claim 16’s preamble includes the term “artificial
intelligence,” a preamble does not generally limit a claim. See Allen Eng’g
Corp. v. Bartell Indus., Inc., 299 F.3d 1336, 1346 (Fed. Cir. 2002). For the
reasons explained below, the preamble term “artificial intelligence” simply
“state[s] a purpose or intended use for the invention.” See Catalina Mktg.
Int’l, Inc. v. Coolsavings.com, Inc., 289 F.3d 801, 809–10 (Fed. Cir. 2002)
(deciding that a preamble stated an intended use). Claim 16’s body defines a
complete invention. Appeal Br. Claims App. 3–4. The steps recited in the
claim body by themselves accomplish a result, i.e., furnishing “information
or control signals.” Id. The “artificial intelligence” recited in claim 16’s
preamble does not participate in accomplishing that result. Id. Deleting the
term “artificial intelligence” from claim 16’s preamble does not affect
Appeal 2018-007636
Application 15/280,022

8
method performance. Id. In addition, that term is not essential to
understand limitations or terms in the claim body. Id. Moreover, the claim
body does not rely on the preamble to provide an antecedent basis for a
claim term. Id.
Second, as explained below, Mosterman employs machine learning,
i.e., one type of artificial intelligence. See Mosterman ¶¶ 37, 62–63; see id.
¶ 59. Mosterman instructs that a formalism represents a way to organize,
define, or structure information and that formalisms include a table, a block
diagram, a state diagram, a mathematical formula, and a chemical structure.
Id. ¶ 13; see id. ¶¶ 8, 97–109, Figs. 7–13. Mosterman discloses a system for
(1) recognizing a formalism type in an input image, (2) processing
information from the input image using formalism-specific processing
techniques, and (3) generating a computer-based representation of the input
image. Id. ¶¶ 11–12, 14–16, Abstract.
Mosterman recognizes a formalism type in an input image by initially
identifying the elements in the input image, e.g., the rectangles, circles,
lines, arrows, mathematical symbols, and text in the input image.
Mosterman ¶¶ 14, 17–19, 30–37, 44–48, Figs. 1A, 2–3. As examples,
Mosterman explains that a block diagram may include “boxes/blocks,
arrows, and text” and that a state diagram may include circles, boxes,
straight arrows, and curved arrows. Id. ¶¶ 19, 33–34, 40–41, 72–73, 97–98,
108, Figs. 1A–1B, 2, 6–7, 12.
To identify the elements in an input image, Mosterman uses
algorithms, e.g., edge-detection algorithms and optical-character-recognition
algorithms. Mosterman ¶¶ 19, 46, 73. To improve performance, the
algorithms may apply the “learned meaning” of visual characteristics to
Appeal 2018-007636
Application 15/280,022

9
“future inputs” or even “different structures within the same input.” Id. ¶ 37.
Mosterman states that “[o]ther types of machine learning may also be
leveraged.” Id.
After identifying the elements in an input image, Mosterman
“determine[s] a likelihood that the elements would coexist” in a specific
formalism type by consulting a probability table. Mosterman ¶¶ 20–22,
52–55, Abstract, Figs. 1A, 4A–4B; see id. ¶ 98. Mosterman’s Figure 4A
(reproduced below) depicts a probability table:

Figure 4A illustrates “an exemplary probability table used to determine a
likelihood that two or more elements coexist in a given formalism based
on an analysis of elements in an input image.” Id. ¶ 4; see id. ¶ 53.
In Figure 4A, probability table 400 includes row 402 of formalism
types and column 404 of element types expected to appear in the formalism
Appeal 2018-007636
Application 15/280,022

10
types. Mosterman ¶¶ 53–54. A cell in the table specifies the probability that
an element type appears in a formalism type. Id. ¶ 56; see id. ¶ 21.
Mosterman discloses machine learning to dynamically update the probability
assigned to each cell. Id. ¶¶ 62–63. For instance, “the probability assigned
to each cell may be updated” based on the rate that each element-formalism
pair “is encountered in operation.” Id. ¶ 62. Further, “the computing system
may update the table” adaptively to “raise or lower the probability” in a cell
“based on feedback, such as user feedback.” Id. ¶ 63.
Using the probabilities in the table, Mosterman “select[s] a formalism
type that is compatible with the coexistence of” the elements in the input
image, e.g., by selecting the highest-scoring formalism type from among the
formalism types evaluated. Mosterman ¶¶ 26, 68; see id. ¶¶ 22–25.
After selecting a formalism type, Mosterman employs a processing
technique specific to the selected formalism type to (1) generate formalism-
specific elements corresponding to the elements in the input image and
(2) produce a computer-based representation of the input image. Mosterman
¶¶ 27–28, Abstract; see id. ¶¶ 8, 71, 96–109, Figs. 7–13. Further, the
computer-based representation may be “executed to carry out the
functionality of the model that was originally drawn in” the input image. Id.
¶ 28; see id. ¶ 96. For instance, if the input image includes a mathematical
equation, the system may produce a computer-based equation and “may
generate a solution to the equation and present the solution.” Id.
¶¶ 100–101, Fig. 8.
Hence, contrary to Appellant’s contention, Mosterman discloses
artificial intelligence, e.g., machine learning to discern visual characteristics
Appeal 2018-007636
Application 15/280,022

11
and adaptively update probabilities for element-formalism pairs. See
Ans. 9–12.
CLAIM 16: THE “CONVERTING” LIMITATION
Appellant argues that the Examiner erred in rejecting claim 16
because Mosterman fails to disclose the following limitation: “converting
input information to patterns, the input information including at least one of:
a word, language, a sentence, knowledge, a numerical expression, a symbol,
an image and sound.” See Appeal Br. 5–6; Reply Br. 2–3. In particular,
Appellant contends that the claimed “patterns” (1) are “separate and distinct
from the input information” and (2) “constitute data that is derived from the
input information.” Appeal Br. 5. Further, Appellant asserts that “[t]he
dictionary definition of ‘convert’ is ‘to change (something) into a different
form or properties; transmute; transform.’” Reply Br. 2 (footnote omitted).
Appellant argues that Mosterman’s “recognition of a line, box, text, shape or
other feature within an image does not transform the image into something
different.” Id. Appellant also argues that the “recognition, or identification,
of a feature is not the same as the ‘conversion’ of that feature into something
different.” Id.
Appellant’s arguments do not persuade us of Examiner error because
Mosterman converts image data (pixels) into computer-usable binary data
(ones and zeros) representing geometric shapes and other features. See
Mosterman ¶¶ 14, 17–19, 30–37, 44–48, 73, Figs. 1A, 2–3; Final Act. 4;
Ans. 12–20. Geometric shapes and other features correspond to the claimed
“patterns.” See Ans. 13. As the Examiner properly reasons, Mosterman
takes “as input an unprocessed raw input image, then appl[ies] preprocessing
algorithms to identify and extract patterns such as lines, boxes, text, shapes
Appeal 2018-007636
Application 15/280,022

12
or other features from that raw input image,” and “this is indeed ‘converting
input information to patterns.’” Id.
Consistent with this analysis, the Specification explains that “it is
possible to perform a process of imag[ing] information” and “transition from
an image pattern to” another pattern. Spec. ¶ 5. In addition, the
Specification describes an “input processor” that converts image data into
patterns. Id. ¶ 12, Fig. 1. Appellant does not explain how the conversion
of image data into patterns described in the Specification differs from
Mosterman’s conversion of image data into patterns. See Appeal Br. 5–6;
Reply Br. 2–3.
Appellant further argues that Mosterman’s preprocessing algorithms
“merely function to recognize existing content within the input image.”
Reply Br. 2. According to Appellant’s definition of “convert,” however, the
preprocessing algorithms “change (something) into a different form.” For
example, the preprocessing algorithms change the boxes/blocks, arrows, and
text in the hand-drawn block diagram shown in Figure 1A’s top panel into
the computer-based representation shown in Figure 1A’s middle panel.
Mosterman ¶¶ 18–19, Fig. 1A; see id. ¶¶ 44–48, 72–82. Hence, the
preprocessing algorithms change something, e.g., a block diagram, into a
different form because a hand-drawn form becomes a computer-based form.
For the reasons discussed above, Mosterman satisfies claim 16’s
“converting” limitation.
CLAIM 16: THE “ANALYZING” LIMITATION
Appellant argues that the Examiner erred in rejecting claim 16
because Mosterman fails to disclose the following limitation:
Appeal 2018-007636
Application 15/280,022

13
analyzing, of the patterns, at least one of: words, meanings,
parts of speech, fields, reliability, newness, validity,
comparison with recorded patterns, types of sentences, and
intention and relations between sentences of the input
information, and the parts of speech including at least one of:
subject, predicate, modifiers of the subject, modifiers of the
predicate and relations between modifiers.
See Appeal Br. 6; Reply Br. 3. Specifically, Appellant asserts that
Mosterman lacks this limitation because Mosterman analyzes the input
elements “relative to predefined formalism types, rather than any patterns
derived from such elements.” Appeal Br. 6 (emphasis omitted). Appellant
also asserts that Mosterman “only discloses processing the original input
information itself.” Reply Br. 3.
Appellant’s arguments do not persuade us of Examiner error because
Mosterman analyzes patterns, e.g., geometric shapes and other features,
derived from elements in an input image by comparing the derived image
elements to elements in a library or database to classify the derived image
elements as, among other things, rectangles, circles, lines, arrows,
mathematical symbols, and text. See Mosterman ¶¶ 14, 17–19, 30–37,
44–48, Figs. 1A, 2–3; Final Act. 4; Ans. 21–22. When Mosterman analyzes
the derived image elements by comparing them to elements in a library or
database, Mosterman analyzes patterns by “comparison with recorded
patterns” according to claim 16’s “analyzing” limitation.
Moreover, after classifying the derived image elements, Mosterman
further analyzes them to “determine a likelihood that the elements would
coexist” in a specific formalism type by consulting a probability table.
Mosterman ¶¶ 20–26, 52–56, Abstract, Figs. 1A, 4A–4B; see id. ¶¶ 4,
68, 98. Using the probabilities in the table, Mosterman “select[s] a
Appeal 2018-007636
Application 15/280,022

14
formalism type that is compatible with the coexistence of” the elements
in the input image. Id. ¶¶ 26, 68.
The Specification explains that “combinations of patterns are also
patterns.” Spec. ¶¶ 5, 13. In Mosterman, each formalism type comprises
a combination of patterns. As an example, a block diagram may include
“boxes/blocks, arrows, and text.” Mosterman ¶¶ 19, 33–34, 40–41,
Figs. 1A–1B, 2. As another example, a state diagram may include circles,
boxes, straight arrows, and curved arrows. Id. ¶¶ 33–34, 97–98, Figs. 2, 7.
Hence, when Mosterman analyzes derived image elements to “determine a
likelihood that the elements would coexist” in a specific formalism type,
Mosterman analyzes patterns by “comparison with recorded patterns”
according to claim 16’s “analyzing” limitation.
For the reasons discussed above, Mosterman satisfies claim 16’s
“analyzing” limitation.
CLAIM 16: THE “RECORDING” LIMITATION
Appellant argues that the Examiner erred in rejecting claim 16
because Mosterman fails to disclose the following limitation: “recording the
patterns, resulting from analysis by the analyzer, and at least one of: a type
of a problem, a method to solve the problem, connective relations and a
relationship between the patterns, an activation history of the patterns, and
an activation history of a program.” See Appeal Br. 6; Reply Br. 4.
Specifically, Appellant asserts that Mosterman fails to disclose “recording
a type of problem, or a method to solve the problem, together with a
formalism that has been identified through its processing techniques.”
Appeal Br. 6; see Reply Br. 4. In addition, Appellant contends that “once a
formalism has been identified” in Mosterman, “there is no disclosure that a
Appeal 2018-007636
Application 15/280,022

15
problem or method to solve the problem is thereafter recorded in
conjunction with that identified formalism.” Reply Br. 4.
Appellant’s arguments do not persuade us of Examiner error because
Mosterman records the claimed “patterns” together with (1) “a type of a
problem” and (2) “connective relations and a relationship between the
patterns” according to claim 16’s “recording” limitation. See Mosterman
¶¶ 27–28, 71, 96–109, Abstract, Figs. 1B, 7–13; Final Act. 5; Ans. 23–30.
In particular, after selecting a formalism type, Mosterman employs a
processing technique specific to the selected formalism type to (1) generate
formalism-specific elements corresponding to the elements in the input
image and (2) produce a computer-based representation of the input image.
Mosterman ¶¶ 27–28, Abstract, Fig. 1B; see id. ¶¶ 8, 71, 96–109, Figs. 7–13.
The computer-based representation of the input image corresponds to “a
type of a problem” solved by the system, i.e., determining what formalism
type or types the input image includes. See Ans. 23–25; Mosterman ¶¶ 18–
28, 30–48, Figs. 1A–1B, 2–3. For instance, if the input image includes a
mathematical equation, the system may produce a computer-based equation
and “may generate a solution to the equation and present the solution.”
Mosterman ¶¶ 100–101, Fig. 8.
The computer-based representation of the input image also
corresponds to “connective relations and a relationship between the patterns”
according to claim 16’s “recording” limitation. See Ans. 25–30. As an
example, a block diagram shows “connective relations and a relationship
between the patterns” of “boxes/blocks, arrows, and text.” Mosterman
¶¶ 19, 33–34, 40–41, 72–73, 108, Figs. 1A–1B, 2, 6A, 12. As another
example, a state diagram shows “connective relations and a relationship
Appeal 2018-007636
Application 15/280,022

16
between the patterns” of circles, boxes, straight arrows, and curved arrows.
Id. ¶¶ 33–34, 97–98, Figs. 2, 7. As yet another example, a mathematical
equation shows “connective relations and a relationship between the
patterns” of symbols on opposite sides of an equals sign. Id. ¶¶ 100–101,
Fig. 8; see id. ¶¶ 12–13, 16; Ans. 25–26.
For the reasons discussed above, Mosterman satisfies claim 16’s
“recording” limitation.
CLAIM 16: THE “TRANSITION” LIMITATION
Appellant argues that the Examiner erred in rejecting claim 16
because Mosterman fails to disclose the following limitation: “a transition
control between the patterns to a goal based on the result of analysis.” See
Appeal Br. 6–7; Reply Br. 4. Specifically, Appellant asserts that transitions
in Mosterman “are from an identified (or extracted) formalism to a particular
program for handling that type of formalism” and not “from one pattern to
another pattern.” Appeal Br. 7; see Reply Br. 4.
Appellant’s arguments do not persuade us of Examiner error because
Mosterman transitions control between pattern analysis and a goal based on
the result of analysis when it “select[s] a formalism type that is compatible
with the coexistence of” the elements in the input image. See Mosterman
¶¶ 22–26, 65–68, Figs. 1A, 5; Final Act. 6; Ans. 30–33. After selecting a
formalism type, Mosterman employs a processing technique specific to the
selected formalism type to (1) generate formalism-specific elements
corresponding to the elements in the input image and (2) produce a
computer-based representation of the input image. Mosterman ¶¶ 27–28,
Abstract; see id. ¶¶ 8, 71, 96–109, Figs. 7–13. Thus, after selecting a
formalism type, control transitions to a processing technique specific to the
Appeal 2018-007636
Application 15/280,022

17
selected formalism type, and that processing technique accomplishes the
goal of producing a computer-based representation of the input image. See
Ans. 30–32.
For the reasons discussed above, Mosterman satisfies claim 16’s
“transition” limitation.
SUMMARY FOR CLAIM 16
For the reasons discussed above, Appellant’s arguments have not
persuaded us that the Examiner erred in rejecting claim 16 under § 102 as
anticipated by Mosterman. Mosterman includes every element of claim 16
arranged as required by the claim and, therefore, anticipates the claim. See
Final Act. 4–6; Ans. 9–33. Hence, we sustain the § 102 rejection of
claim 16.
DEPENDENT CLAIMS 18, 21, 22, AND 24–26
Claims 18, 21, 22, and 24–26 depend from claim 16. Appellant does
not argue patentability separately for these dependent claims. See Appeal
Br. 4–10; Reply Br. 2–4. Thus, we sustain the § 102 rejection of these
dependent claims for the same reasons as claim 16. See 37 C.F.R.
§ 41.37(c)(1)(iv).
DEPENDENT CLAIM 19
Claim 19 depends from claim 16 and further requires “evaluating the
patterns and performing the processing according to the results of analysis
autonomously, the processing including recording the patterns, updating a
knowledge system, executing instructed items, and answering questions.”
Appeal Br. Claims App. 4. Appellant asserts that the Examiner erred in
rejecting claim 19 because Mosterman fails to disclose “updating a
knowledge system.” Appeal Br. 8. Appellant contends that the “update”
Appeal 2018-007636
Application 15/280,022

18
shown in Mosterman’s Figure 6A “add[s] a block to a block diagram model”
and “has nothing to do with updating information in a knowledge system.”
Id.
We disagree. “[D]uring examination proceedings, claims are given
their broadest reasonable interpretation consistent with the specification.”
In re Hyatt, 211 F.3d 1367, 1372 (Fed. Cir. 2000). Here, the Examiner
properly determines that the Specification contains “broad descriptions of a
knowledge system.” Ans. 36 (citing Spec. ¶ 5). For example, the
Specification explains that a knowledge system may include “truth, fact,
rules, common sense, right and wrong, definitions, logic, explanation,
hypotheses, prediction, opinion, impression, [and] rumor.” Spec. ¶ 1;
see id. ¶¶ 5–6, 12–13, 17–18, Fig. 23.
Mosterman constructs a knowledge system by (1) generating
formalism-specific elements corresponding to the elements in the input
image and (2) producing a computer-based representation of the input
image. Mosterman ¶¶ 27–28, Abstract; see id. ¶¶ 8, 71, 96–109, Figs. 7–13;
Ans. 35–37. Mosterman’s Figures 7–13 depict exemplary computer-based
representations comprising knowledge systems, i.e., knowledge systems
resulting from analyses of various formalism types originating from image
data. See Mosterman ¶¶ 8, 97–110, Figs. 7–13; Final Act. 6; Ans. 35–37.
As the Examiner properly reasons, “whenever new raw input data is
formalized into . . . any of the formalism types shown in fig. 7–13, this can
be construed as ‘updating a knowledge system’ since new relationships are
formed that were previous[ly] not present.” Ans. 37.
For the reasons discussed above, Appellant’s arguments have not
persuaded us that the Examiner erred in rejecting claim 19 under § 102 as
Appeal 2018-007636
Application 15/280,022

19
anticipated by Mosterman. Mosterman includes every element of claim 19
arranged as required by the claim and, therefore, anticipates the claim. See
Final Act. 6; Ans. 35–37. Hence, we sustain the § 102 rejection of claim 19.
DEPENDENT CLAIM 20
Claim 20 depends from claim 16 and reads as follows:
20. The computer implemented method for realizing an
artificial intelligence of claim 16, further comprising:
extracting a useful relation by strengthening a relation
between one group of patterns and another group of patterns,
the relations associating the input information with a significant
relation, and
constructing at least one of: a generalized thought, a
prediction, an estimate, and a common sense from the input
information autonomously.
Appeal Br. Claims App. 4.
Appellant asserts that the Examiner erred in rejecting claim 20
because Mosterman fails to disclose “strengthening a relation between one
group of patterns and another group of patterns.” Appeal Br. 8. Appellant
contends that Mosterman’s model for selecting a formalism type based on
probabilities “does not pertain to groups of patterns, nor relationships
between different groups of patterns, let alone the strengthening of such
relationships.” Id. Appellant also contends that Mosterman’s model
“merely serve[s] as a tool for differentiating respective formalisms from one
another as likely candidates.” Id.
We disagree. As the Examiner explains for a block diagram, the
rectangle and associated text for a first block corresponds to a first group
of patterns, and the rectangle and associated text for a second block
corresponds to a second group of patterns. Ans. 38; see Mosterman
Appeal 2018-007636
Application 15/280,022

20
¶¶ 65–69, 108, Figs. 4–5, 12. As the Examiner properly reasons,
Mosterman’s model for selecting a formalism type based on probabilities
strengthens the relation between the two groups of patterns “by applying the
probability table” in Figure 4 “to determine that these two groups of
patterns” relate to the same block diagram. Ans. 38–39.
For the reasons discussed above, Appellant’s arguments have not
persuaded us that the Examiner erred in rejecting claim 20 under § 102 as
anticipated by Mosterman. Mosterman includes every element of claim 20
arranged as required by the claim and, therefore, anticipates the claim. See
Final Act. 6–7; Ans. 37–39. Hence, we sustain the § 102 rejection of
claim 20.
DEPENDENT CLAIM 23
Claim 23 depends from claim 16 and further requires “by using a
knowledge system, analyzing a situation, recognizing a problem, searching
for information to solve the problem, activating a programs [sic] to perform
processing, generating information to solve the problem, installing the
information into the knowledge system, solving the problem and expanding
the knowledge furthermore autonomously.” Appeal Br. Claims App. 5.
Appellant asserts that the Examiner erred in rejecting claim 23 because
(1) “[t]his claim encompasses the self-learning capabilities of the artificial
intelligence mechanism” and (2) Mosterman “does not pertain to artificial
intelligence.” Appeal Br. 9. Appellant contends that Mosterman “does not
disclose the concept of a knowledge system, let alone generating information
to solve a problem, and then installing that information into a knowledge
system to expand a knowledge base.” Id.
Appeal 2018-007636
Application 15/280,022

21
Appellant’s arguments do not persuade us of Examiner error for three
reasons. First, claim 23 does not recite a self-learning capability. Appeal
Br. Claims App. 5. Appellant cannot rely on unclaimed features to establish
patentability. See Self, 671 F.2d at 1348, 1350.
Second, as discussed above for claim 16, Mosterman employs
machine learning, and thus relates to artificial intelligence. See Mosterman
¶¶ 37, 62–63; see id. ¶ 59.
Third, Mosterman “generat[es] information to solve the problem” and
“install[s] the information into the knowledge system” according to
claim 23. See Final Act. 6, 8; Ans. 11–12, 35–37, 39; Mosterman ¶¶ 8,
27–28, 71, 96–109, Abstract, Figs. 7–13. Mosterman does so by
(1) generating formalism-specific elements corresponding to the elements
in the input image and (2) producing a computer-based representation of
the input image. Mosterman ¶¶ 27–28, Abstract; see id. ¶¶ 8, 71, 96–109,
Figs. 7–13. Mosterman’s Figures 7–13 depict exemplary computer-based
representations comprising knowledge systems, i.e., knowledge systems
resulting from analyses of various formalism types originating from image
data. See id. ¶¶ 8, 97–110, Figs. 7–13; Final Act. 6; Ans. 35–37.
For the reasons discussed above, Appellant’s arguments have not
persuaded us that the Examiner erred in rejecting claim 23 under § 102 as
anticipated by Mosterman. Mosterman includes every element of claim 23
arranged as required by the claim and, therefore, anticipates the claim. See
Final Act. 6, 8; Ans. 11–12, 35–37, 39. Hence, we sustain the § 102
rejection of claim 23.
Appeal 2018-007636
Application 15/280,022

22
DEPENDENT CLAIM 27
Claim 27 depends from claim 16 and further requires “constructing a
knowledge system having a relation between the patterns and summarizing
or developing in detail using the relation between the patterns and the
knowledge system autonomously.” Appeal Br. Claims App. 6. Appellant
asserts that the Examiner erred in rejecting claim 27 because Mosterman’s
model for selecting a formalism type based on probabilities does not identify
“a relationship between image elements represented in the layer of nodes
502” shown in Mosterman’s Figure 5. Appeal Br. 9. Appellant contends
that Mosterman contains “no mention of any relationship between the
individual elements, let alone that such information is contained in a
knowledge system.” Id.
We disagree. As discussed above for claim 20, in a block diagram,
the rectangle and associated text for a first block corresponds to a first group
of patterns, and the rectangle and associated text for a second block
corresponds to a second group of patterns. Ans. 38; see Mosterman
¶¶ 65–69, 108, Figs. 4–5, 12. Thus, Mosterman’s model for selecting a
formalism type based on probabilities strengthens the relation between the
two groups of patterns “by applying the probability table” in Figure 4 “to
determine that these two groups of patterns” relate to the same block
diagram. Ans. 38–39; see Mosterman ¶¶ 65–69, 108, Figs. 4–5, 12. And as
discussed above for claim 16’s “recording” limitation, a block diagram
shows “connective relations and a relationship between the patterns” of
“boxes/blocks, arrows, and text.” Mosterman ¶¶ 19, 33–34, 40–41, 72–73,
108, Figs. 1A–1B, 2, 6A, 12.
Appeal 2018-007636
Application 15/280,022

23
For the reasons discussed above, Appellant’s arguments have not
persuaded us that the Examiner erred in rejecting claim 27 under § 102 as
anticipated by Mosterman. Mosterman includes every element of claim 27
arranged as required by the claim and, therefore, anticipates the claim. See
Final Act. 9; Ans. 25–30, 37–40. Hence, we sustain the § 102 rejection
of claim 27.
DEPENDENT CLAIMS 28–30
Claim 28 depends from claim 16 and further requires “recording the
patterns associatively, searching for the pattern about a designated theme
and a topic and generating a conversation according to an intention of a
speaker.” Appeal Br. Claims App. 6. Appellant argues that the Examiner
erred in rejecting claim 28 because Mosterman fails to disclose “generating
a conversation according to an intention of a speaker.” Appeal Br. 10.
Appellant asserts that Mosterman’s “disclosure of audio data is in the
context of applicable forms of input to the formalism-specific processing.”
Id. Appellant also asserts that claim 28 “recites generating a conversation
after searching for a pattern” and requires “a form of output.” Id.
In response, the Examiner states that Mosterman discloses “the ability
to handle audio data input.” Ans. 42. The Examiner explains that “[i]f the
audio data is speech input from a human speaker, e.g., ‘a conversation’, the
recording will be exactly as intended by the speaker since it is an exact
digital recording of what was spoken.” Id.
Based on the record before us, we agree with Appellant that
“generating a conversation” requires a form of output, i.e., outputting one or
more words as part of an exchange. See Spec. ¶¶ 12, 15–16. The Examiner
has not adequately explained how the cited portions of Mosterman disclose
Appeal 2018-007636
Application 15/280,022

24
outputting one or more words as part of an exchange. See Final Act. 9–10;
Ans. 41–42. Hence, we do not sustain the § 102 rejection of claim 28.
Claim 29 requires “generating conversations according to the
context,” and claim 30 similarly requires “generating a conversation
according to the context.” Appeal Br. Claims App. 6. For claims 29 and 30,
Appellant references the arguments for claim 28 and asserts that “there has
been no showing” that Mosterman “discloses generating an output that
consists of a conversation, let alone one directed to any of the specific types
recited in claims 29 or 30.” Appeal Br. 10. In response, the Examiner states
that the “assertions for Claim 28 [are] applicable here with regard to
‘generating a conversation’” in claims 29 and 30. Ans. 43. For the reasons
discussed above for claim 28, we do not sustain the § 102 rejection of
claims 29 and 30.
The § 103 Rejection of Claims 1, 3–15, and 31
CLAIM 1 AND DEPENDENT CLAIMS 3–15
For claim 1 and dependent claims 3–15, Appellant references the
arguments “previously discussed with respect to claims 16 and 18–30” and
does not argue patentability separately for these claims. See Appeal
Br. 4–10; Reply Br. 2–4. Because Appellant does not argue patentability
separately for these claims, we sustain the § 103 rejection of these claims.
See 37 C.F.R. § 41.37(c)(1)(iv).
DEPENDENT CLAIM 31
Claim 31 depends from claim 1 and specifies that:
the controller is configured to analyze the relations between
the pattern converted by the input processor and the recorded
patterns, construct, expand and update a knowledge system by
extracting the pattern converted by the input processor which
Appeal 2018-007636
Application 15/280,022

25
is valuable, and solve problems using the knowledge system,
relations between the patterns and a pattern generated by the
programs.
Appeal Br. Claims App. 6. Appellant argues that the Examiner erred in
rejecting claim 31 because Mosterman lacks “any disclosure relating to
relations between patterns, i.e., image elements,” and “updating a
knowledge system on the basis of such patterns.” Appeal Br. 11.
We disagree. As discussed above for claim 16’s “recording”
limitation, Mosterman discloses determining relations between patterns
originating from image data. And as discussed above for claims 19 and 23,
Mosterman discloses installing and updating a knowledge system based on
patterns originating from image data. Hence, we sustain the § 103 rejection
of claim 31.3
CONCLUSION
We affirm the rejection of claims 1, 3–16, and 18–31 under 35 U.S.C.
§ 112(b).
We affirm the rejection of claims 16 and 18–27 under 35 U.S.C.
§ 102.
We reverse the rejection of claims 28–30 under 35 U.S.C. § 102.

3 In the event of continued prosecution, the Examiner should consider (1) the
applicability of 35 U.S.C. § 112(f) to the limitations in claims 1, 3–15,
and 31 and (2) whether under 35 U.S.C. § 112(b) the Specification discloses
adequate corresponding structure, e.g., a suitable algorithm, for
accomplishing each function recited in a limitation. See, e.g., Advanced
Ground Info. Sys., Inc. v. Life360, Inc., 830 F.3d 1341, 1349–50 (Fed. Cir.
2016); Williamson v. Citrix Online, LLC, 792 F.3d 1339, 1348–54 (Fed. Cir.
2015); EON Corp. IP Holdings LLC v. AT&T Mobility LLC, 785 F.3d 616,
624 (Fed. Cir. 2015).
Appeal 2018-007636
Application 15/280,022

26
We affirm the rejection of claims 1, 3–15, and 31 under 35 U.S.C.
§ 103.
Because we affirm at least one ground of rejection for each claim on
appeal, we affirm the Examiner’s decision to reject all of the claims on
appeal. See 37 C.F.R. § 41.50(a)(1).
In summary:
Claims
Rejected 35 U.S.C. § Reference(s)/Basis Affirmed Reversed
1, 3–16,
18–31 112(b) Indefiniteness
1, 3–16,
18–31
16, 18–30 102 Mosterman 16, 18–27 28–30
1, 3–15, 31 103 Mosterman 1, 3–15, 31
Overall
Outcome
1, 3–16,
18–31
TIME PERIOD FOR RESPONSE
No time period for taking any subsequent action in connection with
this appeal may be extended under 37 C.F.R. § 1.136(a)(1)(iv). See
37 C.F.R. § 41.50(f).
AFFIRMED