Ex Parte Bates

Board of Patent Appeals and Interferences

Mar 30, 2011

11008807 (B.P.A.I. Mar. 30, 2011)

UNITED STATES PATENT AND TRADEMARK OFFICE
____________

BEFORE THE BOARD OF PATENT APPEALS
AND INTERFERENCES
____________

Ex parte CARY LEE BATES
____________

Appeal 2009-009998
Application 11/008,807
Technology Center 2100
____________

Before JEAN R. HOMERE, JOHN A. JEFFERY, and
ST. JOHN COURTENAY III, Administrative Patent Judges.

JEFFERY, Administrative Patent Judge.

DECISION ON APPEAL
Appellant appeals under 35 U.S.C. § 134(a) from the Examiner’s
rejection of claims 1-4, 7, 8, 11-13, 23-27, and 30-33. We have jurisdiction
under 35 U.S.C. § 6(b). We affirm.

STATEMENT OF THE CASE
Appellant’s invention suggests search engine keywords based on
results obtained via a query. See generally Abstract; Spec. 3. Claim 1 is
illustrative with key disputed limitations emphasized:
Appeal 2009-009998
Application 11/008,807

2
1. A computer-implemented method for performing a search, the
method comprising the steps of:
in response to a query that includes one or more keywords, generating
a result set identifying a plurality of results that match the query;
analyzing web pages in the result set to identify at least one additional
keyword that was not part of the query that would narrow the result set,
including identifying a first plurality of keywords omitted from the query
wherein inclusion of each of the first plurality of keywords in the query
would result in narrowing the result set by a respective first percentage,
ranking the first plurality of keywords based at least in part on the proximity
of the respective first percentage to 50%, outputting a ranked list of the first
plurality of keywords, and selecting the additional keyword from among the
ranked list in response to user input; and
narrowing the result set based upon the additional keyword.

The Examiner relies on the following as evidence of unpatentability:
Schuetze US 5,675,819 Oct. 7, 1997
Palmon US 2006/0059134 A1 Mar. 16, 2006
(eff. filed Sept. 10, 2004)

THE REJECTION
The Examiner rejected claims 1-4, 7, 8, 11-13, 23-27, and 30-33 under
35 U.S.C. § 103(a) as unpatentable over Palmon and Schuetze. Ans. 3-7.1

CONTENTIONS
Regarding representative claim 1, the Examiner finds that Palmon
discloses a search method with every recited feature except for analyzing

1 Throughout this opinion, we refer to (1) the Appeal Brief filed August 25,
2008; (2) the Examiner’s Answer mailed November 18, 2008; and (3) the
Reply Brief filed January 21, 2009.
Appeal 2009-009998
Application 11/008,807

3
web pages in a result set to identify narrowing keywords, but cites Schuetze
for teaching analyzing words in documents for narrowing purposes. Ans.
4-18. According to the Examiner, analyzing web pages of a result set as
claimed in lieu of Palmon’s analyzing web page attachments would have
been obvious in light of Palmon’s and Schuetze’s collective teachings. Id.
Appellant argues that the cited prior art does not teach or suggest (1)
analyzing web pages from a result set to identify additional keywords to
narrow the result set, and (2) ranking keywords identified via this analysis
based on how close those keywords would narrow the result set by 50%.
App. Br. 6-9. According to Appellant, Palmon uses known, defined
keyword/URL links unlike the claimed invention, and Schuetze does not
reference web pages at all, let alone analyze them to identify potential
narrowing keywords as claimed. Id. Appellant adds that the Examiner fails
to provide a reason why skilled artisans would modify Palmon’s searching
method using predefined keyword/URL links with Schuetze’s thesaurus
algorithm to arrive at the claimed invention. App. Br. 9.
The issues before us, then, are as follows:

ISSUES
(1) Under § 103, has the Examiner erred in rejecting claim 1 by
finding that Palmon and Schuetze collectively would have taught or
suggested a search method that (1) analyzes web pages from a result set to
identify additional keywords to narrow the result set, and (2) ranks keywords
identified via this analysis based at least in part on how close those
keywords would narrow the result set by 50%?
Appeal 2009-009998
Application 11/008,807

4
(2) Is the Examiner’s reason to combine the teachings of these
references supported by articulated reasoning with some rational
underpinning to justify the Examiner’s obviousness conclusion?

FINDINGS OF FACT (FF)
1. Palmon’s system assists searching for information on the internet
in which a searcher provides a search phrase or selects a “topic” (e.g., a
category or keyword) from a set of topics and, as a result, is provided with
(1) search results in the form of URLs that meet the search request, and (2) a
set of suggested topics to further the search. Each topic has “attachments” to
one or more “items of information” (e.g., web pages, publications, or
documents) that are appropriate for the topic, and the set of suggested topics
is generated by analyzing the attachments of the initial search results. In one
aspect, suggested topics include “refinement topics” that (1) narrow the
search, and (2) are determined according to certain criteria. Palmon, ¶¶
0003, 0005, 0017, 0032, 0034, 0079, 0055-70, 0076-98; Figs. 2-3.
2. Each topic is “attached to” (i.e., associated with) at least one URL
of a set of URLs that form the universe of searchable items. Palmon, ¶
0048; Fig. 1.
3. Palmon determines the set of refinement topics by considering the
set of identified information items {S:P} (i.e., the set of URLs) attached to a
particular search phrase (“S”) and search topic path (“P”). In one aspect, a
potential refinement topic has an attachment to at least one of the identified
information items (e.g., to at least one URL in the set {S:P}). Palmon, ¶¶
0162-63.
Appeal 2009-009998
Application 11/008,807

5
4. Palmon’s system also ranks potential refinement topics to
determine the better topic for a particular search. In one aspect, a
“refinement measure” is determined for each potential refinement topic,
where the refinement measure is the number of members in the set {S:P}
that are in attachments to the potential topic as a percentage of the size of
{S:P}. The best potential refinement topic is that whose refinement measure
is closest to 50%, which would split the results into two equally-sized sets of
URLs. The next best refinement topic is that whose refinement measure is
next closest to 50%, etc. Palmon, ¶ 0165.
5. Palmon’s Figure 6 shows a display for a search for the phrase
“Half Moon Bay” after the user consecutively selects corresponding
refinement topics (1) “Things to do”; (2) “Hiking”; and (3) “Trails.” At this
point in the search, the display includes (1) the current search topic path 603
(“Half Moon Bay > Things to do > Hiking > Trails”); (2) a set of refinement
topics 605 (“family trails,” “long trails,” “short trails,” etc.); and (3) the
results 607 for the first three URLs. Selecting (clicking) any displayed result
displays the web page defined by the URL. Palmon, ¶¶ 0197-0200; Fig. 6.
Palmon’s display for a search for the phrase “Half Moon Bay” after the user
selects corresponding refinement topics in Figure 6 is reproduced below:

Appeal 2009-009998
Application 11/008,807

6

Palmon’s Figure 6 displaying a search for the phrase “Half Moon Bay”
after the user selects corresponding refinement topics

6. Schuetze’s system improves retrieving relevant documents from a
corpus of documents based on a query by constructing a thesaurus based on
lexical co-occurrence of terms in the corpus. To this end, (1) each word’s
lexical co-occurrence pattern is presented as a “thesaurus vector”; (2) a
similarity measure is induced on words by comparing these vectors; and (3)
words’ synonyms are defined as the “nearest neighbors” with respect to the
similarity measure. Schuetze, Abstract; col. 1, ll. 7-12; col. 4, ll. 5-27.
7. Schuetze’s Figure 3 shows a display of an exemplary query
refinement technique that determines the nearest neighbors for terms in the
query “tank artillery.” In this example, the individual terms in the query
Appeal 2009-009998
Application 11/008,807

7
(“tank” and “artillery”) are added to the “pool,” and the “SEARCH
RESULT” section displays those terms’ nearest neighbors (e.g., “tanks,”
“armor,” “howitzers,” etc.). Schuetze, col. 12, ll. 44-61; Fig. 3.
8. Ambiguous terms (e.g., “tank”) used in queries can be clarified to
narrow the search using vector-based ambiguity resolution techniques.
Schuetze, col. 12, l. 62 – col. 13, l. 12; col. 24, ll. 1-11 (Table 2).

ANALYSIS
Based on the record before us, we find no error in the Examiner’s
obviousness rejection of representative claim 1 which calls for, in pertinent
part, a search method that (1) analyzes web pages from a result set to
identify additional keywords to narrow the result set, and (2) ranks keywords
identified via this analysis based on how close those keywords would narrow
the result set by 50%.
We agree with the Examiner (Ans. 9-10) that Palmon’s analyzing the
attachments to a web page (e.g., URLs) to determine “refinement topics,”
this analysis at least suggests analyzing web pages in the result set to
identify additional keywords to narrow the result set. See FF 1-5. Notably,
Palmon’s refinement topics are, by their very nature, intended to refine or
narrow the search. FF 1. And these refinement topics include keywords that
may or may not be part of the initial query. See FF 5 (displaying search
results and refinement topics (1) “Things to do”; (2) “Hiking”; and (3)
“Trails” for search query “Half Moon Bay”).
We also see no error in the Examiner’s reliance on Palmon’s
refinement topic ranking scheme (Ans. 15-16; FF 3-4) for at least suggesting
ranking keywords omitted from the initial search based at least in part on
Appeal 2009-009998
Application 11/008,807

8
how close those keywords would narrow the result set by 50%. As noted
above, the refinement topics’ keywords are not necessarily those that formed
the user’s initial query. See FF 1-5. Palmon’s scheme accounts for these
keywords and associated URLs in determining a refinement measure which
is used as a basis to designate the best refinement topic for a particular
search. FF 3-4. Notably, the best refinement topic is that whose refinement
measure is closest to 50% to split the results into two equally-sized sets of
URLs. FF 4.
Even assuming, without deciding, that Palmon uses known,
predefined keyword/URL links to identify additional keywords as Appellant
argues (App. Br. 8), nothing in the claim precludes this functionality. In
short, Appellant’s arguments in this regard are not commensurate with the
scope of the claim. In any event, to the extent that Appellant’s arguments
are predicated on the notion that Palmon’s analysis does not analyze web
pages, but rather their attachments, we find the Examiner’s position
regarding the obviousness of such a distinction persuasive. Ans. 9. At a
minimum, analyzing web pages to narrow an initial query would at least
suggest analyzing their corresponding attachments as the Examiner suggests.
Nevertheless, we find no error in the Examiner’s reliance on Schuetze
(Ans. 9-10; FF 6-8) to teach analyzing the text of documents to improve
search queries, and combining this teaching with Palmon to conclude that
analyzing the text of the web pages themselves in the disclosed search query
and results analysis would have been obvious. First, Palmon’s search
system is not limited to web pages, but also involves documents and
publications as the Examiner indicates (Ans. 10-11; FF 1). Since Schuetze’s
system improves document search and retrieval by refining queries (FF 6-8),
Appeal 2009-009998
Application 11/008,807

9
we find Schuetze’s and Palmon’s keyword-based document search and
analysis systems reasonably analogous. That Schuetze likewise teaches a
keyword-based analysis technique for narrowing searches (FF 8) only
bolsters this conclusion. In short, analyzing the text of documents in Palmon
in addition to analyzing web page attachments in view of Schuetze’s text-
based analysis scheme would have been obvious. We therefore find the
Examiner’s reason to combine the teachings of Palmon and Schuetze
supported by articulated reasoning with some rational underpinning to
justify the Examiner’s obviousness conclusion.
We are therefore not persuaded that the Examiner erred in rejecting
representative claim 1, and claims 2-4, 7, 8, 11-13, 23-27, and 30-33 not
separately argued with particularity.

CONCLUSION
The Examiner did not err in rejecting claims 1-4, 7, 8, 11-13, 23-27,
and 30-33 under § 103.

ORDER
The Examiner’s decision rejecting claims 1-4, 7, 8, 11-13, 23-27, and
30-33 is affirmed.

Appeal 2009-009998
Application 11/008,807

10
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

pgc