Ex Parte Singhal et al

Board of Patent Appeals and Interferences

Aug 30, 2010

10377117 (B.P.A.I. Aug. 30, 2010)

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.
10/377,117 03/03/2003 Amit Singhal 0026-0024 9085
44989 7590 08/31/2010
HARRITY & HARRITY, LLP
11350 Random Hills Road
SUITE 600
FAIRFAX, VA 22030
EXAMINER
PHAM, MICHAEL
ART UNIT PAPER NUMBER
2167
MAIL DATE DELIVERY MODE
08/31/2010 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 AMIT SINGHAL, MEHRAN SAHAMI, JOHN LAMPING,
MARCIN KASZKIEL, and MONIKA H. HENZINGER
____________________

Appeal 2009-004904
Application 10/377,117
Technology Center 2100
____________________

Before LANCE LEONARD BARRY, JEAN R. HOMERE, and
DEBRA K. STEPHENS, Administrative Patent Judges.

STEPHENS, Administrative Patent Judge.

DECISION ON APPEAL1

1 The two-month time period for filing an appeal or commencing a civil
action, as recited in 37 C.F.R. § 1.304, or for filing a request for rehearing,
as recited in 37 C.F.R. § 41.52, begins to run from the “MAIL DATE”
(paper delivery mode) or the “NOTIFICATION DATE” (electronic delivery
mode) shown on the PTOL-90A cover letter attached to this decision.

Appeal 2009-004904
Application 10/377,117

2
Appellants appeal under 35 U.S.C. § 134 from a rejection of claims 1-
10, 12-24, and 26-33. Claims 11 and 25 have been canceled (App. Br. 2).
We have jurisdiction under 35 U.S.C. § 6(b) (2008).
We REVERSE.

Introduction
Appellants’ invention is directed to search engines that locate
information on a network, such as the World Wide Web (Spec. 1). Alternate
terms are incorporated into a search query for a search engine (Abstract).
The alternate terms are semantically similar to the search query terms and
consider account information derived from the search query (id.).

STATEMENT OF CASE
Exemplary Claim
Claim 1 is an exemplary claim and is reproduced below:
1. A method of modifying a search query, the method
comprising:

determining one or more alternative terms for one or
more terms in the search query;

generating search results by searching an indexed corpus
of documents, the searching being based on the search query,
each search result identifying one or more documents in the
indexed corpus of documents;

defining a query context as a plurality of the documents
identified by the search results;

comparing the query context to the alternative terms to
generate one or more valid ones of the alternative terms: and
Appeal 2009-004904
Application 10/377,117

3

incorporating one or more of the valid ones of the
alternative terms into the search query to obtain a modified
search query.

Prior Art
The prior art relied upon by the Examiner in rejecting the claims on
appeal is:
Bowman US 6,169,986 B1 Jan. 2, 2001
Asmussen US 2002/0042923 A1 Apr. 11, 2002
Li US 2002/0059161 A1 May 16, 2002

REJECTIONS
Claims 1-5, 7, 8, 10, 12, 13, 19, 20, 22-24, 26-28, 30, 32, and 33 stand
rejected under 35 U.S.C. § 103(a) as being unpatentable over Bowman and
Asmussen (Ans. 3).
Claims 6, 9, 14-18, 21, 29, and 31 stand rejected under 35 U.S.C.
§ 103(a) as being unpatentable over Bowman, Asmussen, and Li (Ans. 19).

GROUPING OF CLAIMS

(1) Appellants argue claims 1, 2, 4, 5, 7, 8, 10, 12, and 13 as a
group on the basis of claim 1 (App. Br. 7-13). Claim 32 is argued on the
same basis as claim 1 (id. at 27-31). We also group claims 27, 28, and 30
with this group as analogous arguments are presented. Therefore, we select
independent claim 1 as the representative claim and treat claims 2, 4, 5, 7, 8,
10, 12, 13, 27, 28, 30, and 32 as standing or falling with claim 1.
Appeal 2009-004904
Application 10/377,117

4
(2) Appellants argue claim 3 independently (id. 13-15). Claim 26
is argued on the same basis as claim 3 (id. at 20-21). Therefore, we select
claim 3 as the representative claim and treat claim 26 as falling with
representative claim 3.
(3) Appellants argue claim 19, 20, and 22-24 as a group on the
basis of claim 19 (id. at 15-20). Claim 33 is argued on the same basis as
claim 19 (id. at 31-35). We select independent claim 19 as the
representative claim and treat claims 20, 22-24, and 33 as standing or falling
with representative claim 19.
(4) Appellants argue claim 14-18 as a group based on claim 14 (id.
at 36-38). Therefore, we select claim 14 as the representative claim and treat
claims 15-18 as standing or falling with representative claim 14.
(5) Appellants separately argue claims 6 and 9; claim 21; and claim
29 and 31 (id. at 36-38). We therefore address these claims separately
We accept Appellants’ grouping of the claims. See 37 C.F.R.
§ 41.37(c)(1)(vii).

ISSUE 1
35 U.S.C. § 103(a): claims 1, 2, 4, 5, 7, 8, 10, 12, 13, 27, 28, 30, and 32
Appellants assert their invention is not obvious over Bowman and
Asmussen because neither teaches or suggests “comparing the query context
to alternative terms to generate one or more valid ones of the alternative
terms,” as recited in claim 1 and commensurately recited in claims 27 and 32
(App. Br. 7-13 and 21-31). Specifically, Appellants contend the Examiner
misinterprets Asmussen since the search in Asmussen does not return a
plurality of documents but instead returns the historical analysis report that
Appeal 2009-004904
Application 10/377,117

5
is a summary of statistical data of the user’s previous account activities (id.
at 7-11 and 27-31; Reply Br. 4). Thus, Appellants argue historical data does
not correspond to the query context (App. Br. 10). Moreover, Appellants
contend that although Asmussen’s historical data may include returned
search results, “these search results are not based on a search query for
which one or more of the indexed suggestion keywords were determined”
(id.) (emphasis omitted). Appellants additionally contend “the user profile
and history data of A[smussen] will be not compared to the indexed
suggestion keywords to generate a list of suggested terms” (id.) (emphasis
omitted). Thus, Appellants argue Asmussen does not disclose or suggest
comparing the historical data that is relevant to the user to the indexed
suggestion keywords to generate valid indexed suggestion keywords and
thus, the recited element (id. at 8-11).
The Examiner finds plurality means anything greater than another and
documents are collections of characters (Ans. 25). Therefore, according to
the Examiner, the history analysis report of Asmussen teaches “a plurality of
documents returned by a search because it is a report of information
containing data regarding a search” (id. at 26). Further, the Examiner finds
the history data is compared to a suggestion database of indexed suggestion
keyword to provide suggested search parameters (id. at 26-27).

Issue 1: Have Appellants shown the Examiner erred in finding
Asmussen teaches “comparing the query context to the alternative terms to
generate one or more valid ones of the alternative terms” as recited in claim
1 and commensurately recited in claims 27 and 32?

Appeal 2009-004904
Application 10/377,117

6

FINDINGS OF FACT (FF)
Asmussen
(1) Asmussen teaches a method and system directed toward a
search suggestion engine and corresponding method that provides
suggestions of available content thus allowing a user to maximize use of a
video and digital multimedia aggregator. A search suggestion engine
constructs a database of metadata elements or word items using a crawler
that periodically or continually crawls a content metadata database,
extracting word items from the content metadata database. The word items
are then processed, indexed, and possibly provided with a vector value. A
subsequent search request may include search criteria provided with vector
values. (Abstract.) “A processor within the search suggestion engine
compares the vector values to determine if an indexed word item should be
considered for retrieving the indexed word item’s associated content.” (Id.)
A generated list of suggested content may be ranked and filtered before
being provided to the user. (Id.)
(2) The user history analysis report is a summary of
statistical analysis of the user’s previous account activities that
may include previous search requests, returned search results,
and content download requests. The history analysis report is
compiled by the user database server 511 within the database
administrator 502 and comprises information such as a
percentage breakdown of content types and genres for which a
user has requested searches, statistical breakdown of content
that the user has downloaded, and summary data about content
that was selected for download from earlier suggested content
results.

(Pg. 9, ¶ [0078]).
Appeal 2009-004904
Application 10/377,117

7
(3) The suggestion database processor 307 takes the vector-
formatted search request criteria and performs a comparison against the
index matrix, routine 123 to determine whether a particular index matrix
record falls within a suggestion vector range of the search criteria. (Pg. 9,
¶ [0079]).
(4) Once the comparison has been completed for the entire index
matrix, the cached records that fell within the suggestion vector range are
then filtered according to the user’s profile. The filtered results are then
forwarded and ranked according to the user’s historical analysis report, by
relevance, and by similarities to previously searched and downloaded
content. (Pg. 9, ¶ [0080]).
Dictionary
(5) A document is defined as “[a]ny self-contained piece of work
created with an application program and, if saved on disk, given a unique
filename by which it can be retrieved. Documents are generally thought of
as word-processed materials only. To a computer, however, data is nothing
more than a collection of characters, so a spreadsheet or a graphic is as much
a document as is a letter or report.” (Microsoft Computer Dictionary 171
(5th ed. 2002)).

ANALYSIS
The Examiner finds the history analysis report discloses a plurality of
documents that are returned by a search as the history analysis report is a
report of information containing data regarding a search (Ans. 26). Thus,
the Examiner finds the history analysis report is the recited query context
(id. at 25-26). We find, however, the history analysis report is not a plurality
Appeal 2009-004904
Application 10/377,117

8
of documents. The history analysis report is a summary of statistical
analysis of the user’s previous account activities (FF2). The user’s activities
include previous search reports, returned search results, and content
download requests; however, these activities are then statistically analyzed
and summarized (FF 2-3). Thus, we conclude one of ordinary skill in the art
would not have found the history analysis report to be a plurality of
documents identified by the search results. Instead, we find the history
analysis report is a document that provides a statistical analysis summary of
a user’s activities – not documents identified by the search results.
The Examiner seems to find that “history data” as discussed in
Asmussen and the history analysis report are the same, finding the history
data is compared to the suggestion database that houses the indexed
suggestion keywords (Ans. 26). However, we find no explanation or
support presented by the Examiner, or identification in the Asmussen
reference, to support this equivalence. Nor do we find that the history
analysis report is compared to the alternative terms to generate one or more
valid alternative terms. The history analysis report is compared to the result
to rank the results, not to generate valid alternative terms (FF 4).

CONCLUSION
Thus, after considering the totality of the circumstances before us,
Appellants have persuaded us of error in the Examiner’s finding that
Asmussen teaches “comparing the query context to the alternative terms to
generate one or more valid ones of the alternative terms” as recited in claim
1 and commensurately recited in claims 27 and 32. Accordingly, claims 2,
4, 5, 7, 8, 10, 12, 13, 27, 28, 30, and 32 grouped with claim 1, along with
Appeal 2009-004904
Application 10/377,117

9
claims 1, 27, and 32 are found to be non-obvious over Bowman and
Asmussen.

ISSUE 2
35 U.S.C. § 103(a): claims 19, 20, 22-24 and 26
Appellants assert their invention is not obvious over Bowman and
Asmussen because Asmussen does not teach or suggest “filtering the
additional terms using the query context to remove those of the additional
terms that are not semantically similar to the one or more search terms” as
recited in claim 19 and commensurately recited in claim 33 (App. Br. 15-
21).
The Examiner finds Asmussen teaches that the user history analysis
report includes search request, information, and summary data about content
selected for download from earlier suggested content results (Ans. 39-40).
The Examiner further finds the suggestion database a vector formatted
search request criteria and compares it to the index matrix (id. at 40). The
Examiner then contends that since the suggestion vector range is
automatically adjustable based on a “trend analysis of the user’s previous
download of suggested content” (id.) (emphasis omitted). The Examiner
thus finds that the comparison of word items in the suggestion database
using the history analysis report removes the word items in the suggestion
database that do not favorably compare to the vector quantity to one or more
search parameters that are not valid (id. at 41). Accordingly, the Examiner
finds Asmussen teaches “filtering the additional terms using the query
context to remove those of the additional terms that are not semantically
similar to the one or more search terms” as recited in claim 19. (Id.)
Appeal 2009-004904
Application 10/377,117

10
Issue 2: Have Appellants shown the Examiner erred in finding
Asmussen discloses “filtering the additional terms using the query context to
remove those of the additional terms that are not semantically similar to the
one or more search terms” as recited in claim 19 and commensurately
recited in claim 33?

FURTHER FINDINGS OF FACT (FF)
Asmussen
(6) A listing of all content metadata in the aggregator local
database 501 [is passed] to the suggestion keyword indexer
310. The suggestion keyword indexer 310 indexes each
metadata element. . . . [and] may assign each metadata element
with a vector quantity that describes the metadata element as a
suggestion keyword for a particular content entity. . . .The
indexer 310 then populates the suggestion database 308 with
these indexed suggestion keywords, which are subsequently
used by the suggestion database processor 307 along with
historical data relevant to the user, to supply a list of suggested
search terms.”

(Pg. 6-7, ¶ [0066].

ANALYSIS
As discussed above with respect to Issue 1, the Examiner has not
provided any persuasive argument or evidence, or identified the portion of
Asmussen that “historical data” is equivalent to “history analysis report.”
Therefore that historical data is used by the suggestion database processor is
not persuasive (FF 6). Moreover, we find Asmussen discloses that the
filtering takes place and the filtered results are then forwarded to rank the
results according to the history analysis report, relevance, and similarities to
Appeal 2009-004904
Application 10/377,117

11
previously searched and downloaded content (FF 4). We find ranking is not
filtering.

CONCLUSION
Accordingly, after considering the totality of the arguments and
evidence presented, Appellants have shown error in the Examiner’s finding
that Asmussen describes “filtering the additional terms using the query
context to remove those of the additional terms that are not semantically
similar to the one or more search terms” as recited in claim 19 and
commensurately recited in claim 33. Since claims 20 and 22-24 are grouped
with claim 19, we conclude claims 19, 20, 22-24, and 33 are not obvious
over Bowman and Asmussen.

ISSUE 3
35 U.S.C. § 103(a): claims 14-18
Appellants separately argue claims 14-18 (App. Br. 36-38). Since
claims 14-18 depend from claim 1 and the additional reference Li has not
been argued to cure the deficiencies of claim 1 as set forth in Issue 1 above,
we conclude claims 14-18 are not obvious over Bowman, Asmussen, and Li.

35 U.S.C. § 103(a): claims 6, 9, 21, 29, and 31
Appellants argue claims 6, 9, 21, 29, and 31 are patentable based on
their respective independent claims, but present no further arguments (App.
Br. 36, 38). As set forth above in Issues 2 and 3, their respective
independent claims 1, 19, and 27 have been shown to be patentable over
Bowman and Asmussen. The Examiner has not shown that Li cures the
Appeal 2009-004904
Application 10/377,117

12
deficiencies of Bowman and Asmussen. Therefore, we conclude claims 6, 9,
21, 29, and 31 are not obvious over Bowman, Asmussen, and Li.

DECISION
The Examiner’s rejection of claims 1-5, 7, 8, 10, 12, 13, 19, 20, 22-
24, 26-28, 30, 32, and 33 under 35 U.S.C. § 103(a) as being obvious over
Bowman and Asmussen is reversed.
The Examiner’s rejection of claims 6, 9, 14-18, 21, 29, and 31 under
35 U.S.C. § 103(a) as being obvious over Bowman, Asmussen, and Li is
reversed.

REVERSED

llw

HARRITY & HARRITY, LLP
11350 RANDOM HILLS ROAD
SUITE 600
FAIRFAX VA 22030