Ex Parte Zhang

Board of Patent Appeals and Interferences

May 4, 2011

09895690 (B.P.A.I. May. 4, 2011)

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.
09/895,690 06/29/2001 Alex Xin Zhang 10010296 6868
7590 05/04/2011
HEWLETT-PACKARD COMPANY
Intellectual Property Administration
P.O. Box 272400
Fort Collins, CO 80527-2400
EXAMINER
MERCHANT, SHAHID R
ART UNIT PAPER NUMBER
3694
MAIL DATE DELIVERY MODE
05/04/2011 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 BOARD OF PATENT APPEALS
AND INTERFERENCES
____________

Ex parte ALEX XIN ZHANG
____________

Appeal 2010-003505
Application 09/895,690
Technology Center 3600
____________

Before: MURRIEL E. CRAWFORD, BIBHU R. MOHANTY, and
MICHAEL W. KIM, Administrative Patent Judges.

KIM, Administrative Patent Judge.

DECISION ON APPEAL
Appeal 2010-003505
Application 09/895,690

2
STATEMENT OF THE CASE
This is an appeal from the final rejection of claims 25-45. We have
jurisdiction to review the case under 35 U.S.C. §§ 134 and 6.
The claimed invention is generally directed to methods and systems
for auction-based simulation to extract a demand curve (Spec. 1:5-7). Claim
25, reproduced below, is further illustrative of the claimed subject matter.

25. A computer-implemented method for analyzing new
auction formats and rules comprising:
estimating, performed by a computer system, bidder
characteristics associated with potential auction participants;
estimating, performed by said computer system, a value
of an outcome of a market transaction to generate an estimated
value;
accessing, performed by said computer system, a set of
new auction rules and formats;
accessing, performed by said computer system, data from
a previous auction, said data associated with at least one of said
auction rules and formats;
modeling, performed by said computer system, a new
auction using said new auction rules, said formats and said
bidder characteristics to generate a post modeled auction value;
comparing, performed by said computer system, said post
modeled auction value with said estimated value; and
modifying, performed by said computer system, said
bidder characteristics and repeating said modeling, said
comparing and said modifying until said post modeled auction
value is said estimated value.1

1 We refer to the claims set forth in the Amendment and Response to Final
Office Action under 37 C.F.R. § 1.116(b) filed September 11, 2009, and
entered by the Examiner in the Office Communication mailed January 28,
2010. We do not refer to the claims set forth in the Appendix of the Appeal
Brief filed September 11, 2009.
Appeal 2010-003505
Application 09/895,690

3
Claims 25-45 stand rejected under 35 U.S.C. § 112, first paragraph,
for lack of enablement; claims 25, 29-31, 32, 36-38, 39, and 43-45 stand
rejected under 35 U.S.C. § 103(a) as unpatentable over White (US Pat.
6,088,676, iss. Jul. 11, 2000) in view of Hammond (US Pub. 2002/0082977
A1, pub. Jun. 27, 2002); and claims 26-28, 33-35, and 40-42 stand rejected
under 35 U.S.C. § 103(a) as unpatentable over White in view of Hammond
and Takriti (US Pat. 5,974,403, iss. Oct. 26, 1999).
We REVERSE.

ISSUES
Did the Examiner err in asserting that claims 25-45 are not enabled
under 35 U.S.C. § 112, first paragraph, because it is unclear what values the
recited notation sequence g1, g2,…, gk represent?
Did the Examiner err in asserting that a combination of White and
Hammond renders obvious “modifying, performed by said computer system,
said bidder characteristics and repeating said modeling, said comparing and
said modifying until said post modeled auction value is said estimated
value,” as recited in independent claims 25, 32, and 39?

FINDINGS OF FACT
Specification
The Specification discloses that in one embodiment, g1, g2,…, gk
represent new auction outcomes, where auction outcome g1 refers to the
highest bid; g2 refers to the number of recorded bidders; g3 refers to the
number of rejected bids; and gk refers to the second highest bid (16:6-8;
17:5).
Appeal 2010-003505
Application 09/895,690

4
The Specification also discloses that in step 406, it is determined
whether the newly recorded auction outcomes g1, g2,..., gk are tolerably
close to recorded auction data for statistical confidence. If not, process 400
loops back to repeat step 403 and subsequent steps, until statistical
confidence is obtained. In one embodiment, it may be typical to loop
through repeating steps 403-406 on the order of 10,000 times to achieve a
satisfactory statistical confidence level (17:7-12).
The Specification further discloses that if the known bidder
characteristic f and/or auction outcome g is not found to be tolerably close to
the observed values (step 408), then it is determined whether the assumed
bidder characteristics f1, f2,..., fn agree with the outcome. If they do not,
bidder characteristics f1, f2,..., fn are revised based on the sign and
difference of the recorded bidder characteristics f and auction outcomes g to
the known bidder characteristics f and auction outcomes g. Again, for the
example of g2 = 80 bidders, if the recorded number of bidders is 70, then the
total bidding population size (fn) or the percentage of bidders in a particular
price level (e.g., fl) would be increased (18:5-13).

White
White discloses that at block 28, the relevant data Z, from the past are
specified, wherein t represents a data index, such as a time period from 1 to
T + τ. According to White, each data point Zt is a vector of “‘dependent’”
variables (those to be predicted) and “independent” variables (those to be
used for prediction). For example, when the prediction models under test
are stock market prediction models, the data points Zt might include stock
Appeal 2010-003505
Application 09/895,690

5
market returns as a dependent variable and price trends, high/low prices,
returns on investment, price per earnings ratio, and so on as independent
variables (col. 7, ll. 13-22).
At block 30, an integer number P of prediction observation indices are
specified. According to White, each index t corresponds to an observed data
point (such as a past time period), wherein t=R,…, T and thus P=T-R+1. As
disclosed in detail below with reference to Figure 4, the data identified at
block 28 is associated with its appropriate data index R,...,T in a primary
data matrix. Essentially, the data indices R,...,T define respective matrix
row indices (col. 7, ll. 23-30).
White then discloses that the process then moves to block 32, wherein
a method is specified for generating model specifications indexed by model
indices k, wherein k is an integer=0,...,1. In other words, at block 32, a
method is specified for generating "1" prediction models (col. 7, ll. 31-35).
Moving to block 86, White discloses an efficacy factor p, which is a
statistic that represents the statistical significance of the effectiveness of the
best among the k prediction models relative to the benchmark, is determined.
White discloses that the efficacy factor p is the difference between unity and
the ratio of n to the total number of observation matrices. The efficacy
factor p is an estimate of a “p-value” associated with a statistical test of a
formal hypothesis that a best prediction model has expected performance no
better than that of a benchmark. The efficacy factor varies between zero and
one; the more statistically significant the “best” model vis-a-vis the
benchmark, the lower the p-value (col. 10, ll. 7-19).
The logic then moves to block 88 of White to increment k by one, and
then tests whether the last model has been evaluated at decision diamond 90.
Appeal 2010-003505
Application 09/895,690

6
If not, the logic loops back to block 72. Otherwise, the model ends at state
92 and returns the final value of the efficacy factor p (col. 10, ll. 20-24).

ANALYSIS
Enablement
We are persuaded the Examiner erred in asserting that claims 25-45
are not enabled under 35 U.S.C. § 112, first paragraph, because it is unclear
what values the recited notation sequence g1, g2,…, gk represent (App. Br.
8-12; Reply Br. 1-2). The Examiner is correct that usually, a sequence
notation of g1, g2,…, gk would represent different values having the same
format, e.g., they all represent different values of dollars (Exam’r’s Ans. 9-
11). However, Appellant is entitled to be his/her own lexicographer, and the
Specification clearly defines g1, g2,…, gk as various new auction outcomes,
such as bids and bidders. See In re Paulsen, 30 F.3d 1475, 1480 (Fed. Cir.
1994) (inventor may define specific terms used to describe invention, but
must do so “with reasonable clarity, deliberateness, and precision” and, if
done, must “set out his uncommon definition in some manner within the
patent disclosure' so as to give one of ordinary skill in the art notice of the
change” in meaning) (quoting Intellicall, Inc. v. Phonometrics, Inc., 952
F.2d 1384, 1387-88 (Fed. Cir. 1992)). While Appellant’s use of such a
sequence notation may be unorthodox, it is enabling. Upon reading the
Specification, one of ordinary skill would understand that changing bidder
characteristics f1, f2,…, fn, one ends up with different new auction
outcomes g1, g2,…, gk, where n and k are decoupled (i.e., f1 and g1 can
have unrelated values and formats).
Appeal 2010-003505
Application 09/895,690

7
The Examiner also asserts that by using the sequence notation g1,
g2,…, gk, Appellant is admitting that g4, g5,…, gk exist, yet, because of the
unorthodox format, one of ordinary skill does not know what kind of
format/value to apply to notations following g3 (Exam’r’s Ans. 9-11).
However, Appellant has defined g3, which could potentially be the only
notation between g2 and gk. Accordingly, while again, unorthodox,
Appellant’s disclosure of at least one intermediate notation, g3, means that
the Specification is enabled with respect to the sequence notation g1, g2,…,
gk.

Obviousness
We are persuaded the Examiner erred in asserting that a combination
of White and Hammond renders obvious “modifying, performed by said
computer system, said bidder characteristics and repeating said modeling,
said comparing and said modifying until said post modeled auction value is
said estimated value,” as recited in independent claims 25, 32, and 39 (App.
Br. 13-25; Reply Br. 2-5). Essentially, the aforementioned portions of
independent claims 25, 32, and 39 disclose holding the new auction rules
and formats (model) constant, and modifying the bidder characteristics
(variables) until a desired outcome (estimated value) is reached. By
contrast, White discloses holding the relevant data Z (variables) constant,
and plugging this relevant data Z into different models k by regularly
increasing k by 1 in block 88, until the best efficacy factor p is calculated for
a particular model k given relevant data Z. Accordingly, because
independent claims 25, 32, and 39 and White hold opposing portions of the
model/variables constant, we will not sustain this rejection.
Appeal 2010-003505
Application 09/895,690

8
The decision of the Examiner is REVERSED.

REVERSED
Appeal 2010-003505
Application 09/895,690

9

hh