Ex Parte Chasin

Patent Trial and Appeal Board

May 31, 2013

10888370 (P.T.A.B. May. 31, 2013)

UNITED STATES PATENT AND TRADEMARK OFFICE
________________

BEFORE THE PATENT TRIAL AND APPEAL BOARD
________________

Ex parte C. SCOTT CHASIN

________________

Appeal 2011-000114
Application 10/888,370
Technology Center 2400
________________

Before MARC S. HOFF, ELENI MANTIS MERCADER, and
JOHN G. NEW, Administrative Patent Judges.

NEW, Administrative Patent Judge.

DECISION ON APPEAL

Appeal 2011-000114
Application 10/888,370

2
SUMMARY
Appellant files this appeal under 35 U.S.C. § 134(a) from the
Examiner’s Final Rejection of claims 1-3, 5-10, 12, 13, and 15-23 as
unpatentable under 35 U.S.C. § 103(a).
1
Specifically, claims 1-3, 5-10, 12,
13, and 15-20 stand rejected as unpatentable under 35 U.S.C. § 103(a) as
being obvious over the combination of Riemers (US 6,615,242 B1,
September 2, 2003) (“Riemers”) and Matthew Woitaszek, Muhammad
Shaaban, and Roy Czernikowski, Identifying Junk Electronic Mail in
Microsoft Outlook with a Support Vector Machine, Proceedings of the 2003
Symposium on Applications and the Internet (SAINT’03) (IEEE, 2003)
(“Woitaszek”).
Claims 21-23 stand rejected as unpatentable under 35 U.S.C.
§ 103(a) as being obvious over the combination of Riemers, Woitaszek, and
Abkemeier (US 2003/0023736 A1, January 30, 2003) (“Abkemeier”).
We have jurisdiction under 35 U.S.C. § 6(b).
We AFFIRM.

NATURE OF THE CLAIMED INVENTION
Appellant’s invention is directed to a method for identifying e-mail
messages as being unwanted junk or spam. The method includes receiving
an e-mail message and then identifying contact and link data, such as URL
information, within the content of the received e-mail message. A blacklist
including contact information and/or link information previously associated
with spam is accessed, and the e-mail message is determined to be spam or

1
Claims 4, 11, and 14 are canceled. App. Br. 27, 28, 29.
Appeal 2011-000114
Application 10/888,370

3
to likely be spam based on the contents of the blacklist. The contact or link
data from the received e-mail is compared to similar information in the
blacklist to find a match, such as by comparing URL information from e-
mail content with URLs found previously in spam. If a match is not
identified, the URL information from the e-mail message is processed to
classify the URL as spam or "bad." The content indicated by the URL
information is accessed and spam classifiers or statistical tools are applied.
Abstract.

GROUPING OF CLAIMS

Because Appellant argues that the Examiner erred for substantially
the same reasons with respect to claims1-3, 5-10, 12, 13, and 15-23, we
select claim 1 as representative. App. Br. 14-15, 25. Claim 1 recites:
1. A method for identifying e-mail messages received over
a digital communications network as unwanted junk e-
mail or spam, comprising:

receiving an e-mail message;

identifying at least one of contact data and link data
within content of said received e-mail message;

accessing a list comprising at least one of contact
information and link information associated with previously-
identified spam and non-spam messages such that said
previously identified spam messages are differentiated from
said previously identified non-spam messages; and

determining whether said received e-mail message is
spam by applying a statistical method that determines a
probability that said received e-mail message is spam based on
how often elements of said received e-mail message appear in
Appeal 2011-000114
Application 10/888,370

4
said previously-identified spam and non-spam messages, and
information associated with said previously-identified spam and
non-spam messages, said information associated with said
previously-identified spam messages and said information
associated with said previously-identified nonspam messages
being obtained via said link information associated with said
previously identified spam and non-spam messages.

App. Br. 26.
Appellant argues that the Examiner erred for additional reasons with
respect to dependent claim 5. App. Br. 17. Claim 5 recites:
5. The method of claim 2, further comprising processing
content in said received email message by applying a
combination of a plurality of at least two different
statistical methods and other spam classifier methods.

App. Br. 27.
ISSUES AND ANALYSES
We address Appellant’s arguments seriatim, as presented in
Appellant’s Brief.

Issue 1
Appellant argues that the Examiner erred by improperly combining
Riemers with Woitaszek and Abkemeier, because Riemers would be
inoperable for its intended purpose of determining whether an e-mail is
spam based on the result of analysis using pre-determined strings and
associated pre-determined scores. App. Br. 17. We therefore address the
issue of whether the Examiner so erred.

Appeal 2011-000114
Application 10/888,370

5
Analysis
Appellant argues that although Riemers discloses a number of
methods of performing calculations after predetermined strings are found in
the received e-mail message, Riemers does not disclose or suggest
evaluating elements of the received e-mail message versus previously-
identified spam and/or non-spam messages to determine a probability that
the received e-mail message is spam. App. Br. 16 (citing Riemers, col. 4, ll.
1-27). Appellant contends that although some manipulation of the overall
message score is suggested by Riemers, no statistical analysis of elements
of the received e-mail message in comparison to previously-identified spam
and/or non-spam messages is suggested or contemplated. App. Br. 16.
According to Appellant, evaluating an element of the received e-mail
message based on how often the element appears in previously-identified
spam and/or non-spam messages using statistical methods is distinctly
different from performing calculations after predetermined strings with
predetermined scores are found in the received message. Id.
Consequently, Appellant argues, Riemers would be unfit for its
intended purpose if predetermined strings with pre-determined scores were
not used as the basis of the spam evaluation system that determines a score
indicating whether a received e-mail message is spam. App. Br. 16.
Appellant argues that changing the filter of Riemers to a Support Vector
Machine (“SVM”), as taught by Woitaszek, would alter the basic principle
of operation of Riemers’ system by removing the essential element of the
pre-determined strings and pre-determined scores. Id. Appellant contends
that the basic principle of operation of the Riemers reference would be
Appeal 2011-000114
Application 10/888,370

6
further altered by including means for storage of previously-identified spam
messages in order to permit the SVM to be trained. Id.
The Examiner responds, with respect to Appellant’s arguments
concerning the deficiencies of Riemers, that Appellant cannot show
nonobviousness by attacking references individually where the rejections
are based on combinations of references. Ans. 18 (citing In re Keller, 642
F.2d 413, 426 (C.C.P.A. 1981).
With respect to Appellant’s argument that combining the cited prior
art references would alter the basic principle of operation of Riemers, the
Examiner finds that Appellant provides no evidence, other than Appellant’s
unsupported assertion, that the basic principle of Riemers would be altered.
Ans. 19. The Examiner finds that Riemers does not suggest that no other
spam detection method can be used in conjunction with predetermined
strings with pre-determined score method. Ans. 19. The Examiner finds
further that the basic principle of both Riemers and Woitaszek teaches the
identification of email spam messages. Ans. 19-20. The Examiner finds
that an artisan of ordinary skill would employ the combined teachings of
Riemers and Woitaszek to achieve better prediction of messages and
reduction of false results, thereby improving the overall filtering. Ans. 20.
Moreover, the Examiner finds that Riemers does not explicitly teach or
implicitly suggest that his scoring technique cannot incorporate any other
statistical method to score URLs. Id.
The Examiner also finds that Riemers teaches or suggests storage
media, but does not teach or suggest that these storage media cannot
accommodate additional data. Ans. 20 (citing Riemers, col. 4, ll. 38-44).
The Examiner finds that the “basic principle of Riemers’ operation cannot
Appeal 2011-000114
Application 10/888,370

7
be altered just because an improvement requires extra hardware (memory as
argued by appellant).” Id.
We are persuaded by the Examiner’s reasoning and adopt it as our
own. Although we agree with Appellant’s contention that Appellant
contends that Riemers does not teach or suggest a statistical analysis of
elements of the received e-mail message in comparison to previously-
identified spam and/or non-spam messages is suggested or contemplated,
Woitaszek satisfies this limitation by teaching or suggesting the use of an
SVM to detect unsolicited commercial email. See Woitaszek 1. Nor are we
persuaded by Appellant’s unsupported assertions that incorporating the
SVM taught by Woitaszek would alter the basic function or intended
purpose of Riemers. Appellant adduces no evidence in support of their
contention that modification of Riemers with the method of Woitaszek
would render the prior art invention unsatisfactory for its intended purpose.
See In re Gordon, 733 F.2d 900, 902 (Fed. Cir. 1984). We agree with the
Examiner that both Woitaszek and Riemers teach the detection of spam
emails (see Ans. 21-22), and that it would have been obvious to one of
ordinary skill in the contemporaneous art to combine the SVM the
teachings of Woitaszek and Riemers to increase the efficiency of spam
email detection. We therefore conclude that the Examiner did not err by
improperly combining the cited prior art references.

Issue 2
Appellant argues that the Examiner erred in finding that the
combination of Riemers and Woitaszek teaches or suggests the limitation of
claim 5 reciting “processing content in said received email message by
Appeal 2011-000114
Application 10/888,370

8
applying a combination of a plurality of at least two different statistical
methods and other spam classifier methods.” App. Br. 17. We therefore
address the issue of whether the Examiner so erred.
Analysis
Appellant argues that although Riemers offers a number of variants
of the final message score based on performing calculations on the post-
identification scores for found strings, Riemers only offers a single
conceptual classifier for identifying spam e-mails, viz., finding pre-
determined strings that have pre-determined scores. App. Br. 18.
Appellant contends that Riemers therefore does not teach or suggest the use
of “additional statistical and other spam classifier methods.” App. Br. 19.
Furthermore, Appellant argues, at no point does Woitaszek disclose the use
of additional statistical methods and/or other classifier methods in
combination with the SVM. Id. Therefore, according to Appellant,
Woitaszek also does not disclose, teach or suggest the use of a combination
of a plurality of at least two different statistical methods and other classifier
methods to filter e-mail messages for spam. Id. Appellant argues further
that Abkemeier likewise does not address applying a plurality of statistical
methods and other spam classifier methods. Id.
The Examiner responds that Riemers teaches the use of a classifier
method using scoring utilizing predetermined strings approach and
Woitaszek discloses use of a support vector machine which is another
statistical method. As such, the combination of the two references meets
the requirement of the disputed limitation. Ans. 25. Moreover, the
Examiner finds that an artisan of ordinary skill would be motivated to
combine the references such that the SVM model taught by Woitaszek,
Appeal 2011-000114
Application 10/888,370

9
added to the basic design of Riemers, would improve message filtering
using combination of techniques disclosed by both references. Id.
We agree with the Examiner. All elements of each prior art reference
need not read on the claimed invention, rather, the proper test for
obviousness is what the combined teachings would have suggested to a
person of ordinary skill in the art. In re Kotzab, 217 F.3d 1365, 1370 (Fed.
Cir. 2000). As such, it is not necessary that a single prior art reference
suggest all of the requirements of a limitation, rather, the question is
whether the combination of the references would be obvious to an artisan of
ordinary skill. Appellant makes no substantial argument in support of the
contention that it would not have been obvious to combine the prior art
references, and we are persuaded by the Examiner’s argument that a person
of ordinary skill in the contemporaneous art would have been motivated to
combine the references to improve message filtering using a combination of
the techniques disclosed by both references. Ans. 25. We therefore
conclude that the Examiner did not err in finding that the combination of
Riemers and Woitaszek teaches or suggests the limitation of claim 5
reciting “processing content in said received email message by applying a
combination of a plurality of at least two different statistical methods and
other spam classifier methods.”

DECISION

The Examiner’s rejection of claims 1-3, 5-10, 12, 13, and 15-23 as
unpatentable under 35 U.S.C. § 103(a) is affirmed.

Appeal 2011-000114
Application 10/888,370

10
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).

AFFIRMED

ELD