Ex Parte Turner et al

Patent Trial and Appeal Board

Nov 25, 2014

11522177 (P.T.A.B. Nov. 25, 2014)

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.
11/522,177 09/15/2006 Alan E. Turner 15268-E (BA4-304) 3929
21567 7590 11/25/2014
Wells St. John P.S.
601 West First Avenue
Suite 1300
Spokane, WA 99201-3828
EXAMINER
AMIN, MUSTAFA A
ART UNIT PAPER NUMBER
2176
MAIL DATE DELIVERY MODE
11/25/2014 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 ALAN E. TURNER, ELIZABETH G. HETZLER, and
GRANT C. NAKAMURA
____________

Appeal 2012-002280
Application1 11/522,177
Technology Center 2100
____________

Before ERIC B. GRIMES, LORA M. GREEN, and ULRIKE W. JENKS,
Administrative Patent Judges.

JENKS, Administrative Patent Judge.

DECISION ON APPEAL
This is an appeal under 35 U.S.C. § 134 involving claims directed to a
device and method for text analysis. The Examiner has rejected the claims
as anticipated and obvious. We have jurisdiction under 35 U.S.C. § 6(b).
We affirm-in-part.
STATEMENT OF THE CASE
Claims 11–14, 16–23, 25, 29, 30, 32, 34–36, and 41–45 are on
appeal,2 and can be found in the Claims Appendix of the Appeal Brief (see

1 Appellants identify Battelle Memorial Institute as the Real Party in
Interest. (App. Br. 1.)
Appeal 2012-002280
Application 11/522,177

2
also App. Br. 1). Claims 11, 18, and 32 are representative of the claims on
appeal, and read as follows:
11. An article of manufacture comprising:
a computer readable storage medium configured to store
programming configured to cause processing circuitry to
perform processing comprising:
accessing a first representation of an initial text item;
accessing a second representation of the initial text item,
wherein the second representation is different than the first
representation;
combining the first and second representations of the
initial text item to form a third representation of the initial text
item which is different than the first and second representations;
using the third representation, processing the initial text
item with respect to another text item; and
wherein the combining comprises initial combining, and
wherein the programming is configured to cause the processing
circuitry to perform processing comprising subsequent
combining of the first and second representations of the initial
text item to form a fourth representation of the initial text item
which is different than the first, second and third
representations.

18. A text analysis method comprising:
using processing circuitry, generating a first
representation of a text item using a first measurement basis;
using the processing circuitry, generating a second
representation of the text item using a second measurement
basis different than the first measurement basis, wherein the
second representation is different than the first representation;
using the processing circuitry, analyzing the text item
using the first representation and the second representation; and

2 The Examiner has withdrawn the rejection of claims 1–10 and 38–40 under
35 U.S.C § 103(a) over Calistri-Yeh in view of Ukrainczyk, as well as the
rejection of claims 38 and 40–45 under 35 U.S.C. § 112, first paragraph, and
the rejection of claim 11 under 35 U.S.C. § 112, second paragraph (Ans. 4).
Appeal 2012-002280
Application 11/522,177

3
using the processing circuitry, steering the generating of
at least one of the first and second representations according to
a perspective of interest of a user.

32. A text analysis method comprising:
using processing circuitry, accessing a first representation
of a text item;
using the processing circuitry, accessing a second
representation of the text item;
using the processing circuitry, accessing a blend
constant;
using the processing circuitry, weighting at least one of
the first and second representations using the accessed blend
constant;
after the weighting and using the processing circuitry,
providing a third representation of the text item using the first
and second representations which were weighted using the
accessed blend constant;
using the processing circuitry, changing the blend
constant;
using the processing circuitry, reweighting the at least
one of the first and second representations using the blend
constant after the providing the third representation and after
the changing; and
using the processing circuitry, providing a fourth
representation of the text item using the first and second
representations after the reweighting.
The following grounds of rejection are before us for review:
I. claims 11–14, 16, 17, 32, 35, and 36 under 35 U.S.C § 102(b)
as anticipated by Calistri-Yeh3 (Ans. 5);
II. claims 18–23, 25, 29, 30, and 41–45 under 35 U.S.C § 103(a)
as obvious over Calistri-Yeh in view of Ukrainczyk4 (Ans. 17);
and

3 Calistri-Yeh et al., US 2004/0199505 A1, published Oct. 7, 2004.
Appeal 2012-002280
Application 11/522,177

4
III. claim 34 under 35 U.S.C. § 103(a) as obvious over Calistri-Yeh
in view of Fox5 (Ans. 36).

I. The Issue: Anticipation by Calistri-Yeh
The Examiner finds that Calistri-Yeh meets all claim limitations (see
Ans. 5–9, 38–39).
Appellants contend that the trainable semantic vectors (TSV) as
disclosed in Calistri-Yeh are not determined multiple times using the same
data point as claimed. Specifically, “[t]here are no teachings that a first
TSV is calculated using the formula of paragraph 0055 a plurality of times
for a given data point, and to the contrary, paragraph 0057 discloses that
the TSV formula is utilized to calculate a single TSV value for each data
point.” (Reply Br. 3.)
Does the preponderance of the evidence of record support the
Examiner’s finding that Calistri-Yeh anticipates the claims?

Findings of Fact
1. Calistri-Yeh disclosed “producing a semantic representation of
information in a semantic space” (Calistri-Yeh Abstract). The construction
of trainable semantic vectors (TSV) allows a visual representation of this
information. Fig. 2 is reproduced below:

4 Ukrainczyk et al., US 2006/0143175 A1, published June 29, 2006.
5 Fox et al., US 6,574,632 B2, issued June 3, 2003.
App
App

Fig 2
step
is ind
categ

deter
form

entri
acco
eal 2012-0
lication 11
shows th
S210, a da
icative of
ories.” (C
2. C
mine the s
ula: TSV=
3. “A
es (i.e. the
mplished b
02280
/522,177
e steps inv
ta table is
a relation
alistri-Ye
alistri-Yeh
ignificanc
α(v)+(l-α)
t step S21
data point
y first cal
olved in th
constructe
ship betwe
h 4: ¶ 46.)
disclosed
e of data p
(u)” (Cal
2 [shown
s) in the d
culating th
5
e construc
d. The da
en data po

applying w
oints, acc
istri-Yeh 5
above Fig
ata table is
e percenta
tion of a s
ta table sto
ints and p
eighting
ording to t
: ¶¶ 55–56
. 2], the si
determin
ge of data
emantic d
res inform
redetermin
factor α “t
he followi
; Ans. 14.
gnificance
ed. . . . Th
points occ
ataset. “A
ation that
ed
o
ng
)
of the
is can be
urring in
t

App
App

each
51; A

of a
Ans.
asso

cons
each
categ
poin

word
“FIG
stren
110;

TSV
eal 2012-0
lication 11
category”
ns. 14.)
4. “V
data point
14.) “Bot
ciated with
5. “A
tructed. T
entry, or d
ories). . .
t in the dat
6. C
in a given
. 9 illustr
gth of eac
Ans. 40.)
7. Fi
is “calcul
02280
/522,177
and is rep
” in the T
’s occurren
h u and v
a particul
t step S21
he first TS
ata point,
. Accordin
a table.” (
alistri-Yeh
category
ates a table
h word wi

g. 10 of C
ated based
resented a
SV formu
ce across
represent t
ar categor
4 [shown
V corresp
across the
gly, a firs
Calistri-Y
disclosed
. Fig. 9 is
230 that
th respect
alistri-Yeh
on the act
6
s “u” in th
la represen
all categor
he strengt
y.” (Calis
above in F
onds to a s
semantic
t TSV mu
eh 5: ¶ 57
a table of
reproduce
stores the v
to the cate
, reproduc
ual values
e formula.
ts “the pro
ies.” (Cal
h with wh
tri-Yeh 5:
ig. 2], a fi
emantic re
space (i.e.
st be const
; Ans. 14.)
stored val
d below:
alues that
gories.” (
ed below,
stored in
(Calistri-
bability d
istri-Yeh 5
ich an entr
¶ 53; Ans
rst TSV is
presentati
, the prede
ructed for

ues associ

indicate t
Calistri-Ye
shows tha
table 230.
Yeh 5: ¶
istribution
: ¶ 52;
y is
. 14.)

on for
termined
each data
ated with a
he relative
h 10: ¶
t each

App
App

Acco
repre
“FIG
the w
the s
Mor
α(v)+

of a
Fig.
docu
11: ¶

recei
eal 2012-0
lication 11
rdingly, th
sentation
. 10, a tab
ords . . . .
emantic re
e particula
(l-α)(u).”
8. Fi
plurality o
11 shows
ments wou
114.)
9. C
ved they “
02280
/522,177
e TSVs sh
of the exem
le 250 is s
Table 25
presentati
rly, the fol
(Calistri-
g. 11 of C
f words de
“a two-dim
ld be clus
alistri-Yeh
must be cl
own in ta
plary wo
hown for
0 is a com
on of each
lowing for
Yeh 10: ¶
alistri-Yeh
fined by o
ensional c
tered in th
disclosed
assified w
7
ble 250 co
rds” (Cali
illustrating
bination of
word acro
mula is us
112.)
, reproduc
nly two ca
oordinate
e same ma
that when
ithin the p
rrespond t
stri-Yeh 10
the sema
five TSV
ss the sem
ed to calc
ed below,
tegories.
for each w
nner as w
additiona
reviously
o the actu
: ¶ 113).

ntic repres
s that corr
antic spac
ulate the v
shows the

ord. Not
ords.” (Ca
l items (w
defined ca
al
entation o
espond to
e. . . .
alues.
clustering
e that
listri-Yeh
ords) are
tegories.”
f

Appeal 2012-002280
Application 11/522,177

8
(Calistri-Yeh 8: ¶ 86.) After newly received data items are placed into
categories this eventually leads to “a process wherein the nature of each
category is reevaluated. This iterative approach enables the classification
algorithm to adapt to changes in data and definition over time.” (Calistri-
Yeh 8: ¶ 87.) “TSVs are optionally reconstructed. More particularly, the
reconstructed category TSVs are recalculated according to the method
described earlier to represent the semantic dimension across the space of the
original sample items within that category as well as the newly added items
within that category.” (Calistri-Yeh 8: ¶ 88.)

Principle of Law
“A single prior art reference that discloses, either expressly or
inherently, each limitation of a claim invalidates that claim by anticipation.”
Perricone v. Medicis Pharm. Corp., 432 F.3d 1368, 1375 (Fed. Cir. 2005).

Analysis
Claim 11
Calistri-Yeh disclosed the use of trainable semantic vectors to display
relational information about data sets (FF 1–9). The process involves the
production of data tables for storing relational information about data points
in a particular category (FF 1, 6–8). The number of data points in a given
category provides information about the importance of each data point
within that particular category (FF 3–4). A data point can also fall into more
than one category and the distribution of a data point across the various
categories is also given a value (FF 4). “[A] first TSV must be constructed
for each data point in the data table” (FF 5, 7). “Through empirical evidence
Appeal 2012-002280
Application 11/522,177

9
gathered from experimentation, . . . [it is] determined that the weighted
average of the u and v vectors can produce superior results than achievable
without the use of a weighting factor.” (Calistri-Yeh 5: ¶ 56.) Calistri-Yeh
disclosed the TSV is calculated with the formula TSV=α(v)+(l-α)(u), where
α is a weighting factor (FF 2). Construction of the first TSV requires the
input of values “u” and “v” (FF 1–8).
We find that claim 11 recites a first, second, third, and fourth
representation of an initial text item. The claim also requires that these
representations are different from each other. The claim, however, is silent
with respect to showing these representations simultaneously on a display, or
even that the information about the four representations are simultaneously
accessible in a table.
Appellants contend that “allowing the user to select the value of the
weighting factor which is used in the TSV formula fails to disclose that the
TSV formula is calculated a plurality of times using different values of the
weighting factor to provide different TSV values for the same data point.”
(App. Br. 21; see also Reply Br. 4.)
We are not persuaded by Appellants’ contentions. The Examiner
identifies “u” as being a first representation of an initial text item, and “v” as
being “a second representation of the initial text item” (Ans. 18). The
Examiner identifies the combination of “u and v without a[ ] weighting
factor” to form the third representation of an initial text item (Ans. 38
(emphasis added)). “Calistri-Yeh explicitly discloses that from
experimentation, the weighted average of u and v vectors provide superior
results than without the use of the weighting factor (i.e. for a certain data
point, a first TSV was calculated without weighting a factor and for the
Appeal 2012-002280
Application 11/522,177

10
same data point, a second TSV was calculated with the weighting factor)”
(Ans. 38). Finally, according to the Examiner, the combination of “u and v
with a[ ] weighting factor” is identified as the fourth representation (Ans.
38–39 (emphasis added)).
The Examiner’s position for establishing that the initial values “u” and
“v” are used multiple times to construct a TSV comes from Calistri-Yeh’s
explanation that the superiority of the weighting factor was empirically
determined. “Through empirical evidence gathered from experimentation,
the inventors [Calistri-Yeh] have determined that the weighted average of
the u and v vectors can produce superior results than achievable without the
use of a weighting factor.” (Calistri-Yeh 5: ¶ 56.) We agree with the
Examiner’s position that in order to compare the results with and without the
use of a weighting factor, necessarily requires the use the same initial data
points “u” and “v.” Thus, we agree with the Examiner’s finding that
Calistri-Yeh disclosed a first and second representation as well as a third and
fourth representation that are formed by combining the same first and second
representations either with or without a weighting factor.
Accordingly, we find that the Examiner has adequately explained how
Calistri-Yeh, within the meaning of 35 U.S.C. § 102(b), describes a third and
fourth representation derived from the same initial text item as required by
claim 11, we affirm the Examiner’s anticipation rejection of claim 11.
Appellants do not separately argue dependent claims 12–14, 16, and 17.
(App. Br. 19–22.) Therefore, these claims fall with claim 11. 37 C.F.R.
§ 41.37(c)(1).

Appeal 2012-002280
Application 11/522,177

11
Claim 32
Appellants contend that claim 32 recites “changing the blend
constant” as well as “reweighting” of the first and second representations,
while Calistri-Yeh merely discloses at paragraph 0056 that a user may select
the alpha value which is used during the single execution of the TSV
formula to provide the single TSV for the data item per paragraph 0057.
Appellants contend that Calistri-Yeh fails to provide any teachings that
multiple TSVs are calculated for a given text item using multiple alpha
values and to the contrary merely discloses that a single TSV is calculated
for an individual data item. (App. Br. 27–28.)
We are not persuaded by Appellants’ contention. Calistri-Yeh
disclosed that there are instances in which additional data is added to
categories in a table requiring a recalculation of the TSV items (FF 9). The
TSV is calculated based on a formula requiring a variable weighting factor,
and as recognized by Calistri-Yeh the use of the weighting factor produced
superior results (FF 2, 9; Calistri-Yeh 5: ¶ 56). “The variable α is a
weighting factor that can be determined based on the information being
represented and analyzed. . . . Other values [for the weighting factor] can be
selected depending on various factors such as the type and quantity of
information, or the level of detail necessary to represent the information.”
(Calistri-Yeh 5: ¶ 56; see also FF 1, 6–9). Most importantly, Calistri-Yeh
provides that the addition of additional data to a table requires the
recalculation of the TSV (FF 9). We agree with the Examiner that “by
simply changing ‘alpha’ value, one can create infinitely many vectors (e.g.
TSVs) that are different from each other depending on the level detail
Appeal 2012-002280
Application 11/522,177

12
desired at a first point in time and a different level of detail at a different
time.” (Ans. 42.)
Accordingly, we find that the Examiner has adequately explained how
Calistri-Yeh, within the meaning of 35 U.S.C. § 102(b), discloses a third and
fourth representation using different weighting factors as required by claim
32, and we affirm the Examiner’s anticipation rejection of claim 32.
Appellants do not separately argue dependent claims 35 and 36. (App. Br.
26–29.) Therefore, these claims fall with claim 32. 37 C.F.R. § 41.37(c)(1).

II. The Issue: Obviousness over Calistri-Yeh and Ukrainczyk
The Examiner finds that Calistri-Yeh teaches most of the recited claim
limitations of claim 18 (see Ans. 17–19). The Examiner, however,
acknowledges that Calistri-Yeh “fails to expressly disclose - and using the
processing circuitry, steering the generating of at least of the first and second
representations according to a perspective of interest of a user.” (Id. at 19)
(emphasis omitted.) The Examiner relies on Ukrainczyk for this teaching,
and concludes that Ukrainczyk “‘provide[s] an effective method for
classifying text in which user knowledge may be used’, in other words
Ukrainczyk improves upon the teaching of Calistri-Yeh by incorporating
user input (i.e. word list and concepts) in order to improve classification of
texts or documents.” (Ans. 40–41.)
Appellants contend that the combination of references fails to suggest
“weighting of values from plural vectors of the same text item which are
generated according to different perspectives of interest of a user and
combining the weighted values of the first and second vectors as positively-
claimed.” (App. Br. 15.) Appellants contend that the combination fails to
Appeal 2012-002280
Application 11/522,177

13
suggest “the weighted values of one of these vectors with the weighted
values of the other of the vectors to produce a third vector as explicitly
claimed.” (App. Br. 16.)
Does the preponderance of evidence of record support the Examiner’s
conclusion that the combination of Calistri-Yeh and Ukrainczyk teaches
“generating of at least one of the first and second representations according
to a perspective of interest of a user”?

Findings of Fact
10. Ukrainczyk disclosed a method: “for automatically classifying
text into categories. . . . A weight is then coupled to each feature, wherein
the weight indicates a degree of association between the feature and the
category.” (Ukrainczyk Abstract.) “The REE table 50 consists of user-
inputted, concept evidence vector edits. Concept evidence vector edits are
user modified feature vectors 100 with associated flags and weights. The
information in REE table 50 is used to modify feature vectors 100 in topic
spotter matrix 90. (Ukrainczyk 3: ¶ 29; Ans. 19.)
11. Ukrainczyk disclosed:
Weight—the strength of the association between a term
(human and learned) and a concept node (manually or
automatically specified). In the case of human evidence, the
weight can be assigned by the system or it can be assigned by a
person. . . . . A user may assign the weights such that a feature
contributes to the decision to assign a document to a particular
concept node. . . . In the event a user does not specify the
weight, it will be assigned based on the feature’s distribution in
the training data (possibly modulated by its prior probability of
occurrence in the language).
(Ukrainczyk 6: ¶ 52; Ans. 19.)
Appeal 2012-002280
Application 11/522,177

14
Principle of Law
“The combination of familiar elements according to known methods
is likely to be obvious when it does no more than yield predictable results.”
KSR Int’l Co. v. Teleflex Inc., 550 U.S. 398, 416 (2007). As expressly
recognized in KSR, any art recognized need or problem can provide a reason
for combining claim elements. Id.

Analysis
Claim 18
Appellants contend that Calistri-Yeh fails to disclose “a measurement
basis let alone plural acts of generating first and second representations of
the same text item using first and second measurement bases as claimed.”
(App. Br. 22; see also Reply Br. 5.)
We are not persuaded. Claim 18 requires “a first representation of a
text item using a first measurement basis” as well as “a second
representation of the text item using a second measurement basis.” The
Examiner interprets the “measurement basis” to be the formula that produces
values “u” and “v” in table 230 of Calistri-Yeh (Ans. 40; FF 3, 4, 6).
[T]he formula for vector “u” can fairly be equated to “first
measurement basis” that measures relative strength of each
word with respect to categories. The results as shown in figure
9 (e.g. (u(w1), cat2) = 0.74) and can fairly be interpreted as
“first representation” thus clearly Calistri-Yeh discloses
“generating a first representative of a text item using a first
measurement basis”. Furthermore the formula for vector “v”
can fairly be equated to “second measurement basis” that
measure the probability distribution of a data points occurrence
across all categories and the results as shown in figure 9 (e.g.
Appeal 2012-002280
Application 11/522,177

15
(v(w1), cat2) = .071) can fairly be interpreted as “second
representation.”
(Ans. 39–40.)
“[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). The Specification provides that
“measurement basis 58 may indicate a plurality of associations or
relationships of individual ones of the measurement features with respect to
the dimension anchors . . . . Rows of the array may be the measurement
features and columns of the array may be the dimension anchors.” (Spec.
16: ¶ 49 (emphasis added).) The Specification, therefore, reasonably
provides that the measurement basis can indicate a relationship based on
multiple measurement features or based on a single measurement feature.
We interpret that based on the Specification a “measurement basis” can
reasonably be calculated from an individual (single) “measurement feature,”
such as a word.
We recognize, but are not persuaded by, Appellants’ contention that
“u” and “v” of Calistri-Yeh are “both void of array teachings let alone
teachings of arrays including measurement features, dimension anchors and
association values which associate the two.” (Reply Br. 7.) As discussed
above “measurement features,” in light of the Specification, can reasonably
be interpreted to encompass a single feature. Thus, we find that the
Examiner has reasonably interpreted word (w1) found in table 230 of
Calistri-Yeh to be the measurement feature and that formulas resulting in the
vectors “u” and “v” are reasonably interpreted to meet the “measurement
basis” requirement of the claim (Ans. 40).
Appeal 2012-002280
Application 11/522,177

16
We are also not persuaded by Appellants’ contention that the
Examiner has failed to articulate a reason for making the combination, and
that Calistri-Yeh teaches away from the combination (App. Br. 8–9). The
Examiner explains that “Calistri-Yeh and Ukrainczyk both deal with
clustering and/or organizing document by categories/similarities” (Ans. 40).
“The combination of familiar elements according to known methods is likely
to be obvious when it does no more than yield predictable results.” KSR,
550 U.S. at 416. Calistri-Yeh disclosed an automated method for organizing
text (FF 1–9), while Ukrainczyk also disclosed a method for organizing text
(FF 10–11). Ukrainczyk’s method provides that the organization can be
either made automatically, supplemented manually by the user, or may
utilize a combination to achieve the desired level of organization (FF 10–
11). Ukrainczyk acknowledges problems with the “prior art systems that
utilize training data, most do not have the capability to interactively take
advantage of human knowledge” (Ukrainczyk 1: ¶ 10). Thus, the inability to
interact with users when organizing data and utilize their knowledge to
improve data organization is an art recognized problem. As expressly
recognized in KSR, any art recognized need or problem can provide a reason
for combining claim elements. KSR, 550 U.S. at 416. Here, Ukrainczyk
“provide[s] an effective method for classifying text in which user knowledge
may be utilized very early in the construction of the statistical model”
(Ukrainczyk 2: ¶ 11; FF 10–11). Accordingly, we find that the Examiner
has met the burden of articulating a reason for making the combination
based on Calistri-Yeh and Ukrainczyk.
A reference is said to “teach away” from a claimed invention when it
“suggests that the line of development flowing from the reference’s
Appeal 2012-002280
Application 11/522,177

17
disclosure is unlikely to be productive of the result sought by the applicant.”
In re Gurley, 27 F.3d 551, 553 (Fed. Cir. 1994). That is not the case here. If
anything, these references merely illustrate the trade-offs that those of
ordinary skill in this art recognize and routinely consider in designing
computer based organizational applications.
We find that the Examiner has adequately articulated how the
combination of Calistri-Yeh and Ukrainczyk meets the limitation of the text
analysis method as set out in claim 18. For the reasons set out above and
those articulated by the Examiner in the Answer, we affirm the Examiner’s
obviousness rejection of claim 18. Appellants do not separately argue
claims 19–23, 25, 27–30, and 41. (App. Br. 22–26.) Therefore, these claims
fall with claim 18. 37 C.F.R. § 41.37(c)(1).

Claims 42
Appellants contend that the combination of Calistri-Yeh and
Ukrainczyk does not disclose the limitation of “accessing at least one text
pattern of interest to the user” as recited in claim 42 (App. Br. 33–34).
We are not persuaded. Ukrainczyk describes that “[t]he method
comprises the steps of manually or automatically associating a plurality of
features with at least one category, wherein the plurality of features
contribute to a decision to classify a document into said at least one
category.” (Ukrainczyk 2: ¶ 13 (emphasis added.) As explained by the
Examiner the recorded evidence edits (REE) table of Ukrainczyk contains
user input and is used to modify the evidence vectors (Ans. 43–44; FF 10).
We agree with the Examiner’s conclusion that the user input as disclosed in
Appeal 2012-002280
Application 11/522,177

18
Ukrainczyk meets the limitation of providing a search based on the interest
of the user.

Claim 43
Appellants contend that the combination of Calistri-Yeh and
Ukrainczyk does not meet the limitation of “associations of a plurality of
measurement features” as recited in claim 43 (App. Br. 34–36).
With respect to claim 43, we find that Appellants have the better
position. Although we agree with the Examiner that Calistri-Yeh discloses a
first and second measurement basis, we do not agree with the position that
the first and second measurement basis as disclosed in Calistri-Yeh is based
on “a plurality of measurement features.” In sum, we are not persuaded that
the Examiner has adequately explained how Calistri-Yeh’s use of a single
measurement feature in combination with Ukrainczyk teaches the limitation
of a measurement basis based on a plurality of measurement features.
Accordingly, we reverse the rejection of claim 43.

Claim 44
Appellants contend that the combination of Calistri-Yeh and
Ukrainczyk does not meet the limitation “generating the one of the first
measurement basis and the second measurement basis wherein the at least
one text pattern . . . comprises one of the measurement features and one of
the dimension anchors” (App. Br. 36–37).
We are not persuaded. As explained above (Claim 18) we agree with
the Examiner that the combination of Calistri-Yeh and Ukrainczyk teaches
utilizing at least one text pattern of interest to the user in order to organize
Appeal 2012-002280
Application 11/522,177

19
texts (Ans. 47; FF 10–11). We also agree with the Examiner that Calistri-
Yeh teaches generating a first and second measurement basis from a text
(e.g., word) (FF 6–8). “As shown in figure 9, ‘measurement basis’ clearly
include ‘measurement feature’ (e.g. word) and ‘dimension anchors’ (e.g.
categories)” (Ans. 48). We conclude that the evidence cited by the
Examiner supports a prima facie case of obviousness with respect to
claim 44.

III. The Issue: Obviousness over Calistri-Yeh and Fox
The Examiner has rejected claim 34 based on the combination of
Calistri-Yeh and Fox (Ans. 36–38). Appellants do not address the merits of
this rejection in their Appeal Brief or Reply Brief. “When the appellant fails
to contest a ground of rejection to the Board, . . . the Board may treat any
argument with respect to that ground of rejection as waived.” Hyatt v.
Dudas, 551 F.3d 1307, 1314 (Fed. Cir. 2008).

SUMMARY
We affirm the rejection of claims 11–14, 16, 17, 32, 35, and 36 under
35 U.S.C § 102(b) over by Calistri-Yeh.
We affirm the rejection of claims 18–23, 25, 29, 30, 41, 42, 44, and 45
under 35 U.S.C § 103(a) over Calistri-Yeh in view of Ukrainczyk.
We reverse the rejection of claim 43 under 35 U.S.C § 103(a) over
Calistri-Yeh in view of Ukrainczyk.
We affirm the rejection of 34 under 35 U.S.C. § 103(a) over Calistri-
Yeh in view of Fox.
Appeal 2012-002280
Application 11/522,177

20
Only those arguments actually made by Appellants have been
considered in this decision. Arguments which Appellants could have made
but chose not to make in the Brief have not been considered and are deemed
to be waived. See 37 C.F.R. § 41.37(c)(1).

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).

AFFIRMED-IN-PART

cdc