Ex Parte Burdick et al

Board of Patent Appeals and Interferences

Mar 2, 2012

11007247 (B.P.A.I. Mar. 2, 2012)

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.
11/007,247 12/09/2004 Brian Burdick MFCP.115021 9577
45809 7590 03/05/2012
SHOOK, HARDY & BACON L.L.P.
(MICROSOFT CORPORATION)
INTELLECTUAL PROPERTY DEPARTMENT
2555 GRAND BOULEVARD
KANSAS CITY, MO 64108-2613
EXAMINER
HICKS, MICHAEL J
ART UNIT PAPER NUMBER
2165
MAIL DATE DELIVERY MODE
03/05/2012 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 BRIAN BURDICK, JOSHUA J. FORMAN,
KEVIN P. KORNELSON, MURALI VAJJIRAVEL,
and RAJEEV PRASAD
____________

Appeal 2009-012633
Application 11/007,247
Technology Center 2100
____________

Before SCOTT R. BOALICK, CARL W. WHITEHEAD, JR.,
and DENISE M. POTHIER, Administrative Patent Judges.

POTHIER, Administrative Patent Judge.

DECISION ON APPEAL

Appeal 2009-012633
Application 11/007,247

2
STATEMENT OF THE CASE
Appellants appeal under 35 U.S.C. § 134(a) from the Examiner’s
rejection of claims 1-7, 9-18, and 20-33. Claims 8 and 19 have been
canceled. App. Br. 6. We have jurisdiction under 35 U.S.C. § 6(b). We
reverse but enter a new ground of rejection for claims 1-7, 9, and 10 under
37 C.F.R. § 41.50(b).

Invention
Appellants’ invention relates to selecting advertisements using
indexing and pre-filtering techniques in response to a user search query. See
Spec. ¶ 0003. Claim 1 is reproduced below with key disputed limitations
emphasized:
1. A delivery system for selecting advertisements for presentation to
a user in response to a user search query, the system comprising:
a keyword server for parsing the user search query, wherein the
keyword server parses the user search query into individual word
components and locates a linguistic locale identifier of the user search
query;
an index server for receiving the parsed search query, the index
server comprising an inverted index of advertising phrases and
corresponding linguistic local identifiers purchased by advertisers and
pre-filtering components for comparing the linguistic locale
identifiers and advertising phrases purchased by the advertisers to
each individual word component and [a] linguistic locale identifier of
the parsed user search query to discard non-matching linguistic locale
identifiers and advertising phrases purchased by the advertisers and
locate matching linguistic locale identifiers and advertising phrases
purchased by the advertisers;
a listing server for sorting through the matching linguistic locale
identifiers and advertising phrases purchased by the advertisers
located by the index server and further filtering the matching entries
for retrieval and presentation of associated advertisements to the user;
and
Appeal 2009-012633
Application 11/007,247

3
an advertisement server for retrieving, from an advertisement
source, advertisements corresponding to the purchased advertising
phrases sorted and filtered by the listing server.

The Examiner relies on the following as evidence of unpatentability:
Burrows US 5,745,890 Apr. 28, 1998
Koningstein US 2005/0228797 A1 Oct. 13, 2005
(filed Dec. 31, 2003)
Korry Douglas & Susan Douglas, POSTGRESQL: A COMPREHENSIVE GUIDE
TO BUILDING, PROGRAMMING, AND ADMINISTERING POSTGRESQL
DATABASES 73-80 (2003) (“Douglas”).

THE REJECTIONS
1. Claims 1, 9, and 10 stand rejected under 35 U.S.C. § 102(e) as
anticipated by Koningstein. Ans. 3-8.1
2. Claims 2-5, 11-16, 20-28, and 31-33 stand rejected under 35
U.S.C. § 103(a) as unpatentable over Koningstein and Douglas. Ans. 9-22.
3. Claims 6 and 7 stand rejected under 35 U.S.C. § 103(a) as
unpatentable over Koningstein and Burrows. Ans. 22-24.
4. Claims 17, 18, 29, and 30 stand rejected under 35 U.S.C. § 103(a)
as unpatentable over Koningstein, Douglas, and Burrows. Ans. 24-26.

THE ANTICIPATION REJECTION OVER KONINGSTEIN
Regarding independent claim 1, the Examiner finds that Koningstein
discloses an inverted index (e.g., 350) having both advertising phrases (e.g.,

1 Throughout this opinion, we refer to (1) the Appeal Brief filed March 3,
2009; (2) the Examiner’s Answer mailed May 13, 2009; and (3) the Reply
Brief filed June 23, 2009.
Appeal 2009-012633
Application 11/007,247

4
keywords 354) and corresponding linguistic locale2 identifiers (e.g.,
categories 352). See Ans. 5, 26-27. Appellants argue linguistic locale
identifiers correspond to specific geographic locales of interest to advertisers
(App. Br. 13) and that mapping Koningstein’s categories to the recited
linguistic locale identifier is not consistent with the ordinary meaning of
linguistic locale identifier (see Reply Br. 5-6). The Examiner further finds
that Koningstein discloses the recited pre-filtering components in different
paragraphs and refers particularly to Koningstein’s relevancy determination.
Ans. 27-28. Appellants argue that Koningstein’s ad system is not included
in the index server with the prefiltering components as recited. See App. Br.
13.

ISSUE
Under § 102, has the Examiner erred in rejecting claim 1 by finding
that Koningstein discloses pre-filtering components in the index server for
comparing the linguistic locale identifiers and advertising phrases purchased
by the advertisers to each individual word component and a linguistic locale
identifier of the parsed user search query?

ANALYSIS
Based on the record before us, we find error in the Examiner’s
anticipation rejection of claim 1. Although we disagree with Appellants that
the categories in Koningstein’s inverted index cannot reasonably be mapped
to linguistic locale identifiers as broadly as recited (see App. Br.13; Reply

2 Claim 1, line 9, mistakenly recites “local identifiers,” as evidenced by the
later recitations to “locale identifiers” in lines 10, 13, 14, and 16.
Appeal 2009-012633
Application 11/007,247

5
Br. 5-6), Koningstein does not disclose explicitly or implicitly the recited
prefiltering components in the index server for comparing the linguistic
locale identifiers and advertising phrases purchased by the advertisers to
each individual word component and linguistic locale identifiers of the
parsed user search query.
Regarding the limitation, “linguistic locale identifiers,” Appellants
assert that the plain meaning requires the identifier to correspond to a
geographic region. App. Br. 13; Reply Br. 5-6. However, Appellants have
not defined the phrase, “linguistic locale identifiers.” See generally
Specification. While the Specification describes a linguistic locale ID
(LCID) containing a value (e.g., 1033) and being associated with a
geographic region (e.g., United States) (see Spec. ¶¶ 0021, 0038, cited in
Reply Br. 6), we will not read this particular embodiment into the “linguistic
locale identifiers” recited claim 1. See SuperGuide Corp. v. DirecTV
Enters., Inc., 358 F.3d 870, 875 (Fed. Cir. 2004).
Koningstein discloses that a category 352 in its inverted index 350 is
used as a lookup key to associated keywords. ¶ 0069; Fig. 3. Each category
is a key that identifies or locates keywords or acts as a “locale” identifier.
Moreover, as the category contains words describing the keywords (e.g.,
automobiles (¶ 0091)), the category also contains a linguistic component and
thus can reasonably be considered a “linguistic locale identifier.”
Nonetheless, Koningstein’s index server does not contain prefiltering
components as recited and as Appellants contend (App. Br. 13).
The Examiner states the prefiltering components of Koningstein are
the relevancy determination operations discussed at Paragraphs 0045, 0049,
Appeal 2009-012633
Application 11/007,247

6
and 00523 and that the relevancy determination includes a look-up for
advertiser phrases in the inverted index as discussed at Paragraph 0069. See
Ans. 28. Koningstein discloses a system 120 that receives request for ads,
including a parsed search query and geolocation information (e.g., zip code,
area code), from a search engine or ad consumer 130, and the search engine
may combine the search results with an advertisement and forward the
results and ads to the user requested content. ¶¶ 0042, 0045-46, 0058; Fig.
1. Additionally, Koningstein teaches that an exemplary ad system (e.g.,
120’) may include relevancy determination operations 235 to determine
candidate ads for a given request. ¶¶ 0049, 0052. While the relevancy
operation prefilters ads (e.g., prefiltering components) and can arguably be a
component of the index server, there are no details in Koningstein about how
the relevancy determination is made, including whether there is any
comparing of the linguistic locale identifiers and advertising phrases to each
individual word component and linguistic locale identifier of the parsed user
search query.
Also, Koningstein discusses generating an inverted index 350 that
associates keywords 354 and categories 352 and that a category 352 serves
as a lookup key to associated keywords. ¶ 0069; Fig. 3. Koningstein further
discloses how the inverted index is generated and also how the index is used
to suggest or supply keywords for targeting an ad. ¶¶ 0071-72; Fig. 4. For
example, ad information 420 (e.g., keywords) are used to determine and
generate categories 450 at 410 (¶¶ 0071, 0090-92; Fig. 4), and keyword
suggestion/provision operations 460 may use the generated categories 450 as

3 In the Examiner’s Answer (Ans. 28), we believe the Examiner intended to
cite to Paragraph 0052, while mistakenly discussing Paragraph 0051.
Appeal 2009-012633
Application 11/007,247

7
keys to lookup keywords 354 in the inverted index to suggest determined
keywords to the user (¶ 0072; Fig. 4). While this filtering operation (e.g.,
determining keywords to suggest to a user) uses the category and keyword
with a keyword or individual word component of user search query to
suggest other keywords, Koningstein does not discuss also comparing the
linguistic locale identifiers and advertising phrases to both individual word
component and a linguistic locale identifier of the parsed user search query
as recited.
For the foregoing reasons, Appellants have persuaded us of error in
the anticipation rejection of independent claim 1and dependent claims 9 and
10 for similar reasons.

New Ground of Rejection Under 37 C.F.R. § 41.50(b)
However, under 37 C.F.R. § 41.50(b), we enter a new ground of
rejection under 35 U.S.C. § 112, first paragraph, for claims 1-7, 9, and 10 as
failing to satisfy the written description requirement. Appellants’ disclosure
does not convey with reasonable clarity to an ordinarily skilled artisan that
Appellants had possession of the claimed invention “the index server
comprising . . . pre-filtering components for comparing the linguistic locale
identifiers and advertising phrases purchased by the advertisers to each
individual word component and [a] linguistic locale identifier of the parsed
user search query” as of the filing date. See Ariad Pharms., Inc. v. Eli Lilly
& Co., 598 F.3d 1336, 1351 (Fed. Cir. 2010).
Appellants map the above-quoted limitation to Paragraphs 0039-41 in
the Specification for support. App. Br. 8. Paragraphs 0039 and 0040
describe the index 350 may be structured alphabetically based on a first
Appeal 2009-012633
Application 11/007,247

8
word 352 and a LCID 354, and an index search component 360 that
traverses the index 350 and examines each word of a user input query to
pass listings having a matching first word to the pre-filtering components
370. See Spec. ¶¶ 0039-40; Fig. 6. These paragraphs do not describe the
pre-filtering components; instead, they describe only that the information is
passed to the pre-filtering components. See id.
Paragraphs 0041-48 of the Specification describe the pre-filtering
components 370. Appellants state that pre-filtering components 370 include
a phrase-length pre-filtering component 372 and a word hash pre-filtering
component 374. See Spec. ¶ 0041; Fig. 3. The components 370 may
evaluate the terms contained within index 350, such as the first word
matches passed by the index search component 360. See id. This passage
supports that the pre-filtering components compare the linguistic identifier
and advertising phrases found in the inverted index 350 (see Spec. ¶ 0039;
Fig. 6) to an individual word of the parsed user search query (see Spec. ¶
0041; Fig. 3).
For example, Appellants discuss using a phrase match approach,
where if the user searches “BMW dealers in Chicago” and the advertiser
bids on “BMW dealers” as the phrase match, the pre-filtering component
370 will find a match. See Spec. ¶ 0044; Fig. 3. Similarly, Appellants’
other discussions of phrase length pre-filtering filters out advertisers’ bids by
evaluating words of the search query against the inverted index but fails to
mention comparing linguistic locale identifiers and advertising phrases to a
linguistic locale identifier of the parsed user search query as claim 1
requires. See Spec. ¶¶ 0042-45; Fig. 3. Notably, there is no discussion or
support in these passages for comparing a locale identifier and advertising
Appeal 2009-012633
Application 11/007,247

9
phrases to a linguistic locale identifier of the parsed user search query as
claim 1 requires.
Similarly, the word hash pre-filtering component 374 uses the index
350 to filter out advertisers’ bids. See Spec. ¶ 0047; Fig. 3. For example, an
advertiser bids on the phrase “Toshiba laptop sale,” and the user searches on
the phrase “Toshiba laptop price.” In this scenario, while the first two words
are the same, the third word will have different hashes, and the component
374 filters the listing. See Spec. ¶ 0048; Fig. 3. Again, there is no
discussion or support in these passages for comparing a locale identifier and
advertising phrases to a linguistic locale identifier of the parsed user search
query as claim 1 requires. Also, original claim 1 does not include this
limitation.
Thus, for the above reasons, Appellants’ disclosure does not convey
with reasonable clarity to an ordinarily skilled artisan that Appellants had
possession of the claimed invention “the index server comprising . . . pre-
filtering components for comparing the linguistic locale identifiers and
advertising phrases purchased by the advertisers to each individual word
component and [a] linguistic locale identifier of the parsed user search
query” as recited in claim 1 as of the filing date.
Claims 2-7, 9, and 10 depend from claim 1 and also include the
above-quoted limitation. We therefore also find that Appellants fail to meet
the written description requirement for these claims.

THE OBVIOUSNESS REJECTION OVER KONINGSTEIN AND DOUGLAS
Regarding independent claim 11, the Examiner finds that Koningstein
in Paragraphs 0045 and 0058-59 teaches and suggests storing metadata (e.g.,
Appeal 2009-012633
Application 11/007,247

10
information about the ad) in each index entry of the inverted index. Ans. 31.
The Examiner additionally finds Douglas’s teaching of a string matching
operation, when combined with Koningstein, further teaches employing such
operations to record and compare string lengths and query keywords used in
the filtering operations. See id. Appellants argue that Koningstein does not
store metadata as recited in each index entry (App. Br. 18) and that Douglas’
string matching operations fail to cure the deficiency in Koningstein. See id.
Regarding independent claim 23, the Examiner finds that
Koningstein’s discussion of geolocation information discloses demographic
information is parsed from a user search request and used to filter the
candidate matches. Ans. 19-20, 32-33. Appellants assert that Koningstein
discloses a search engine that receives advertisements for display and returns
search results to the user but does not teach or suggest traversing the
inverted index of content entries to match individual words and demographic
information and to pass content entries to a pre-filtering component as
recited. See Ans. 21. Appellants further contend that Douglas does not cure
this deficiency. See id.

ISSUES
Under § 103, has the Examiner erred by finding that Koningstein and
Douglas collectively would have taught or suggested:
(1) storing metadata that includes advertiser phrase length in number
of characters and advertiser phrase length in number of words in each index
entry relevant to each corresponding purchased advertiser phrase as recited
in claim 11?
Appeal 2009-012633
Application 11/007,247

11
(2) traversing an inverted index of contact entries to match each
individual word and the demographic information with any appropriate
content entries and to pass the content entries to a pre-filtering component as
recited in claim 23?
ANALYSIS
Claims 2-5
At the outset, we note that claims 2-5 depend from claim 1 and are
rejected based on Koningstein and Douglas under § 103. See Ans. 9-12.
The Examiner relies on Koningstein to teach the recited limitation of an
index server having pre-filtering components for comparing the linguistic
locale identifiers and advertising phrases purchased by the advertisers to
each individual word component and a linguistic locale identifier of the
parsed user search query. See id. As stated above, Koningstein does not
disclose the limitation of the index server having the pre-filtering
components as recited. Further, the Examiner does not rely on Douglas to
cure Koningstein’s deficiency. See id. We therefore will not sustain the
rejection of claims 2-5 for the above-stated reasons.

Claims 11-16 and 20-22
Based on the record before us, we find error in the Examiner’s
rejection of claim 11. As stated above, Koningstein discloses an inverted
index that includes index entries (e.g., categories 352, keywords 354). See
¶ 0069; Fig. 3. Koningstein further teaches that various information (e.g.,
information based on the query, information associated with the search
results) is submitted to the system 120 as a request for ads. See ¶¶ 0044-45.
Yet, even assuming that this information is metadata, Koningstein fails to
Appeal 2009-012633
Application 11/007,247

12
teach or suggest this information is stored in each index entry of the inverted
index. See id. Koningstein only teaches that the category and keyword
entries are part of the inverted index. See ¶ 0069; Fig. 3.
Also, we agree with Appellants that Douglas does not cure this
deficiency and combining Douglas’ teaching with Koningstein does not
describe or suggest storing metadata that includes advertiser phrase length in
number of characters and advertiser phrase length in number of words in
each index entry relevant to each corresponding purchased advertiser phrase.
App. Br. 18. That is, Douglas teaches comparing string values but does not
teach or suggest using or storing this comparison information as metadata in
any type of index. See Douglas 76-77. We therefore do not see how
combining Douglas’ teaching of performing a comparison of string values
with Koningstein teaches or suggests storing comparison information as
metadata that includes advertiser phrase length in number of characters and
advertiser phrase length in number of words in each of Koningstein’s index
entry relevant to each corresponding purchased advertiser phrase as recited.
For the foregoing reasons, Appellants have persuaded us of error in
the rejection of independent claim 11 and dependent claims 12-16 and 20-22
for similar reasons.

Claims 23-28 and 31-33
Also, based on the record before us, we find error in the Examiner’s
rejection of claim 23. The Examiner relies on Koningstein’s discussion of
geolocation information to teach the recited demographic information is
parsed from a user search request and used to traverse and filter the
candidate matches. See Ans. 19-20, 32-33. Koningstein discusses the
Appeal 2009-012633
Application 11/007,247

13
request for ads submitted to the system 120 can include a parsed query (e.g.,
individual words) as well as geolocation information. See ¶ 0045; Fig. 1.
However, even assuming that this geolocation information is demographic
information as recited, Koningstein fails to teach that the inverted index is
traversed to match each individual word and the demographic information
with any appropriate content entries and the entries are passed to a
pre-filtering component.
As discussed above, Koningstein discloses the inverted index contains
categories and keywords. See ¶ 0069; Fig. 3. Koningstein also teaches that
the inverted index is searched against keywords that could be obtained by
parsing the search query and to obtain suggested and/or automatically
supplied keywords. See ¶¶ 0045, 0069-72; Figs. 3-4. Koningstein teaches
and suggests traversing the index to match individual words from a user
search request with the content entries (e.g., keywords) in the index. See id.
But, Koningstein fails to teach or suggest also traversing the inverted index
to match the demographic information with appropriate entries and pass the
entries to a pre-filtering component as recited.
For the foregoing reasons, Appellants have persuaded us of error in
the rejection of independent claim 23 and dependent claims 24-28 and 31-33
for similar reasons.

THE REMAINING OBVIOUSNESS REJECTIONS
Claims 6, 7, 17, 18, 29, and 30 depend directly or indirectly from
independent claims 1, 11, and 23. The Examiner finds that: (1) Koningstein
and Burrows teach all the limitations in claims 6 and 7 (Ans. 22-24) and (2)
Koningstein, Douglas, and Burrows teach all the limitations of claims 17,
Appeal 2009-012633
Application 11/007,247

14
18, 29, and 30 (Ans. 24-26). The Examiner has not relied on Burrows to
cure Koningstein’s above-noted deficiencies for claims 1, 11, and 23. See
Ans. 22-26. Because we reverse the Examiner’s rejections of independent
claims 1, 11, and 23, we likewise will not sustain the rejections of claims 6,
7, 17, 18, 29, and 30.

CONCLUSION
The Examiner erred in rejecting claims 1, 9, and 10 under § 102 and
claims 2-7, 11-18, and 20-33 under § 103.
We newly reject claims 1-7, 9, and 10 under § 112, first paragraph.

DECISION
The Examiner’s decision rejecting claims 1-7, 9-18, and 20-33 is
reversed. We have entered a new ground of rejection under 37 C.F.R.
§ 41.50(b) for claims 1-7, 9, and 10.
Because this decision also contains a new ground of rejection pursuant
to 37 C.F.R. § 41.50(b), this section provides that “[a] new ground of
rejection . . . shall not be considered final for judicial review.”
Section 41.50(b) also provides that Appellants, WITHIN TWO
MONTHS FROM THE DATE OF THE DECISION, must exercise one of
the following two options with respect to the new ground of rejection to
avoid termination of the appeal as to the rejected claims:
(1) Reopen prosecution. Submit an appropriate
amendment of the claims so rejected or new evidence relating
to the claims so rejected, or both, and have the matter
reconsidered by the examiner, in which event the proceeding
will be remanded to the examiner. . . .
Appeal 2009-012633
Application 11/007,247

15
(2) Request rehearing. Request that the proceeding be
reheard under § 41.52 by the Board upon the same record. . . .
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).

REVERSED
37 C.F.R. § 41.50(b)

babc