Ex Parte Zhang et al

Patent Trial and Appeal Board

Jul 30, 2014

12133766 (P.T.A.B. Jul. 30, 2014)

UNITED STATES PATENT AND TRADEMARKOFFICE
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.
12/133,766 06/05/2008 Li Zhang CN920050017US2 1289
75234 7590 07/31/2014
RYAN, MASON & LEWIS, LLP
48 South Service Road
Suite 100
Melville, NY 11747
EXAMINER
CHOI, YUK TING
ART UNIT PAPER NUMBER
2164
MAIL DATE DELIVERY MODE
07/31/2014 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 PATENT TRIAL AND APPEAL BOARD
____________

Ex parte LI ZHANG, LI PING YANG, and SHIXIA LIU
____________

Appeal 2012-004994
Application 12/133,7661
Technology Center 2100
____________

Before MARC S. HOFF, BRETT C. MARTIN, and
JEREMY J. CURCURI, Administrative Patent Judges.

HOFF, Administrative Patent Judge.

DECISION ON APPEAL
Appellants appeal under 35 U.S.C. § 134(a) from a final rejection of
claims 1–7 and 10–16 . Claims 8 and 9 have been canceled. We have
jurisdiction under 35 U.S.C. § 6(b).
We affirm-in-part.

STATEMENT OF THE CASE
Appellants’ invention is a method, apparatus, and system for linking
documents into a World Wide Web (WWW)-like virtual network. The
method forms a plurality of document clusters each consisting of one or
more documents. A cluster page is built for each cluster to represent the

1 The real party in interest is International Business Machines Corporation.
Appeal 2012-004994
Application 12/133,766

2
documents in the cluster. Links are built based on an analysis of the
contents of the clusters and the documents (Spec. 2-3). When a similarity of
a topic of a first document to a second document is greater than a threshold
(similarity being a cosine function of an angle between a topic vector of the
first document and the document vector of the second document), a link is
built from the first document to the second document (Spec. 11).
Independent claim 1, reproduced below, is representative of the
subject matter on appeal.
1. A method for linking documents, comprising a computer
performing the steps of:

obtaining a set of documents, wherein documents in the set of
documents are not interlinked with other documents via one or more
hyperlinks;

forming a plurality of clusters from the set of documents, each cluster
comprising one or more documents;

building a cluster page for each cluster to represent the documents in
the cluster; and

building links based on analysis of the contents of the clusters and the
documents;

wherein, when a similarity of a topic of a first document of the set of
documents to a second document of the set of documents is greater than a
threshold, a link is built from the first document to the second document, the
similarity being a cosine function of an angle between a topic vector of the
first document and the document vector of the second document.

REFERENCES
Schuetze US 6,598,054 B2 July 22, 2003
Ando US 6,775,677 B1 Aug. 10, 2004
Miwa US 7,213,205 B1 May 1, 2007
Appeal 2012-004994
Application 12/133,766

3
Harik US 7,231,393 B1 June 21, 2007
Zhang US 7,499,934 B2 Mar. 3, 2009
(filed May 23, 2006)

REJECTIONS
Claims 1–7 and 10–16 stand rejected under 35 U.S.C. § 112, second
paragraph, as being indefinite.
Claims 1, 2, and 10–16 stand rejected under 35 U.S.C. § 103(a) as
being unpatentable over Schuetze in view of Ando.
Claims 3, 4, 6, and 7 stand rejected under 35 U.S.C. § 103(a) as being
unpatentable over Schuetze in view of Ando, Harik, and Miwa.
Claim 5 stands rejected under 35 U.S.C. § 103(a) as being
unpatentable over Schuetze in view of Ando and Harik.

ISSUES
Appellants disagree with the Examiner that independent claims 1, 12,
and 16 are indefinite because “the document vector” lacks proper antecedent
basis (App. Br. 8).
Appellants contend that neither Schuetze nor Ando disclose
evaluating the similarity between a document vector and a topic vector, nor
using that similarity to determine whether to build a link between a first
document and a second document (App. Br. 9-10).
Appellants further assert that Schuetze does not disclose providing a
plurality of clusters by splitting a plurality of unlinked documents into
different sets (App. Br. 11).
Appellants contend that Harik and Miwa fail to disclose a
determination whether to generate a link from a first cluster to a second
Appeal 2012-004994
Application 12/133,766

4
cluster, or instead from a second cluster to a first cluster, based on whether
the proportion of commonly owned documents in the first cluster is greater
than that in the second cluster (App. Br. 13).
Appellants argue that Schuetze does not disclose the computation of
similarity between a document vector of a document and a centroid vector of
a cluster (App. Br. 13).
Appellants assert that neither Harik nor Ando discloses the claim
limitation of building a link “from a cluster page to each document therein”
(App. Br. 14-15).
Appellants’ arguments present us with the following issues:
1. Is the phrase “the document vector” in the independent claims
indefinite within the meaning of § 112?
2. Does the combination of Schuetze and Ando fairly suggest
evaluating the similarity between a document vector and a topic vendor, and
using that similarity to determine whether to build a link from a first
document to a second document?
3. Does Schuetze disclose providing a plurality of clusters by splitting
a plurality of unlinked documents into different sets?
4. Do Harik and Miwa disclose or suggest a determination whether to
generate a link from a first cluster to a second cluster, or from a second
cluster to a first cluster, based on whether the proportion of commonly
owned documents in the first cluster is greater than that in the second
cluster?
5. Does Schuetze disclose the computation of similarity between a
document vector of a document and a centroid vector of a cluster?
Appeal 2012-004994
Application 12/133,766

5
6. Do Harik and Ando disclose or fairly suggest building a link from a
cluster page to each document therein?

PRINCIPLES OF LAW
The test for definiteness under 35 U.S.C. § 112, second paragraph, is
whether “those skilled in the art would understand what is claimed when the
claim is read in light of the specification.” Orthokinetics, Inc. v. Safety
Travel Chairs, Inc., 806 F.2d 1565, 1576 (Fed. Cir. 1986) (citations
omitted). Claims must “particularly point[] out and distinctly claim[] the
subject matter which the applicant regards as his invention.” 35 U.S.C.
§ 112, ¶ 2. However, “[o]nly claims not amenable to construction or
insolubly ambiguous are indefinite. . . . A claim term is not indefinite just
because it poses a difficult issue of claim construction.” Star Scientific, Inc.
v. R.J. Reynolds Tobacco Co., 537 F.3d 1357, 1371 (Fed. Cir. 2008)
(internal citations and quotation marks omitted).
The 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 v. Teleflex Inc., 550 U.S. 398, 416 (2007).

ANALYSIS
§ 112, SECOND PARAGRAPH REJECTION
We are persuaded by Appellants’ arguments and we find error in the
Examiner’s § 112 rejection of claims 1–7 and 10–16. We do not agree with
the Examiner that “the document vector of the second document” is
“insolubly ambiguous.” See Star Scientific, 537 F.3d at 1371. Contrary to
the Examiner’s arguments, we find that the recited “the document vector”
Appeal 2012-004994
Application 12/133,766

6
cannot reasonably be read as referring to the first document, or to a “new
document vector of the second document” when no such “new” vector was
disclosed or claimed elsewhere in the application. We do not sustain the
Examiner’s rejection.

§ 103 REJECTION OF CLAIMS 1, 2, AND 11–16
We are not persuaded that Ando discloses a similarity between two
document vectors, not between a document vector and a topic vector (App.
Br. 9). Ando explicitly discloses topic-document relevance, even giving
examples of topic-document relevance computation (col. 11, ll. 14-29).
We further disagree with Appellants’ contention that Ando does not
disclose or suggest using topic-document relevance to determine whether a
link is built between a first document and a second document (App. Br. 10).
Ando discloses building a link (an “edge”) when the similarity
between two documents is above a threshold (col. 9, ll. 47-56). Further,
Ando teaches creating a document map as a result of topic-document
relevance results, graphically indicating the similarity between documents
and topics (col. 11, ll. 32-45). We find that applying Ando’s building of
links as a result of document-document comparison to the concept of
document-to-topic comparison, in view of Ando’s disclosure of constructing
a document map that shows the similarity between documents and topics,
would have been a predictable result of combining familiar elements
according to known methods. See KSR, 550 U.S. at 416.
We find that the combination of Schuetze and Ando discloses all the
limitations of the claimed invention. We sustain the Examiner’s § 103
rejection of claims 1, 2, and 11–16.
Appeal 2012-004994
Application 12/133,766

7
CLAIM 10
We are not persuaded of Examiner error by Appellants’ assertion that
Schuetze does not disclose providing a plurality of clusters by splitting a
plurality of unlinked documents into different sets (App. Br. 11). Schuetze
discloses that “the objects of a cluster are characterized by similarities
among the objects on features of text, usage, collection topology (inlinks and
outlinks), and URL” (col. 29, ll. 3-6). We do not agree with Appellants’
contention that Schuetze describes a technique that relies on links among
pages (App. Br. 12); documents in a cluster need not be linked to be
clustered.
We find that the Examiner did not err in rejecting claim 10 over
Schuetze in view of Ando. We sustain the § 103 rejection.

CLAIMS 3 AND 7
Appellants do not separately argue the patentability of claims 3 and 7,
relying on the arguments made for claim 1. We, therefore, sustain the § 103
rejection of claims 3 and 7 for the reasons given supra, with respect to claim
1.
CLAIMS 4 AND 6
We agree with Appellants that Harik and Miwa fail to disclose a
determination whether to generate a link from a first cluster to a second
cluster, or making such a determination based on whether the proportion of
commonly owned documents in the first cluster is greater than that in the
second cluster (App. Br. 13). We find that Miwa fails to disclose
determining whether the proportion of commonly owned documents in a
Appeal 2012-004994
Application 12/133,766

8
first cluster is greater than the proportion of commonly owned documents in
a second cluster.
We further agree with Appellants that Schuetze does not disclose the
computation of similarity between a document vector of a document and a
centroid vector of a cluster (App. Br. 13).
We do not sustain the Examiner’s § 103 rejection of claims 4 and 6
over Schuetze in view of Ando, Harik, and Miwa.

Claim 5
We agree with Appellants that neither Harik nor Ando, relied upon by
the Examiner, discloses the claim limitation of building a link “from a
cluster page to each document therein” (App. Br. 14-15). Harik builds links
from one cluster to another cluster. Ando builds links from one document to
another document.
Thus, we do not sustain the Examiner’s § 103 rejection of claim 5.

CONCLUSIONS
1. The phrase “the document vector” is not indefinite within the
meaning of § 112.
2. The combination of Schuetze and Ando fairly suggests evaluating
the similarity between a document vector and a topic vendor, and using that
similarity to determine whether to build a link from a first document to a
second document.
3. Schuetze discloses providing a plurality of clusters by splitting a
plurality of unlinked documents into different sets.
Appeal 2012-004994
Application 12/133,766

9
4. Harik and Miwa do not disclose or suggest a determination whether
to generate a link from a first cluster to a second cluster, or from a second
cluster to a first cluster, based on whether the proportion of commonly
owned documents in the first cluster is greater than that in the second
cluster.
5. Schuetze does not disclose the computation of similarity between a
document vector of a document and a centroid vector of a cluster.
6. Harik and Ando do not disclose or fairly suggest building a link
from a cluster page to each document therein.

DECISION
The Examiner’s decision rejecting claims 1–3, 7, and 10-16 is
affirmed.
The Examiner’s decision rejecting claims 4–6 is reversed.
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-IN-PART

rwk