Ex Parte Howard et al

Board of Patent Appeals and Interferences

May 25, 2010

10404583 (B.P.A.I. May. 25, 2010)

UNITED STATES PATENT AND TRADEMARK OFFICE
____________

BEFORE THE BOARD OF PATENT APPEALS
AND INTERFERENCES
____________

Ex parte MICHAEL L. HOWARD and WILLIAM R. HARPER, JR.
____________

Appeal 2009-005947
Application 10/404,583
Technology Center 2400
____________

Decided: May 25, 2010
____________

Before JOHN A. JEFFERY, LEE E. BARRETT, and JAY P. LUCAS,
Administrative Patent Judges.

JEFFERY, Administrative Patent Judge.

DECISION ON APPEAL
Appellants appeal under 35 U.S.C. § 134(a) from the Examiner’s
rejection of claims 1-8, 10, 12-15, 17-25, 27, and 29-36. Claims 9, 11, 16,
26, and 28 have been canceled. Br. 3. We have jurisdiction under 35 U.S.C.
§ 6(b). We reverse.

Appeal 2009-005947
Application 10/404,583

2
STATEMENT OF THE CASE
Appellants invented a computer system related to rating web services
using historical data. See generally Spec. ¶¶ 01, 44. Claim 1 is reproduced
below with the key disputed limitations emphasized:
1. A system for rating web services, the system comprising:

a computer including a network interface for electronic
communications with a communications network, wherein the
computer is configured to process rating information input regarding a
web service and use the rating information input, and wherein the
computer is also configured to respond to a request for rating
information;

a plurality of web service identifications for identifying web
services;

a prediction engine for predicting behavior regarding the
performance of the web service based on prediction inputs;

web service information that relates to the web services, wherein
the web service information was obtained through use of the rating
information input, wherein both the web service information and the
prediction inputs comprise historical data obtained by the system;

a plurality of factors that determine an overall rating of the web
services;

one or more weights that correspond to the plurality of factors in
order to weight the factors according to user input;

a web services directory including a list of a plurality of web
services and a uniform resource locator (URL) for each of the web
services within the plurality of web services; and

a new web service, wherein the system does not have past
performance data for the new web service, wherein the system uses a
web service information database that is preloaded with related past
Appeal 2009-005947
Application 10/404,583

3
performance data of a related web service, wherein the related web
service provides similar services as the new web service, wherein the
system uses the related past performance data of the related web
service to rate the new web service.

The Examiner relies on the following as evidence of unpatentability:
Maltz US 2002/0141351 A1 Oct. 3, 2002
Moore US 2004/0122926 A1 June 24, 2004
(filed Dec. 23, 2002)

Schneider US 6,819,746 B1 Nov. 16, 2004
(filed July 10, 2000)

Fremantle US 2005/0278417 A1 Dec. 15, 2005
(filed Mar. 26, 2003)

THE REJECTIONS
1. The Examiner rejected claims 1-4, 6, 7, 10, 12-15, 17-21, 23, 24,
27, and 29-36 under 35 U.S.C. § 103(a) as unpatentable over Moore,
Fremantle, and Schneider. Ans. 3-13.1
2. The Examiner rejected claims 5, 8, 22, and 25 under 35 U.S.C.
§ 103(a) as unpatentable over Moore, Fremantle, Schneider, and Maltz.
Ans. 13-15.

OBVIOUSNESS REJECTION OVER MOORE, FREMANTLE, AND SCHNEIDER
Regarding independent claim 1, the Examiner finds that Moore
teaches all its limitations, except for the web service directory including a
uniform resource locator (URL) for each web service and rating a new web

1 Throughout this opinion, we refer to (1) the Appeal Brief filed January 9,
2008 and (2) the Examiner’s Answer mailed February 19, 2008.
Appeal 2009-005947
Application 10/404,583

4
service without past performance data. Ans. 3, 4, 16, 17. The Examiner
cites Fremantle to teach a known web service directory having URLs to
allow the client to interface with the service. Ans. 4. The Examiner further
states the combined Moore/Fremantle system does not teach rating new
services without performance data and relies on Schneider to teach this
limitation and to provide a basis for combining with the Moore/Fremantle
system so as to predict a service level. See Ans. 4-5.
Appellants argue that Schneider teaches a new loop or copper-wire
loop network and not a web service as commonly understood by an
ordinarily skilled artisan. Br. 11-13. Appellants further contend that the
“Moore and Fremantle are not analogous art to Schneider” (Br. 12), and
there is no motivation to combine Schneider with the Moore/Fremantle
system. Br. 12 and 14.
The issues before us, then, are as follows:

ISSUES
(1) Under § 103, has the Examiner erred in combining Moore,
Fremantle, and Schneider collectively to teach the invention in claim 1,
which calls for a new web service, wherein the system: (1) does not have
past performance data for the new web service and (3) uses related web
service’s past performance data of related web services provided similar
service to rate the new web service? The main issue turns on the following
pivotal subissues:

Appeal 2009-005947
Application 10/404,583

5
(a) Is Schneider analogous art?
(b) Has the Examiner provided some articulated reason having some
rational underpinning to combine Schneider with the Moore/Fremantle
system to support the Examiner’s obviousness conclusion?

FINDINGS OF FACT (FF)
1. The Specification states the invention relates to rating web
services. Spec. ¶ 01.
2. The Specification defines a “web service” as “an interface that
describes a collection of operations, services and/or data that is accessible
through a communication network.” Spec. ¶ 06.
3. Moore automates ranking of web services when a client queries a
search engine to perform a service. Moore’s search engine crawls contract
data to determine web services that match the client’s basic operating
requirements. The engine then crawls the reputation data 308 to determine
which contract-meeting web services have the best reputation and ranks the
services accordingly. Moore, ¶¶ 0007-09, 0030, 0035.
4. Moore’s reputation data 308 is empirically determined and
includes web services’ responsiveness, latency of a specific operation,
uptime indicating site or device availability, and business-oriented
behavioral attributes (e.g., cost to use a web service, business’
trustworthiness). Moore’ contract data or requirements relates to basic
operating requirements for web service type and whether a web service will
work with the client, such as interfaces to call, timing relationships, and
specifications for freeing the resource when finished. Moore, ¶¶ 0009,
0035-39.
Appeal 2009-005947
Application 10/404,583

6
5. Moore states the terms “web service” and “resource” are
equivalent and include external hardware devices and software and internal
software and hardware components. Moore, ¶ 0031, 0033.
6. Schneider discusses an expert system that classifies loops (e.g., a
copper-wire loops in a network) into one of several classes corresponding to
digital subscriber line (DSL) service grades (e.g., xDSL service grade)
offered through the network. In response to data characterizing a new loop
to be qualified, the system predicts a performance level using an operational
database that has loop characteristic and performance data. Schneider, col.
1, ll. 6-14 and 24-30, col. 4, ll. 1-15, and col. 6, ll. 19-30.
7. To predict a performance level for a new loop requesting xDSL
service, Schneider’s system uses pre-existing metrics and similar line
characteristic data (e.g., lengths, gauges, binder groups, and bridged taps)
from DSL lines already in service. The requested new loop is assigned an
appropriate DSL performance and service grade level. Schneider, col. 6, l.
63 – col. 7, l. 8.

PRINCIPLES OF LAW
As stated in In re Kahn, 441 F.3d 977 (Fed. Cir. 2006), “[t]he
analogous-art test requires that the Board show that a reference is either in
the field of the applicant's endeavor or is reasonably pertinent to the problem
with which the inventor was concerned in order to rely on that reference as a
basis for rejection.” Id. at 986-87 (internal citations omitted).

Appeal 2009-005947
Application 10/404,583

7
“It is impermissible to use the claimed invention as an instruction
manual or ‘template’ to piece together the teachings of the prior art so that
the claimed invention is rendered obvious . . . .” In re Fritch, 972 F.2d
1260, 1266 (Fed. Cir. 1992).

ANALYSIS
Based on the record before us, we find error in the Examiner’s
obviousness rejection of representative claim 1 which calls for, in pertinent
part, a new web service, wherein the system: (a) does not past performance
data for the new web service; (b) uses a web service information database
having related past performance data of a related web service that provides
similar services as the new web service; and (c) uses the related web
service’s past performance to rate the new web service. The Examiner relies
on Schneider to teach this limitation. Ans. 4-5. Appellants contend that
Schneider is nonanalogous art. Br. 12.
To demonstrate that Schneider is analogous, Schneider must be either
in the field of Appellants’ endeavor or be reasonably pertinent to the
problem with which the inventor was concerned. See Kahn, 441 F.3d at
986-87. Appellants’ invention relates to rating web services. FF 1.
Schneider discloses a system for classifying a new loop among one of
several xDSL service grades. FF 6. Because a loop provides an interface
for operations and services accessible through a communication network,
Schneider’s loop can be considered a “web service” in accordance with
Appellants’ definition. See FF 2 and 6. Moreover, by classifying

Appeal 2009-005947
Application 10/404,583

8
Schneider’s new loop, Schneider, in essence, rates a “web service” (e.g., a
new loop) over other web services (e.g., xDSL service grades) and, from this
perspective, is in Appellants’ field of endeavor. We therefore disagree with
Appellants (Br. 12), and find Schneider is analogous art.
However, this common aspect hardly provides a rational basis to
modify Moore/Fremantle system as the Examiner proposes. Ans. 4-5.
Moore automatically ranks web services, including resources, such as
hardware, when a user requests a service. See FF 3 and 5. Moore ranks the
web services first using web services’ contract data, and then using the
reputation data 308 about the web services. FF 3-4. Moore is silent
regarding how a new web service is ranked. See id. The Examiner however
finds that Moore rates both old and new services alike, by indicating that
Moore does not teach rating new web services without performance data.
See Ans. 16-17. Presuming the Examiner is correct, Moore would have
therefore, at best, taught or suggested rating new web services in the same
manner as old web services. That is, Moore uses the web service’s operating
requirements and reputation data. See FF 3-4. Moore therefore does not
suggest rating a web service, whether new or old, without containing some
past performance data about web services.
On the other hand, Schneider’s system uses pre-existing DSL’s
metrics and line characteristic data (e.g., lengths, gauges, binder groups, and
bridged taps) from similar DSL lines to assign a requested new loop (e.g., a
web service) to the appropriate DSL performance and service level. See FF
6-7. As explained above, Schneider teaches rating hardware (e.g., a loop) or
Appeal 2009-005947
Application 10/404,583

9
a “web service” among other hardware so as to predict its performance and
service level. However, we fail to find a reasonable basis to combine this
teaching with Moore’s system as the Examiner proposes.
First, Moore uses reputation data related to the web services’ latency,
responsiveness and business-oriented behavioral attributes for ranking. See
FF 4. Since Moore’s reputation data used for rating has no relationship to
Schneider’s metrics and line characteristics, we find Schneider’s teaching
would not have suggested to an ordinarily skilled artisan to use Schneider’s
DSL’s metrics and line characteristics to assist with rating Moore’s web
services.
Second, Schneider’s lines characteristics and metrics may arguably
relate to Moore’s contract data or basic operating requirements used to rank
the web services. See FF 4 and 7. But upon closer inspection, Moore’s
exemplary operating requirements include interfaces to call, timing
relationships, and specifications for freeing resources. Schneider’s line
metrics are simply not the same type of operating requirements that Moore
uses to rank web services, and thus fails to reasonably suggest to an
ordinarily skilled artisan using Schneider’s line characteristics and metrics to
assist in ranking any of Moore’s web services.
Moreover, even if these characteristics were considered to be the same
as Moore’s operating requirements, the predicting service level taught by
Schneider is not the actual service level the loop provides. That is,
Schneider prediction may be used to select a service in the ranking of web
services that does not actually meet the basic operating requirements of
Moore’s client. We thus, find that Schneider does not reasonably teach nor
suggest that its predicting technique using related service data would assist
Appeal 2009-005947
Application 10/404,583

10
or improve Moore’s system that ranks web services by first determining web
services that meet the client’s basic operating requirements (FF 3). See KSR
Int’l Co., v. Teleflex Inc., 550 U.S. 398, 417 (2007).
Additionally, combining these teachings would result in a system that
ranks new and old web services differently and would effectively change
Moore’s principle of operation by rating some web services (e.g., new web
services) without historical data. See In re Ratti, 270 F.2d 810, 813 (CCPA
1959). We thus, also find that the Examiner’s reason to combine the
references (i.e., to predict a service level (Ans. 4-5)) has no rational
underpinning on this record. See KSR, 550 U.S. 398, 418. In fact, the only
reason we can find to combine these teachings and arrive at the recited new
web service that has no past performance data and uses related past
performance data of related web services to rate the new web service as
recited in claim 1 is by impermissibly using the claim itself as an instruction
to piece the teachings together. In short, we see no rational basis to combine
the respective teachings of the cited references as the Examiner proposes
apart from impermissible hindsight reconstruction of the invention using
Appellants’ own invention as a blueprint. See Fritch, 972 F.2d at 1266.
Independent claims 17, 18 and 32 are commensurate in scope with
claim 1 and include a limitation to a new web service and the system using
the related past performance of related web service to rate the new web
service. We therefore find the Examiner erred in rejection claims 17, 18,
and 32 for the same reasons as discussed above in connection with claim 1.
For the foregoing reasons, Appellants have shown error in the
obviousness rejection of independent claims 1, 17, 18, and 32 based on the
combination of Moore, Fremantle, and Schneider. We therefore will not
Appeal 2009-005947
Application 10/404,583

11
sustain the rejection of claims 1, 17, 18, and 32, and dependent claims 2-4,
6, 7, 10, 12-15, 19-21, 23, 24, 27, 29-31, and 33-36 for similar reasons.

OBVIOUSNESS REJECTION OVER MOORE, FREMANTLE,
SCHNEIDER, AND MALTZ

Claims 5, 8, 22, and 25 depend from independent claims 1 and 18
respectively. We will therefore not sustain the rejection of these claims for
the same reasons previously discussed regarding claims 1 and 18.

CONCLUSION
The Examiner erred in rejecting claims 1-8, 10, 12-15, 17-25, 27, and
29-36 under § 103.

ORDER
The Examiner’s decision rejecting claims 1-8, 10, 12-15, 17-25, 27,
and 29-36 is reversed.

REVERSED

pgc

AUSTIN RAPP & HARDMAN
170 South Main Street, Suite 735
SALT LAKE CITY, UT 84101