Ex Parte Chao

Board of Patent Appeals and Interferences

Jun 15, 2011

10410999 (B.P.A.I. Jun. 15, 2011)

UNITED STATES PATENT AND TRADEMARK OFFICE
____________
BEFORE THE BOARD OF PATENT APPEALS
AND INTERFERENCES
____________
Ex parte HUI CHAO
____________
Appeal 2009-011629
Application 10/410,9991
Technology Center 2100
____________

Before DENISE M. POTHIER, DAVID M. KOHUT, and MICHAEL R.
ZECHER, Administrative Patent Judges.

ZECHER, Administrative Patent Judge.

DECISION ON APPEAL

1 Filed on April 10, 2003. The real party in interest is Hewlett-Packard
Development Co., L.P. (App. Br. 1.)
Appeal 2009-011629
Application 10/410,999

2
I. STATEMENT OF THE CASE
Appellant appeals under 35 U.S.C. § 134(a) (2002) from the
Examiner’s final rejection of claims 1, 3, 5, 6, 8-10, 13, 14, 16, 23-33, 35
and 36. (App. Br. 1-2; Reply Br. 1.) Claims 2, 4, 11, 12, and 19-22 have
been cancelled. (Id.) Claim 7 has been allowed. (Id.) Claims 15, 17, 18,
and 34 are objected to as being dependent upon a rejected base claim, but
would be allowable if rewritten in independent form including all of the
limitations of the base claim and any intervening claims. (Id.) We have
jurisdiction under 35 U.S.C. § 6(b) (2008).
We affirm.
Appellant’s Invention
Appellant invented a method, computer-readable medium, and system
for classifying document elements into categories, such as content and
background. (Spec. 1, ll. 4-6.)
Illustrative Claims
Independent claims 1, 5, and 13 further illustrate the invention as
follows:
1. A computer-implemented method of classifying
elements of a document, comprising:
receiving a data file defining an arrangement of elements
on a page of the document, wherein the data file associates
respective attribute values with each of the elements;
selecting one or more of the elements on the page;
determining for each of the selected elements a
respective content type value, at least one respective size value
indicating a respective amount of the page occupied by the
element, and at least one respective distance value indicating a
Appeal 2009-011629
Application 10/410,999

3
respective location of the element in relation to a reference
location on the page from the associated attribute values;
respectively assigning each of the selected elements to
either a content category or a background category, wherein
each assignment of a respective one of the selected elements is
based on the respective content type value and at least one of (i)
a comparison of the at least one respective size value with a
respective threshold value and (ii) the at least one respective
distance value;
generating a classification record comprising for each of
the selected elements an indication of the respectively assigned
category; and
storing the classification record in a computer-readable
storage medium.

5. A computer-implemented method of classifying
elements of a document, comprising:
receiving a data file defining the document, wherein the
document has at least one page and a plurality of elements,
wherein a type, location, and size of each element is defined by
the file;
classifying at least one selected element of a selected
page into one of a plurality of categories based on at least one
of the type, location, size, and dimension of the selected
element, wherein the classifying comprises determining
whether, at least along one dimension, the selected element has
a size in relation to that of the page that exceeds a
pre-determined threshold and assigning the selected element to
a background element category in response to a determination
that the selected element has a size in relation to that of the
page that exceeds the pre-determined threshold;
generating a classification record comprising an
indication of the category into which each of the selected
elements is respectively classified; and
storing the classification record in a computer-readable
storage medium.
Appeal 2009-011629
Application 10/410,999

4

13. A computer -implemented method of classifying
elements of a document, comprising:
receiving a data file defining a respective arrangement of
elements on each of a plurality of pages of the document,
wherein the data file associates respective attribute values with
each of the elements;
classifying a selected one of the elements on a selected
one of the pages into one of a content category and a
background category based on frequency of occurrence of
similar elements on one or more pages of the document other
than the selected page;
generating a classification record comprising an
indication of the category into which the selected element is
respectively classified; and
storing the classification record in a computer-readable
storage medium.

Prior Art Relied Upon
The Examiner relies on the following prior art as evidence of
unpatentability:
Pirolli US 5,895,470 Apr. 20, 1999
Fan US 2003/0072487 A1 Apr. 17, 2003
Wu US 2004/0066530 A1 Apr. 8, 2004
Bottou US 2005/0207666 A1 Sept. 22, 2005
(effectively filed Jan. 24, 2002

Rejections on Appeal
The Examiner rejects the claims on appeal as follows:
Claims 1, 3, 8-10, 26-33, and 36 stand rejected under 35 U.S.C.
§ 103(a) as being unpatentable over the combination of Wu and Bottou.
Claims 5 and 6 stand rejected under 35 U.S.C. § 103(a) as being
unpatentable over the combination of Wu and Fan.
Appeal 2009-011629
Application 10/410,999

5
Claims 13, 14, 16, and 23-25 stand rejected under 35 U.S.C. § 103(a)
as being unpatentable over the combination of Wu, Bottou, and Pirolli.
Claim 35 stands rejected under 35 U.S.C. § 103(a) as being
unpatentable over the combination of Wu, Bottou, and Fan.
Appellant’s Contentions
1. Appellant contends that Bottou’s disclosure of classifying
image elements is performed independently of content type and page
location value and, therefore, does not teach or fairly suggest “each
assignment of a respective one of the selected elements is based on the
respective content type value,” as recited in independent claim 1. (App. Br.
11; Reply Br. 2-3.) Further, Appellant argues that Bottou’s disclosure of a
ratio between unoccluded and occluded perimeters of an element does not
constitute a size value indicating a respective amount of page occupied by an
element. (App. Br. 11; Reply Br. 3-4.) Moreover, Appellant alleges that
Bottou’s disclosure of image element coordinates does not amount to a
“distance value,” as claimed. (App. Br. 12; Reply Br. 4-5.) Therefore,
Appellant alleges that Bottou does not teach or fairly suggest “each
assignment of a respective one of the selected elements is based on … at
least one of (i) a comparison of the at least one respective size value with a
respective threshold value and (ii) the at least one respective distance value,”
as recited in independent claim 1. (App. Br. 11-12; Reply Br. 3-4.)
Additionally, Appellant contends that there is insufficient rationale to
combine Wu and Bottou, and the Examiner improperly engages in
impermissible hindsight reconstruction. (App. Br. 14-17; Reply Br. 5-7.)
2. Appellant contends that Fan’s disclosure of the dimensions of
the largest region grown does not contain any information about “a size in
Appeal 2009-011629
Application 10/410,999

6
relation to that of the page,” as recited in independent claims 5 and 6. (App.
Br. 19, 24-25; Reply Br. 7-8, 10-11.) Appellant also argues that Fan’s
largest region grown is an absolute size and does not teach or fairly suggest
“a size in relation to that of the page,” as claimed. (Reply Br. 8-11.)
Moreover, Appellant argues that Fan’s disclosure of comparing the
dimensions of the largest region grown with a threshold does not teach or
fairly suggest “determining whether, at least along one dimension, the
selected element has a size in relation to that of the page that exceeds a
pre-determined threshold,” as recited in independent claim 5 and as similarly
recited in independent claim 6. (App. Br. 19-20; Reply Br. 8-11.) Further,
Appellant alleges that there is insufficient rationale to combine Wu and Fan,
and the Examiner improperly engages in impermissible hindsight
reconstruction. (App. Br. 20-23, 25; Reply Br. 9, 11.)
3. Appellant contends that Pirolli’s disclosure of classifying entire
web pages does not amount to classifying individual elements on a web page
into one of a content or background category. (App. Br. 27-28; Reply Br.
11-13.) Therefore, Appellant argues that Pirolli does not teach or fairly
suggest “classifying a selected one of the elements on a selected one of the
pages into one of a content category and a background category based on
frequency of occurrence of similar elements on one or more pages of the
document other than the selected page,” as recited in independent claim 13.
(Id.) Further, Appellant alleges that there is insufficient rationale to combine
Wu, Bottou, and Pirolli, and the Examiner improperly engages in
impermissible hindsight reconstruction. (App. Br. 28-30; Reply Br. 13.)
Appeal 2009-011629
Application 10/410,999

7
Examiner’s Findings and Conclusions
1. The Examiner finds that Bottou’s disclosure of classifying an
image element as either a foreground or background layer teaches or fairly
suggests “respectively assigning each of the selected elements to either a
content category or a background category,” as recited in independent claim
1. (Ans. 21.) Further, the Examiner finds that Bottou’s disclosure of
determining the image element coordinates amounts to determining a
location of an image in relation to a reference location (i.e., edges) of a page.
(Id. at 21, 23.) Moreover, the Examiner finds that since independent claim 1
explicitly recites “at least one of,” Bottou only needs to teach or fairly
suggest one of the two alternatives claimed. (Id.) Consequently, the
Examiner finds that Bottou’s disclosure of classifying an image element
based on both the content type and image element coordinates teaches or
fairly suggests “wherein each assignment of a respective one of the selected
elements is based on the respective content type value and at least one
of…(ii) the at least one respective distance value,” as claimed. (Id. at 22.)
Further, the Examiner finds that there is sufficient rationale to combine Wu
and Bottou. (Id. at 23-25.)
2. The Examiner finds that Fan’s disclosure of comparing the
dimensions of the largest region grown with a predetermined threshold
teaches or fairly suggests “determining whether, at least along one
dimension, the selected element has a size in relation to that of the page that
exceeds a pre-determined threshold,” as recited in independent claim 5 and
as similarly recited in independent 6. (Id. at 26-27.) Further, the Examiner
disagrees with Appellant’s assertion the Fan’s largest region grown does not
constitute an area in relation to a page. (Id. at 26-28.) The Examiner finds
Appeal 2009-011629
Application 10/410,999

8
that Fan’s largest region grown amounts to a size in relation to a page
because it is the largest region or element on the page. (Id.) Additionally,
the Examiner finds that there is sufficient rationale to combine Wu and Fan.
(Id. at 28-29.)
3. The Examiner finds that Pirolli’s disclosure of classification
based on usage frequency and similarity of objects among several Web
pages is only relied upon to teach or fairly suggest “classifying… based on
frequency of occurrence of similar elements on one or more pages of the
document other than the selected page,” as recited in independent claim 11.
(Id. at 29-31.) The Examiner maintains that Bottou’s disclosure of
classifying an image element as either a foreground or background layer
teaches or fairly suggests “classifying a selected one of the elements on a
selected one of the pages into one of a content category and a background
category,” as claimed. (Id.) Further, the Examiner finds that there is
sufficient rationale to combine Wu, Bottou, and Pirolli. (Id. at 31.)

II. ISSUES
1. Has Appellant shown that the Examiner erred in concluding
that the combination of Wu and Bottou renders independent claim 1
unpatentable? In particular, the issue turns on whether:
(a) the combination of Wu and Bottou teaches or fairly suggests
“respectively assigning each of the selected elements to either a content
category or a background category, wherein each assignment of a respective
one of the selected elements is based on the respective content type value
and at least one of (i) a comparison of the at least one respective size value
Appeal 2009-011629
Application 10/410,999

9
with a respective threshold value and (ii) the at least one respective distance
value,” as recited in independent claim 1; and
(b) there is sufficient rationale to combine Wu and Bottou.
2. Has Appellant shown that the Examiner erred in concluding
that the combination of Wu and Fan renders independent claims 5 and 6
unpatentable? In particular, the issue turns on whether:
(a) the combination of Wu and Fan teaches or fairly suggests
“classifying at least one selected element of a selected page into one of a
plurality of categories based on at least one of the type, location, size, and
dimension of the selected element, wherein the classifying comprises
determining whether, at least along one dimension, the selected element has
a size in relation to that of the page that exceeds a pre-determined
threshold,” as recited in independent claim 5 and as similarly recited in
independent claim 6; and
(b) there is sufficient rationale to combine Wu and Fan.
3. Has Appellant shown that the Examiner erred in concluding
that the combination of Wu, Bottou, and Fan renders independent claim 13
unpatentable? In particular, the issue turns on whether:
(a) the combination of Wu, Bottou, and Fan teaches or fairly
suggests “classifying a selected one of the elements on a selected one of the
pages into one of a content category and a background category based on
frequency of occurrence of similar elements on one or more pages of the
document other than the selected page,” as recited in independent claim 13;
and
(b) there is sufficient rationale to combine Wu, Bottou, and Fan.

Appeal 2009-011629
Application 10/410,999

10
III. FINDINGS OF FACT
The following Findings of Fact (hereinafter “FF”) are shown by a
preponderance of the evidence.
Wu
FF 1. Wu’s figure 6 depicts a document analysis procedure.
(¶ [0053].) In particular, upon receipt of a printing document (601), Wu
discloses that a document analyzing section (33) instructs a document
manager (35) to create a data management table for the document (602).
(Id.; fig. 3) Further, Wu’s figure 7 illustrates that a data management table is
utilized to manage each document element contained in the document.
(¶ [0054].)
Bottou
FF 2. According to Bottou, there is a need to effectively classify
image elements and, in particular, to render image elements from structured
page information, such as electronic documents. (¶ [0006].)
FF 3. Bottou discloses an image classification system that assigns
predetermined types of image elements to either the foreground or
background layer. (¶ [0026].) For example, Bottou discloses pre-assigning
text and symbols to the foreground layer in order to improve their
readability. (Id.)
FF 4. Bottou discloses that the image classification system processes
image elements using the corresponding structured page information by
reading and comparing the image element coordinates to the foreground and
background layers. (¶ [0029].)
Appeal 2009-011629
Application 10/410,999

11
Fan
FF 5. According to Fan, “it is desirable to develop a segmentation
algorithm that is suitable for many, if not all, image processing
applications.” (¶ [0002].)
FF 6. Fan discloses “a method for segmenting an image into a main
background, local backgrounds, and objects.” (¶ [0006].)
FF 7. Fan’s figure 8 depicts a background-based segmentation
process (100). (¶ [0012].) At step (546), Fan discloses analyzing,
segmenting, and classifying a first composite object. (¶ [0027].) At step
(548), Fan discloses performing edge detection on the composite object
using a standard edge detector and, at step (550), removing the edge pixels
from the composite object. (Id.) Next, at step (552)2, using each pixel on
the outer boundary of the removed edge as a seed, Fan discloses applying a
region-growing algorithm. (Id.) If the color difference between an adjacent
pixel and a region is less than a predetermined threshold, Fan discloses
marking the adjacent pixel as belonging to the region. (Id.) If the largest
region grown has dimensions larger than a predetermined threshold, Fan
discloses identifying the largest region grown as the local background. (Id.)
Pirolli
FF 8. Pirolli discloses facilitating information searching and
improving design of a web locality using a system that analyzes the
topology, content, and usage of linked collections of documents, such as
those found on the World Wide Web. (Col. 1, l. 65-col. 2, l. 2.)

2 Fan’s figure 8 mistakenly numbers this step as (550).
Appeal 2009-011629
Application 10/410,999

12
FF 9. Pirolli discloses applying techniques for information retrieval to
calculate a text similarity matrix, which represents the inter-document text
similarities among Web pages. (Col. 7, ll. 31-33.)
FF 10. In order to perform categorizations, Pirolli discloses that each
Web page at a Web locality is represented by a vector of features
constructed from the respective topology, meta-information, usage statistics,
usage paths, and text similarities. (Col. 8, ll. 4-7.)

IV. ANALYSIS
35 U.S.C. § 103(a) Rejection—Combination of Wu and Bottou
Claim 1
Independent claim 1 recites, inter alia:
respectively assigning each of the selected elements to either a
content category or a background category, wherein each
assignment of a respective one of the selected elements is based
on the respective content type value and at least one of (i) a
comparison of the at least one respective size value with a
respective threshold value and (ii) the at least one respective
distance value.

As detailed in the Findings of Fact section above, Bottou discloses an
image classification system that assigns image elements to either the
foreground or background layers. (FF 3.) In particular, Bottou discloses
assigning both text and symbols to the foreground layer in order to improve
readability. (Id.)
We find Bottou’s disclosure of assigning image elements to either the
foreground or background layers teaches or fairly suggests assigning image
elements located within a document to either a content or background
category. In particular, we find that Bottou’s disclosure of assigning
Appeal 2009-011629
Application 10/410,999

13
content, such as text and symbols, to the foreground amounts to assigning
image elements within a document to a content category based on a content
type. Thus, we find that Bottou teaches or fairly suggests “respectively
assigning each of the selected elements to either a content category or a
background category, wherein each assignment of a respective one of the
selected elements is based on the respective content type value” as recited in
independent claim 1.
Further, Bottou discloses that the image classification system
processes image elements by reading and comparing the image element
coordinates with the foreground and background layers. (FF 4.) We find
that Bottou’s disclosure of image element coordinates amounts to the
claimed “distance value” because such coordinates indicate the location of
an image element in relation to a reference location, such as the foreground
and background. Moreover, we agree with the Examiner that since
independent claim 1 explicitly recites “at least one of,” Bottou only needs to
teach or fairly suggest one of the two alternatives claimed. (Ans. 21.)
Accordingly, we find that an ordinarily skilled artisan would have readily
appreciated that Bottou’s image classification system assigns image
elements located within a document to either a content or background
category based on both the content type (i.e., text or symbols) and
corresponding image element coordinates. Thus, we find that Bottou
teaches or fairly suggests “wherein each assignment of a respective one of
the selected elements is based on the respective content type value and at
least one of (i) a comparison of the at least one respective size value with a
respective threshold value and (ii) the at least one respective distance value,”
as recited independent claim 1.
Appeal 2009-011629
Application 10/410,999

14
Rationale to Combine
We are not persuaded by Appellant’s argument that there is
insufficient rationale to combine Wu and Bottou. (App. Br. 14-17; Reply
Br. 5-7.) The U.S. Supreme Court has held that “[t]he combination of
familiar elements according to known methods is likely to be obvious when
it does no more than yield predictable results.” KSR Int'l Co. v. Teleflex,
Inc., 550 U.S. 398, 416 (2007). The Court further instructs that:
[o]ften it will be necessary for a court to look to interrelated
teachings of multiple patents; . . . and the background
knowledge possessed by a person having ordinary skill in the
art, all in order to determine whether there was an apparent
reason to combine the known elements in a the fashion claimed
by the patent at issue.

Id. at 418. Additionally, the Court instructs that:
“[[r]]ejections on obviousness grounds cannot be sustained by
mere conclusory statements; instead, there must be some
articulated reasoning with some rational underpinning to
support the legal conclusion of obviousness”… however, the
analysis need not seek out precise teachings directed to the
specific subject matter of the challenged claim, for a court can
take account of the inferences and creative steps that a person
of ordinary skill in the art would employ.

Id. (citation omitted).
Upon reviewing the record before us, we find that the Examiner’s
suggestion for the proposed modification in the prior art suffices as an
articulated reason to establish the prima facie case of obviousness. In
particular, Wu discloses a document analysis procedure that creates a data
management table for a document. (FF 1.) Further, Wu discloses that the
data management table is utilized to manage each element contained within
Appeal 2009-011629
Application 10/410,999

15
the document. (Id.) In summary, we find that an ordinarily skilled artisan in
the art of document production and publishing, at the time of the claimed
invention, would have combined Wu’s document analysis procedure (id),
with Bottou’s disclosure of assigning image elements located within a
document to either a content or background category based on both the
content type (i.e., text or symbols) and corresponding image element
coordinates. (FFs 3 and 4.) This proffered combination would result in
effectively classifying image elements and, in particular, rendering image
elements from structured page information. (FF 2.)
Further, as prescribed by the controlling case law, while it is often
necessary for an Examiner to identify a reason for combining the familiar
elements obtained from the prior art in establishing a prima facie case of
obviousness, the identification of such a reason is not a sine qua non
requirement. So long as the Examiner provides an articulated reasoning
with some kind of a rational underpinning to substantiate the obviousness
rejection, such a conclusion is proper. In this case, the Examiner provided
more than just a mere conclusory statement. The Examiner noted that it
would have been obvious to an ordinary skilled artisan at the time of the
claimed invention to modify Wu with the teachings of Bottou in order to
enable a way to classify image elements, in general, and image elements
rendered from structured page information, e.g., electronic documents, in
particular. (Ans. 5; see also FF 2.) In our view, such a statement suffices as
an articulated reason with a rational underpinning to support the cited
combination. As noted above, the case law allows the Examiner to look to
the state of the prior art, including the knowledge of the ordinarily skilled
artisan, to surmise such a reason for combining the known elements of the
Appeal 2009-011629
Application 10/410,999

16
prior art. Consequently, the Examiner’s reliance upon both Wu and Bottou
in order to arrive at an articulated reason with a rational underpinning to
support the proffered combination is proper.
For the same reasons as set forth above, we are not persuaded by
Appellant’s argument that the Examiner improperly engages in
impermissible hindsight reconstruction. (App. Br. 16.) It follows that
Appellant has not shown that the Examiner erred in concluding that the
combination of Wu and Bottou renders independent claim 1 unpatentable.
Claims 3, 8-10, 26-33, and 36
Appellant does not provide separate and distinct arguments for
patentability with respect to dependent claims 3, 8-10, 26-33, and 36.
Therefore, we select independent claim 1 as representative of the cited
claims. Consequently, Appellant has not shown error in the Examiner’s
rejection of dependent claims 3, 8-10, 26-33, and 36 for the reasons set forth
in our discussion of independent claim 1. See 37 C.F.R. § 41.37(c)(1)(vii).
35 U.S.C. § 103(a) Rejection—Combination of Wu and Fan
Claims 5 and 6
Independent claim 5 recites, inter alia:
classifying at least one selected element of a selected page into
one of a plurality of categories based on at least one of the type,
location, size, and dimension of the selected element, wherein
the classifying comprises determining whether, at least along
one dimension, the selected element has a size in relation to that
of the page that exceeds a pre-determined threshold.

As detailed in the Findings of Fact section above, Fan discloses a
background-based segmentation process that analyzes, segments, and
Appeal 2009-011629
Application 10/410,999

17
classifies a composite object. (FF 7.) In particular, Fan discloses using a
standard edge detector to perform edge detection on the composite object
and, thereafter, applying a region-growing algorithm. (Id.) If the largest
region grown with respect to the object has dimensions that are larger than a
predetermined threshold, Fan discloses identifying the composite object as
the local background. (Id.)
We find that Fan’s disclosure teaches classifying a selected composite
object or element located within a document into a background category if
the largest region grown with respect to the element has dimensions that
exceed a predetermined threshold. In particular, we find that an ordinarily
skilled artisan would have appreciated that since the selected element is
located within the document, Fan’s disclosure of the dimensions of the
largest region grown with respect to the element necessarily encompasses
ascertaining the size of such element in relation to the document. Thus, we
find that Fan teaches or fairly suggests the disputed limitation.
Rationale to Combine
We are not persuaded by Appellant’s argument that there is
insufficient rationale to combine Wu and Fan. (App. Br. 20-23, 25; Reply
Br. 9, 11.) We find more than adequate suggestion for the proposed
modification in the prior art. In summary, we find that an ordinarily skilled
artisan in the art of document production and publishing, at the time of the
claimed invention, would have combined Wu’s document analysis procedure
(FF 1), with Fan’s disclosure of classifying a selected element located within
a document into a background category if the largest region grown with
respect to the element has dimensions that exceed a predetermined threshold.
(FF 7.) This proffered combination would result in the development of a
Appeal 2009-011629
Application 10/410,999

18
segmentation algorithm that is suitable for many, if not all, image processing
applications. (FF 5.) Moreover, this proffered combination would
predictably result in a method for segmenting an image or element into
various categories. (FF 6.)
Further, the Examiner provided more than just a mere conclusory
statement in order to establish a prima facie case of obviousness. The
Examiner noted that at the time of the claimed invention, an ordinarily
skilled artisan would have modified Wu with the teachings of Fan in order to
enable a segmentation algorithm suitable for many image processing
applications. (Ans. 10.) The Examiner also notes that the proffered
combination would enable a method for segmenting an image into
categories, such as main background, local background, and objects. (Id; see
also FF 6) In our view, such statements suffice as articulated reasons with
rational underpinnings to support the cited combination. Consequently, the
Examiner’s reliance upon Wu and Fan in order to arrive at articulated
reasons with rational underpinnings to support the proffered combination is
proper.
For the same reasons as set forth above, we are not persuaded by
Appellant’s argument that the Examiner improperly engages in
impermissible hindsight reconstruction. (App. Br. 22.) It follows that
Appellant has not shown that the Examiner erred in concluding that the
combination of Wu and Fan renders independent claim 5 and independent
claim 6, which recites commensurate limitations, unpatentable.
Appeal 2009-011629
Application 10/410,999

19
35 U.S.C. § 103(a) Rejection—Combination of Wu, Bottou, and Pirolli
Claim 13
Independent claim 13 recites, inter alia, “classifying a selected one of
the elements on a selected one of the pages into one of a content category
and a background category based on frequency of occurrence of similar
elements on one or more pages of the document other than the selected
page.”
As detailed in the Findings of Fact section above, Pirolli discloses
techniques for information retrieval that include calculating text similarities
among Web pages. (FF 9.) Further, Pirolli discloses representing each Web
page at a Web locality by a vector of features constructed from topology,
meta-information, usage statistics, usage paths, and text similarities. (FF
10.)
We find that an ordinarily skilled artisan would have recognized that
Pirolli’s disclosure of information retrieval techniques that calculate the
similarities and usage statistics of text among several Web pages
encompasses determining the frequency of occurrence of similar elements
on the pages of a respective document other than a selected page. In
summary, we find that an ordinarily skilled artisan would have readily
appreciated classifying a selected image element located on a page of a
document into either Bottou’s content or background category based on
Pirolli’s disclosure of determining the frequency of occurrence of similar
elements on the pages of the respective document other than the selected
page. Thus, we find that the combination of Bottou and Pirolli teaches or
fairly suggests the disputed limitation.
Appeal 2009-011629
Application 10/410,999

20
Rationale to Combine
We are not persuaded by Appellant’s argument that there is
insufficient rationale to combine Wu, Bottou, and Pirolli. (App. Br. 28-30;
Reply Br. 13.) As set forth above, we find that there is sufficient rationale to
combine Wu and Bottou. Moreover, we find that an ordinarily skilled
artisan in the art of document production and publishing, at the time of the
claimed invention, would have combined Wu’s document analysis procedure
(FF 1), with Pirolli’s disclosure of determining the frequency of occurrence
of similar elements on the pages of a respective document other than a
selected page. (FFs 9 and 10.) This proffered combination would
predictably result in a system that facilitates information searching by
analyzing the topology, content, and usage of a linked collection of
documents. (FF 8.)
Further, the Examiner provided more than just a mere conclusory
statement in order to establish a prima facie case of obviousness. The
Examiner noted that at the time of the claimed invention, an ordinarily
skilled artisan would have modified Wu with the teachings of Pirolli in order
to enable a system for analyzing the topology, content and usage of linked
collections of documents on the web to facilitate information searching.
(Ans. 13-14; see also FF 8.) In our view, such a statement suffices as an
articulated reason with a rational underpinning to support the cited
combination. Consequently, the Examiner’s reliance upon Wu, Bottou, and
Pirolli in order to arrive at articulated reasons with rational underpinnings to
support the proffered combination is proper.
For the same reasons as set forth above, we are not persuaded by
Appellant’s argument that the Examiner improperly engages in
Appeal 2009-011629
Application 10/410,999

21
impermissible hindsight reconstruction. (App. Br. 29.) It follows that
Appellant has not shown that the Examiner erred in concluding that the
combination of Wu, Bottou, and Pirolli renders independent claim 13
unpatentable.
Claims 14, 16, and 23-25
Appellant offers the same arguments set forth in response to the
obviousness rejections of independent claims 1 and 13 to rebut the
obviousness rejections of independent claims 16, 23, and 25, and dependent
claims 14 and 24. (App. Br. 30-34; Reply Br. 14.) We have already
addressed these arguments in our discussion of independent claims 1 and 13,
and we found them unpersuasive. It follows that Appellant has not shown
that the Examiner erred in concluding that the combination of Wu, Bottou,
and Pirolli renders claims 14, 16, and 23-25 unpatentable.
35 U.S.C. § 103(a) Rejection—Combination of Wu, Bottou, and Fan
Claim 35
Appellant offers the same arguments set forth in response to the
obviousness rejection of independent claim 1 to rebut the obviousness
rejection of dependent claim 35. (App. Br. 34-35; Reply Br. 15.) We have
already addressed these arguments in our discussion of independent claim 1,
and we found them unpersuasive. It follows that Appellant has not shown
that the Examiner erred in concluding that the combination of Wu, Bottou,
and Fan renders dependent claim 35 unpatentable.

Appeal 2009-011629
Application 10/410,999

22
V. CONCLUSION OF LAW
Appellant has not shown that the Examiner erred in rejecting claims 1,
3, 5, 6, 8-10, 13, 14, 16, 23-33, 35, and 36 as being unpatentable under 35
U.S.C. § 103(a).

VI. DECISION
We affirm the Examiner’s decision to reject claims 1, 3, 5, 6, 8-10, 13,
14, 16, 23-33, 35, and 36 as being unpatentable under 35 U.S.C. § 103(a).
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

msc