U.S. patent application number 10/142393 was filed with the patent office on 2004-01-01 for automatic display of web content to smaller display devices: improved summarization and navigation.
Invention is credited to Emmett, David, Rahman, Ahmad.
Application Number | 20040003028 10/142393 |
Document ID | / |
Family ID | 29778486 |
Filed Date | 2004-01-01 |
United States Patent
Application |
20040003028 |
Kind Code |
A1 |
Emmett, David ; et
al. |
January 1, 2004 |
Automatic display of web content to smaller display devices:
improved summarization and navigation
Abstract
A system and method are disclosed for modifying a document
format. In one embodiment, a structure of a first document is
extracted to form a first data structure, including multiple
subdocuments, each subdocument having a label assigned thereto. A
table of contents listing the labels of the subdocuments is then
generated. The various labels are then ordered according to the
amount of text of the associated subdocument, a comparison of the
text of a previous link and the text of the associated subdocument,
a comparison of the text of the previous document and the text of
the associated subdocument, a comparison of the text of the
previous link and the text of the associated subdocument, or a
combination of these.
Inventors: |
Emmett, David; (Palo Alto,
CA) ; Rahman, Ahmad; (San Jose, CA) |
Correspondence
Address: |
KOLISCH HARTWELL, P.C.
520 S.W. YAMHILL STREET
SUITE 200
PORTLAND
OR
97204
US
|
Family ID: |
29778486 |
Appl. No.: |
10/142393 |
Filed: |
May 8, 2002 |
Current U.S.
Class: |
709/203 ;
707/E17.121; 709/246; 715/205; 715/273 |
Current CPC
Class: |
G06F 16/9577
20190101 |
Class at
Publication: |
709/203 ;
709/246; 715/513 |
International
Class: |
G06F 015/16 |
Claims
What is claimed is:
1. A method of ranking entries in a table of contents, the method
comprising: transmitting a first document to a client device, the
first subdocument including a text and a link; receiving from the
client device a request for a second document associated with the
link; dividing the second document into subdocuments, each of the
subdocuments including text; assigning a label to each of a
plurality of the subdocuments; comparing the text of the first
document with the text of each of the subdocuments to generate a
document-document value for each of the plurality of subdocuments
according to the degree of association between the first document
and the corresponding one of the subdocuments; ranking in ranked
order the plurality of subdocuments based, at least in part, on the
document-document values.
2. The method of claim 1, further comprising transmitting a table
of contents page listing the labels associated with the plurality
of subdocuments in the ranked order.
3. The method of claim 1, wherein the link comprises text, the
method further comprising: comparing the text of the link with the
text of each of the subdocuments to generate a link-text value for
each of the plurality of subdocuments according to the degree of
association between the first document and the corresponding one of
the subdocuments; ranking in ranked order the plurality of
subdocuments based, at least in part, on the link-text values.
4. The method of claim 1, wherein the link comprises text, the
method further comprising: comparing the text of the link with the
label assigned to each of the plurality of subdocuments to generate
a link-label value for each of the plurality of subdocuments
according to the degree of association between the first document
and the corresponding one of the subdocuments; ranking in ranked
order the plurality of subdocuments based, at least in part, on the
link-label values.
5. The method of claim 1, further comprising; displaying the first
document at the client device; detecting user selection of the
link; transmitting the request for a second document associated
with the link in response to the detecting user selection of the
link.
6. The method of claim 1, further comprising: generating a size
value indicative of an amount of text in each of the plurality of
subdocuments; ranking in ranked order the plurality of subdocuments
based, at least in part, by the size value.
7. The method of claim 3, further comprising: generating a size
value indicative of an amount of text in each of the plurality of
subdocuments; ranking in ranked order the plurality of subdocuments
based, at least in part, by the size value.
8. The method of claim 4, further comprising: generating a size
value indicative of an amount of text in each of the plurality of
subdocuments; ranking in ranked order the plurality of subdocuments
based, at least in part, by the size value.
9. The method of claim 3, wherein the link comprises text, the
method further comprising: comparing the text of the link with the
label assigned to each of the plurality of subdocuments to generate
a link-label value for each of the plurality of subdocuments
according to the degree of association between the first document
and the corresponding one of the subdocuments; ranking in ranked
order the plurality of subdocuments based, at least in part, on the
link-label values.
10. The method of claim 1, further comprising modifying the ranked
order according to a predefined order defined by a data
structure.
11. A method of ranking entries in a table of contents, the method
comprising: transmitting a first document to a client device, the
first subdocument including a text and a link; receiving from the
client device a request for a second document associated with the
link; dividing the second document into subdocuments, each of the
subdocuments including text; assigning a label to each of a
plurality of the subdocuments; comparing the text of the link with
the text of each of the subdocuments to generate a link-text value
for each of the plurality of subdocuments according to the degree
of association between the first document and the corresponding one
of the subdocuments; ranking in ranked order the plurality of
subdocuments based, at least in part, on the link-text values.
12. The method of claim 11, wherein the link comprises text, the
method further comprising: comparing the text of the link with the
label assigned to each of the plurality of subdocuments to generate
a link-label value for each of the plurality of subdocuments
according to the degree of association between the first document
and the corresponding one of the subdocuments; ranking in ranked
order the plurality of subdocuments based, at least in part, on the
link-label values.
13. The method of claim 11, further comprising; displaying the
first document at the client device; detecting user selection of
the link; transmitting the request for a second document associated
with the link in response to the detecting user selection of the
link.
14. The method of claim 11, further comprising: generating a size
value indicative of an amount of text in each of the plurality of
subdocuments; ranking in ranked order the plurality of subdocuments
based, at least in part, by the size value.
15. A method of ranking entries in a table of contents, the method
comprising: transmitting a first document to a client device, the
first subdocument including a text and a link; receiving from the
client device a request for a second document associated with the
link; dividing the second document into subdocuments, each of the
subdocuments including text; assigning a label to each of a
plurality of the subdocuments; comparing the text of the link with
the label assigned to each of the plurality of subdocuments to
generate a link-label value for each of the plurality of
subdocuments according to the degree of association between the
first document and the corresponding one of the subdocuments;
ranking in ranked order the plurality of subdocuments based, at
least in part, on the link-label values.
16. The method of claim 15, further comprising; displaying the
first document at the client device; detecting user selection of
the link; transmitting the request for a second document associated
with the link in response to the detecting user selection of the
link.
17. The method of claim 16, further comprising: generating a size
value indicative of an amount of text in each of the plurality of
subdocuments; ranking in ranked order the plurality of subdocuments
based, at least in part, by the size value.
18. A method of ranking entries in a table of contents, the method
comprising: transmitting a first document to a client device, the
first subdocument including a text and a link; receiving from the
client device a request for a second document associated with the
link; dividing the second document into subdocuments, each of the
subdocuments including text; comparing the text of the first
document with the text of each of the subdocuments to generate a
document-document value for each of the plurality of subdocuments
according to the degree of association between the first document
and the corresponding one of the subdocuments; comparing the text
of the link with the label assigned to each of the plurality of
subdocuments to generate a link-label value for each of the
plurality of subdocuments according to the degree of association
between the first document and the corresponding one of the
subdocuments; comparing the text of the link with the text of each
of the subdocuments to generate a link-text value for each of the
plurality of subdocuments according to the degree of association
between the first document and the corresponding one of the
subdocuments; ranking in ranked order the plurality of subdocuments
based, at least in part, on the document-document, link-label, and
link-text values.
19. The method of claim 18, further comprising: generating a size
value indicative of an amount of text in each of the plurality of
subdocuments; ranking in ranked order the plurality of subdocuments
based, at least in part, by the size value.
20. A method of ranking entries in a table of contents, the method
comprising: transmitting a first document to a client device, the
first subdocument including a text and a link; receiving from the
client device a request for a second document associated with the
link; dividing the second document into subdocuments, each of the
subdocuments including text; assigning a label to each of a
plurality of the subdocuments; generating a size value indicative
of an amount of text in each of the plurality of subdocuments;
ranking in ranked order the plurality of subdocuments based, at
least in part, by the size value.
Description
CROSS REFERENCE TO RELATED APPLICATIONS
[0001] This application claims the benefit and priority of U.S.
Provisional Patent Application entitled "Automatic Display of Web
Content to Smaller Display Devices: Improved Summarization and
Navigation" filed May 10, 2001 and of U.S. patent application Ser.
No. 10/076,786 entitled "System and Method for Modifying a Document
Format)" filed Feb. 14, 2002, the disclosures of which are hereby
incorporated by reference in their respective entireties.
TECHNICAL FIELD
[0002] The present invention relates to a system and method for
modifying a document format.
BACKGROUND
[0003] Handheld devices, including Personal Digital Assistants
(PDAs) and cellular telephones, offer connectivity to the Internet
and permit access to documents available over the Internet.
Wireless Application Protocol (WAP) is a standard for providing
cellular phones, PDAs, pagers and other handheld devices with
secure access to web pages. WAP features the Wireless Markup
Language (WML), which generally serves as a medium for translating
web-based HTML content into a format that accommodates small form
factor displays and key sets found on conventional handheld
devices. WML also allows handheld device manufacturers to include
microbrowsers in their products that accept WML input from a
WAP-based system across vast regions of the world.
[0004] The proliferation of wireless PDAs has also created a
popular means for handheld Internet access. However, presenting
IP-based content, and other content developed for display on large
form factor devices (e.g., PC monitors), on small form factor
screens of handheld devices has, in the past, been problematic. Two
primary methods of presenting such content to handheld devices have
been employed.
[0005] The first such method can be termed "fixed mapping." Fixed
mapping typically involves rewriting an existing document, such as
an HTML-based web page, to conform to a specific standard, such as
WAP, J-PHONE, or i-Mode, or to a small display device. A web server
must then maintain the rewritten web site as a separate site with
its own URL in addition to the original document. As new content is
added to the original document, a web site operator must manually
trim, edit, and condense the new content by rewriting the new
content into a format that will accommodate the interface
parameters of handheld devices. This method is limited in that
considerable time and expense are typically required to maintain
the two web sites in parallel. Further, the manual editing of the
rewritten web site can be time-consuming, burdensome, and
expensive.
[0006] The second method may be termed "transcoding." Transcoding
typically involves the use of software that takes the entire
content of a web site as input, converts the entire content into a
format of a specific handheld wireless standard for transmission to
handheld devices. The entire content, as formatted according to a
handheld wireless standard, is then transmitted to the handheld
device. This conversion may be performed "on-the-fly" (i.e.,
automatically in real time) or may be performed manually.
[0007] Transcoding has the advantage of reducing the investment to
reach wireless markets since it leverages existing web sites. From
a user standpoint, transcoding is desirable in that it preserves
all the text-based information from the originating site. For large
volumes of text, however, using this approach may overwhelm the
handheld device user with large volumes of text to be viewed on a
small form factor display. Further, the unorganized transcoded
content makes changes or modifications to the wirelessly enabled
web site more difficult for the web site operator.
[0008] In addition, many wireless handheld devices have limited
bandwidth. Thus, downloading an entire web page designed for
viewing on a large form factor device at data rates common to
handheld wireless devices may require large download times. These
large download times may be burdensome to the user who must wait
while the entire web page downloads, even though the user may only
desire to view a portion of the web page. Further, these large
download times may be expensive for users who pay for wireless
service based on the amount of time or the number of packets
downloaded. For example, service plans are time-based or
packet-based. These service plans charge on either the time
connected or number of packets received, respectively. Thus, large
downloads under these service plans will be more expensive than
smaller downloads.
[0009] Additional background details are disclosed in U.S. Pat. No.
6,336,124, the disclosure of which is hereby incorporated by
reference.
SUMMARY
[0010] Accordingly, a need exists to provide a system and method
for presenting content developed for display on large form factor
devices (e.g., PC monitors) on small form factor screens of
handheld devices. In particular, a need exists for a system and
method for permitting a handheld device user to easily navigate
material available over a network, such as an Internet web
site.
[0011] Pursuant to one embodiment of the present invention, a
method of ranking entries in a table of contents for display at a
client device includes transmitting a first document from an
application server over a network, such as the Internet, to the
client device. The first document includes text and at least one
link. The application server then receives a request for a second
document associated with the link from the client device. Next, the
application server divides the second document into subdocuments
and assigns a label to each of a plurality of the subdocuments. The
application server also performs a comparison of the text of the
first document with the text of each of the plurality of
subdocuments to generate a document-document value for each of the
plurality of subdocuments according to the degree of association
between the first document and the corresponding one of the
subdocuments. After performing this comparison, the application
server ranks the plurality of subdocuments based, at least in part,
on the document-document values.
[0012] In another embodiment, the application server performs a
comparison of the text of the link with the text of each of the
subdocuments to generate a link-text value for each of the
plurality of subdocuments according to the degree of association
between the first document and the corresponding one of the
subdocuments. After performing this comparison, the application
server ranks the plurality of subdocuments based, at least in part,
on the link-text values.
[0013] In yet another embodiment, the application server performs a
comparison of the text of the link with the label assigned to each
of the plurality of subdocuments to generate a link-label value for
each of the plurality of subdocuments according to the degree of
association between the first document and the corresponding one of
the subdocuments. After performing this comparison, the application
server ranks the plurality of subdocuments based, at least in part,
on the link-label values.
[0014] In still another embodiment, the application server
generates a size value indicative of an amount of text in each of
the plurality of subdocuments for each of the plurality of
subdocuments. After generating a size value for each of the
plurality of subdocuments, the application server ranks the
plurality of subdocuments based, at least in part, by the size
value.
[0015] In this manner, subdocuments likely to be relevant to the
first document, the selected link, or both, are listed at or near
the top of a table of contents to facilitate user selection of the
same. Hence, users may easily follow a text that spans multiple
documents by having table of contents of a requested page list the
subdocuments containing continuing portions of the text listed at
or near the top of the table of contents.
[0016] Additional details regarding the present system and method
may be understood by reference to the following detailed
description when read in conjunction with the accompanying
drawings.
BRIEF DESCRIPTION OF THE DRAWINGS
[0017] FIG. 1 is a block diagram of a document delivery system in
accordance with one embodiment of the present invention.
[0018] FIG. 2 is a block diagram of the formatter of FIG. 1 in
accordance with one embodiment of the present invention.
[0019] FIG. 3 is a block diagram of the mapper of FIG. 2 in
accordance with one embodiment of the present invention.
[0020] FIG. 4 illustrates a tree data structure in accordance with
one embodiment of the present invention.
[0021] FIG. 5 is a block diagram of the control module of FIG. 2 in
accordance with one embodiment of the present invention.
[0022] FIG. 6 is a flowchart illustrating a method in accordance
with one embodiment of the present invention.
[0023] FIG. 7 is a flowchart illustrating a method in accordance
with another embodiment of the present invention.
[0024] FIG. 8 illustrates a progression of material displayed at
the display of the client device of FIG. 1 in accordance with an
embodiment of the present invention.
[0025] Common reference numerals are used throughout the drawings
and detailed description to indicate like elements.
DETAILED DESCRIPTION
[0026] FIG. 1 illustrates a document delivery system 100 in
accordance with one embodiment of the present invention. The
document delivery system 100 permits a client 102 to access content
of documents (not shown) stored at server 104, server 106, or other
servers 108 over a network 110, such as the Internet, and over a
network 111, such as an intranet.
[0027] In one embodiment, the client 102 comprises a handheld
device, such a PDA (Personal Digital Assistant), a mobile
telephone, or the like, having a small form factor display 112. The
client 102 also includes a web browser 114. The web browser 114 may
comprise a microbrowser designed for small display screens on
web-enabled cellular telephones, PDAs and other handheld devices,
including wireless handheld devices.
[0028] The client 102 may exchange data with the network 110 in a
wireless fashion via a wireless station 120 and a gateway 122 in
accordance with WAP (Wireless Application Protocol), i-Mode, or
other suitable protocol or service. Optionally, the client 102 may
exchange data with the network 110 via a wired connection (not
shown).
[0029] The client 102 may also exchange data with the network 111
in a wireless fashion via a wireless station 121 and a gateway 123
in accordance with WAP (Wireless Application Protocol), i-Mode, or
other suitable protocol or service for delivery of web content to
small display devices. Optionally, the client 102 may exchange data
with the network 111 via a wired connection (not shown).
[0030] In one embodiment, the gateways 122, 123 are network devices
that connect a wireless network with a wired network, such as the
networks 110, 111. Access between the client 102 and application
server 124 may also pass through one or more other firewalls (not
shown), other gateway devices (not shown), or the like.
[0031] Pursuant to one embodiment, the client 102 transmits
requests for documents stored on one or more of the servers 104,
106, 108 to the application server 124. The request for content may
comprise an HTTP request or other suitable type of request.
Moreover, the application server 124 may alternatively receive the
request for a document from the client 102 from any network (e.g.,
110, 111). The application server 124, among other functionality,
functions as a proxy server and receives requests for documents
from client devices, such as the client 102, over the networks 110,
111 and provides associated content in response to such requests by
transmitting the associated content over at least one of the
networks 110, 111.
[0032] In response to a request for a document from the client 102,
the application server 124 requests the document identified by the
request from one or more of the servers 104, 106, 108. Upon receipt
of the document identified by the request, the application server
124 modifies the format of the document identified by the request
for content using a formatter 126.
[0033] In one embodiment, the document identified by the request is
an HTML or XML web page, although other document types, such as PDF
(Portable Document Format), may also be requested. The application
server 124 then transmits at least a portion of the reformatted
content of the document identified by the request to the client 102
in a format compatible with the browser 114 for display at the
display 112 of the client 102.
[0034] The formatter 126 includes a database (see FIG. 5) that may
be configured from a client admin computer 140 via a database
modifier 128. The database modifier 128 may comprise a JavaScript
module that permits a user at the client admin computer to visually
modify a data structure of a document into a desired format. The
modification may be performed by, for example, adding labels,
re-ordering, moving, deleting, or otherwise changing portions of
the data structure and stores the changed, or modified version of
the data structure in the database.
[0035] In particular, the client admin computer 140 includes a web
browser 142, such as Internet Explorer.TM. by Microsoft Corporation
or other suitable web browser for permitting a user at the client
admin computer 140 to view pages at the database modifier 128
hosted at the application server 124. The pages at the database
modifier 128 of the application server 124 permit user
configuration of the FIG. 5 database, as discussed in more detail
below.
[0036] In general, the formatter 126 receives the document
identified by the request from one of the servers 104, 106, 108,
divides the document into multiple blocks, and assigns labels to
individual blocks. The formatter 126 then generates a list
containing the content of the various blocks. If a data structure
associated with the document is stored in the database, the
formatter 126 then uses the data structure to generate output files
from the generated list of content. The output file may contain a
Table of Contents (TOC) page and subdocuments. The TOC page lists
labels associated with the subdocuments and may contain links to
the subdocuments. The formatter 126 then transmits the TOC page, a
headline, an image, or other content specified by a database at the
application server 124 to the client 102 over at least one of the
networks 110, 111. Details of the operation of the formatter 126
are discussed in more detail below.
[0037] FIG. 2 illustrates details of the formatter 126 of FIG. 1
according to one embodiment of the invention. As shown, the
formatter 126 includes a mapper 202, and a control module 206,
which may comprise software written in C++ or other suitable
programming language. The mapper 202 receives the requested
document and reformats the document as a list of document content
204. The control module 206 then generates an output file using the
list document content 204. Additional details regarding the mapper
202, the list of document content 204, and the control module 206
are discussed below.
[0038] FIG. 3 illustrates details of the mapper 202 of FIG. 2
according to one embodiment of the invention. The mapper 202
includes a number of software modules stored in a computer readable
medium. In particular, the mapper 202 includes a network interface
302, a parser 304, a label engine 306, a data structure converter
308, and a ranking engine 310. The network interface 302 receives
the document requested from the network. As mentioned above, the
document requested may comprise a web page, such as an HTML
document, and XML document, or the like.
[0039] The parser 304 parses and decomposes the document into a
tree data structure. FIG. 4 illustrates an example tree data
structure 400, which may comprise a structural representation of a
document, such as an HTML web page. As shown, the tree data
structure 400 includes a root node 402 associated with the
document. The parser 304 (FIG. 3) divides the document into
multiple blocks and represents each block of the document as a
table node 404 in the tree data structure 400. Each table node 404
has at least one row node 406 as a child node. Individual row nodes
406 each have at least one column node 408 as a child node. The
column nodes 408 may then have additional table nodes as children.
At this point, the tree data structure 400 may be recursive.
[0040] Thus, the document is divided into blocks, which may be
defined by the structure of the document. The primary content for
each of the blocks, or tables, is stored in the column nodes 408
and the remaining structure of the various blocks is represented in
the other portions of the tree data structure 400.
[0041] Referring again to FIG. 3, the label engine 306 then assigns
labels to individual blocks and may assign a classification to each
block according to the contents of the block. In one embodiment,
the label engine 306 assigns a classification to each block based
on the block contents. For example, if the document is a web page,
the web page may include links, text, forms, and pictures, as well
as other classes of content.
[0042] The label engine 306 optionally analyzes individual blocks
and assigns a classification to the block indicating the type, or
class, of content in the block. Hence, a block that contains
primarily links may be assigned a "navigation" classification, a
block that contains primarily text may be assigned a "story"
classification, a block that contains primarily pictures may be
assigned an "image" classification, and a block that contains form
information like an address may be assigned a "form"
classification. The label engine 306 inserts a classifier
associated with the assigned classification for each block into the
table node of each block.
[0043] After classifying the blocks, the label engine 306
optionally merges, or combines, column nodes of each block that
have the same classification. For example, if a given block has
multiple column nodes having the classification of "story," the
label engine 306 may merge, or combine, the content of these column
nodes. Likewise, if a given block has multiple columns having the
classification of "navigation," the label engine 306 may merge, or
combine, the content of these column nodes.
[0044] In one embodiment, the label engine 306 may merge, or
combine, column nodes in accordance with predetermined merging
rules stored at the label engine 306. An example merging rule is
that a large "story" node is not merged with another large "story"
node. Another example merging rule is that a small "story" node may
get merged with a "navigation" node. Thus, according to these
rules, a large story, which is likely to be substantial enough to
be viewed in isolation, will not be combined with another large
story. However, a small story would not be isolated. Rather, the
user experience may be improved by merging other nodes, such as a
small "navigation" node or a small "story" node. The specifics of
these merging rules may vary and may be customized according to
particular applications. The classifying and merging are optional
according to some embodiments of the invention.
[0045] The label engine 306 also assigns a label to each block
according to the block contents. In one embodiment, the label
engine 306 uses the first several words of text of a block
including text as the label for that block. In another embodiment,
the label engine 306 assigns a label to a block based on the
classification of the block. The label engine 306 then adds the
assigned label to the table node of the associated block.
[0046] With continued reference to FIG. 3, a data structure
converter 308 of the mapper 202 next "flattens" the tree data
structure by converting the tree data structure into a linear,
one-dimensional list containing the content of the column nodes
408. The table nodes 404 and the row nodes 406 are not included in
the one-dimensional list. Individual entries in the one-dimensional
list include the content of an associated column nodes 408.
[0047] A ranking engine 310 then ranks the entries in the
one-dimensional list according to the content of the individual
entries. In one embodiment, the ranking engine 310 analyzes
characteristics of each entry and assigns a "weight" value to each
entry. The weight assigned to each entry may be based on a variety
of parameters. These parameters may include, for example, the size
of the font used in the entry, whether the text in the entry is
boldface, the color of the text, whether the text is flashing,
whether the text is underlined, and the position of the item in the
document.
[0048] The ranking engine 310 may also generate a size value
indicative of an amount of text in each of the plurality of
subdocuments. Pursuant to this embodiment, the size value may be
larger for subdocuments comprising large amounts of text and the
size value may be smaller for subdocuments comprising smaller
amounts of text. Ranking the entries in the table of contents, at
least in part, according to the size value tends to make entries
associated with larger amount of text appear higher on the list of
entries in the table of contents (i.e., or more important or more
relevant).
[0049] In one embodiment, the weight assigned to each entry may
also depend on the content of the link leading to the document, the
text of the previous document, the text of the subdocument
associated with the entry, the text of the label associated with
the entry, or a combination of these. Additional details regarding
this embodiment are described below with reference to FIGS. 7 and
8. Based on parameters such as these, the ranking engine 310
assigns a weight to individual entries in the one-dimensional list
and then re-orders the one-dimensional list according to the
weighted rankings.
[0050] In one embodiment, the ranking engine 310 reorders the list
in an order of decreasing weight values such that the first entry
in the re-ordered list is the entry having the largest weight value
and the last entry in the list the entry having the smallest weight
value. The re-ordered list is then stored as the list of document
content 204 (FIG. 2). Thus, in some embodiments, entries having
large or bold text may be ranked before entries having smaller or
plain text. Also, entries having a graphic may be ranked higher
than entries having primarily links.
[0051] FIG. 5 illustrates details of the control module 206 of FIG.
2 in accordance with one embodiment of the present invention. In
general, the control module 206 receives the list of document
content 204 and creates a new document structure according to a
navigation rules database 502 and the list of document content
204.
[0052] The navigation rules database 502 contains a tree data
structure for one or more documents. In one embodiment, contents of
the navigation rules database 502 may be modified by accessing the
formatter 126 (FIG. 1) from a client computer, such as the client
admin computer 140 (FIG. 1). The database modifier 128 may modify
the contents of the navigation rules database 502 described
above.
[0053] In particular, the client admin computer 140 includes
browser 142 and permits a user to access the database modifier 128
and to modify the contents of the navigation rules database 502. To
modify the contents of the navigation rules database 502, a user at
the client admin computer 140 directs the browser 142 to the
database modifier 128. The database modifier 128 then presents the
user with a GUI (Graphical User Interface) via the browser 142 that
permits the user to view a default tree data structure, as
constructed by the mapper 202, for a given document, such as an
HTML or XML web page document. The default tree structure may be
the structure of the document at issue as determined by parsing the
document.
[0054] The user may then delete entries in the tree data structure.
The user may alternatively move tree data structure entries from
one location to another within the tree data structure. Further,
the user may change the label or classification assigned to given
nodes within the tree data structure. After the user has thus
modified, or customized, the tree data structure, the control
module 206 stores the modified tree data structure as an entry in
the navigation rules database 502 associated with the document.
[0055] The control module 206 also includes a URL (Uniform Resource
Locator) checker 504. The URL checker 504 receives the list of
document content 204 from the mapper 302 and determines whether the
navigation rules database 502 includes a tree data structure
associated with the list of document content 204. In one
embodiment, the URL checker determines whether the URL associated
with the list of document content 204 matches a URL associated with
an entry in the navigation rules database 502. If such a match
exists, an output file generator 506 retrieves the tree data
structure in the navigation rules database 502 associated with the
list of document content 204. The output file generator 506 then
creates one or more output files 508 based on the retrieved tree
data structure using the content of list of document content
204.
[0056] The output files 508, in one embodiment, include a table of
contents (TOC) page that lists the labels of the document. The
output files 508 also contain one or more subdocuments. Individual
sub-pages are associated with individual entries in the TOC. One or
more of the labels, or entries, of the TOC may include links to
associated subdocuments.
[0057] If the URL checker 504 determines that the navigation rules
database 502 does not include a tree data structure associated with
the list of document content 204, then the output file generator
506 generates an output files 508 that include a TOC page that
lists the labels of the document. One or more of the labels, or
entries, of the TOC may include links to associated
subdocuments.
[0058] The formatter 126 then transmits the TOC page over at least
one of the networks 110, 111 to the client 102. Upon receipt of the
TOC page at the client 102, the client 102 displays the TOC page at
the display 112 of the client 102. The user may then select a link
associated with one of the entries of the TOC, which requests an
associated subdocument from the output files 508. In response to a
request for a subdocument in the output files 508, the formatter
transmits the requested subdocument to the client 102 over at least
one of the networks 110, 111 for display at the display 112 of the
client 102.
[0059] FIG. 6 illustrates a flowchart 600, which depicts a method
according to one embodiment of the present invention. The method
commences at block 602 where application server 124 receives a
request for document from the client 102 (FIG. 1), the requested
document residing on at least one of the servers 104, 106, 108. The
request for document may be directed to the application server 124
directly. Alternatively, the request for document may be directed
directly to one of the servers 104, 106, 108, which, in turn,
redirects the request for document to the application server 124.
The request for document may comprise an HTTP request or other
suitable request. Moreover, the requested document may comprise a
document in HTML, XML, PDF, or other suitable format.
[0060] Next, at block 604, the application server 124 retrieves the
requested document from one or more of the servers 104, 106, 108 on
which the document resides. This retrieval may be accomplished by
the application server 124 transmitting an HTTP request to the
server 104, 106, 108 at which the requested document is stored. For
example, if the requested document resides at the server 104, the
application server 124 requests the document from the server 104
over the network 110 and receives the requested document over the
network 110.
[0061] Then, at block 606, the formatter 126 of the application
server 124 extracts a structure of the retrieved document. In one
embodiment, a parser 304 (FIG. 3) parses the retrieved document and
generates a tree data structure representing the structure of the
retrieved document. An example of such a tree data structure is
illustrated in FIG. 4 and is described above.
[0062] For individual nodes of the tree data structure that include
document content, the formatter 126 next analyzes the content of
the nodes and assigns one of a set of predefined classifiers to
each of the nodes based on the content of the nodes, pursuant to
block 608. As discussed above, for a node having content comprising
primarily text, the label engine 306 of the formatter 126 may
assign a "story" classifier to the node. The classifier may
comprise a text string or other identifier added to the node.
[0063] At block 610, the label engine 306 of the formatter 126
assigns labels to individual nodes of the tree data structure that
include document content. The label engine 306 may assign a label
based on the content of the node, the assigned classification of
the node, or both. In one embodiment, the label engine 306 uses the
first several words of nodes having text content as the label for
the associated node. The label may indicate the content of the node
being labeled.
[0064] At block 612, the label engine 306 merges nodes having
content according to their classification. For example, if a pair
of nodes having content both have the classification "navigation,"
then the label engine 306 merges the content of these nodes to form
a single node that includes the content of the merged nodes. Block
612 may alternatively, or additionally, be performed after block
616. In one embodiment, the merging is performed before and after
ranking.
[0065] At block 614, the data structure converter 308 of the mapper
202 converts the tree data structure to a list. The data structure
converter 308 extracts the nodes of the tree data structure that
include content and generates a list comprising the nodes of the
tree data structure that include content, without the other
associated nodes, such as table and row nodes, which do not include
content.
[0066] Next, at block 616, the ranking engine 310 (FIG. 3) of the
mapper 202 reorders the entries of the list generated at block 614.
In one embodiment, the ranking engine 310 assigns a weight value to
each of the entries in the list according to certain parameters of
the content of the entries, the classification of the list entry,
or a combination thereof. Then, the ranking engine 310 reorders the
list according to the weight value of the list entries. For
example, the ranking engine 310 may order the list entries in order
of decreasing weight value. The ranking engine 310 then stores the
re-ordered list as the list of document content 204 (FIG. 2).
[0067] The control module 206 (FIG. 5) then determines whether the
navigation rules database 520 includes an entry associated with the
list of document content 204, pursuant to block 618. In one
embodiment, the URL checker 504 of the control module 206
determines whether a URL associated with the list of document
content 204 matches a URL associated with an entry in the
navigation rules database 502. The URL checker 504 determines that
the navigation rules database 502 contains an entry associated with
the list of document content if such a match exists and execution
proceeds to block 620, else execution proceeds to block 622.
[0068] At block 620, the output file generator 506 creates a new
data tree structure using the list of document content 204 and the
associated entry of the navigation rules database 502. The entry of
the navigation rules database 502 may specify labels to be assigned
to the various nodes, the location of the various nodes within the
new data tree structure, and whether certain nodes are included in
the new data tree structure. The output file generator 506 then
creates a new data tree structure according to the entry in the
navigation rules database 502 and inserts the associated content
from the list of document content 204 to form a new data tree,
which may be stored as the output files 508.
[0069] At block 622, the output file generator 506 stores the new
data tree structure as the output files 508 if the navigation rules
database 502 contains as entry associated with the list of document
content 204. Otherwise, the output file generator 506 stores the
list of document content as the output files 508 or processes the
list of document content from memory. Moreover, the output file
generator 506 may generate device-specific output.
[0070] The output files 508 include a table of contents (TOC) page
that lists the labels of the nodes having content and subdocuments
that include the content of blocks associated with the labels. Each
of the subdocuments is associated with one of the links so that a
user at the client 102 may request a subdocument by selecting the
link associated therewith.
[0071] Lastly, pursuant to block 624, the formatter 126 transmits
the TOC page to the client 102.
[0072] FIGS. 7 and 8 illustrate details of one embodiment of the
operation of the ranking engine 310 described above and illustrated
in FIG. 3. Since, according to some embodiments, each document is
analyzed individually and independently, when a body of text is
followed from one document to another, tracking the body of text is
a consideration for the ease of reading the body of text and
navigating a set of documents. Indeed it is common for a story to
begin on a first document and extend to a second document. Hence,
in some applications, it may be desirable to facilitate
identification of the continuing portion of the story within the
second document, which may be divided into multiple
subdocuments.
[0073] With reference to FIG. 8, the display 112 of the client
device 102 (FIG. 1) displays a subdocument 802 containing text 806
and one or more links 804. The link 804 is a selectable connection
(e.g., a hyperlink) from a word, a set of words, or other
information object, to another. One implementation of the link 804
is a highlighted set of words, or text, that can be selected by a
user, such as with a mouse or by touch-screen control, resulting in
the immediate delivery and view of another file. The highlighted
text may be referred to as an anchor.
[0074] FIG. 7 is a flowchart illustrating a method in accordance
with one embodiment of the present invention. FIG. 8 illustrates an
example sequence of material displayed at the display 112 (see,
FIG. 1). In general, user selection of the link 804 causes the
client 102 to transmit a request for an associated file, such as a
document, from the application server 124. As discussed above with
reference to FIG. 6, when a document is thus requested, the
application server124 generates a table of contents page 810,
including a list of labels, with each label being associated with a
subdocument.
[0075] It is desirable in some applications that the label
associated with the selected link 804 be at or near the top of the
list of labels in the table of contents page 810 to facilitate
navigation and to permit the user to easily locate the label
associated with the selected link. Thus, it is desirable that the
label 812 of the table of contents page 810 be associated with the
selected link to permit the user to quickly and easily identify the
subdocument associated with the selected link 804. The user may
then select the label 812, which comprises a link to the
subdocument 820 containing the text 822.
[0076] Referring to FIGS. 7 and 8, the user at a client 102 (FIG.
1) views a subdocument 802 at a display 112 of the client 102. As
shown in FIG. 8, the subdocument 802 includes text 806 and one or
more links 804. Pursuant to block 701 of FIG. 7, the user selects
one of the links 804 of the subdocument 802.
[0077] The user selection of the link 804 pursuant to block 701
causes the client 102 to transmit a request for a document
associated with the link 804 selected by the user. Pursuant to
block 702, the application server 124 receives the request for
document from the client 102 (FIG. 1), the requested document
residing on at least one of the servers 104, 106, 108. The request
for document may be directed to the application server 124 directly
or to one of the servers 104, 106, 108, which, in turn, redirects
the request for document to the application server 124.
[0078] Next, at block 704, the application server 124 retrieves the
requested document from one or more of the servers 104, 106, 108 on
which the document resides. This retrieval may be accomplished as
described above. At block 706, the formatter 126 of the application
server 124 extracts a structure of the retrieved document as
described above.
[0079] For individual nodes of the tree data structure that include
document content, the formatter 126 next analyzes the content of
the nodes and assigns one of a set of predefined classifiers to
each of the nodes based on the content of the nodes, pursuant to
block 708 as discussed above. At block 710, the label engine 306 of
the formatter 126 assigns labels to individual nodes of the tree
data structure that include document content as discussed above. At
block 712, the label engine 306 merges nodes having content
according to their classification and, at block 714, the data
structure converter 308 of the mapper 202 converts the tree data
structure to a list, as discussed above.
[0080] At block 716, the ranking engine 310 (FIG. 3) compares the
text 806 of the previous subdocument 802 to each of the
subdocuments of the requested document using conventional document,
or text, matching techniques to determine the extent to which the
previous subdocument is associated with each of the subdocuments of
the requested document. The ranking engine 310 may employ an
n-dimensional vector matching technique for comparing the text of
the previous subdocument 802 to each of the subdocuments of the
requested document. Modern Information Retrieval, by R.
Baeza-Yates, et al, published by Addison-Wesley Pub Co; 1999, ISBN:
020139829X, discloses related techniques and is incorporated herein
by reference.
[0081] In comparing the text 806 of the previous subdocument 802 to
each of the subdocuments of the requested document, the ranking
engine 301 generates a document/document value for each of the
subdocuments of the requested document. The document/document value
indicates the degree to which there is an association between the
text 806 of the previous subdocument 802 to each of the
subdocuments of the requested document. For example, if the text
806 of the subdocument 802 included the terms such as "XYZ,"
"merger," "corporate," "shareholders" and the like, the ranking
engine 301 would assign a higher degree of association, and thus
either a higher or lower document/document value, to subdocuments
in the requested page that include the same or similar terms.
[0082] At block 718, the ranking engine 310 (FIG. 3) compares the
text of the selected link 804 to each of the subdocuments of the
requested document. For example, if the selected link 804 comprised
the text "XYX merger" the ranking engine 310 would determine the
degree to which the text "XYZ merger" is present in each of the
subdocuments of the requested document. The ranking engine 301
generates a link/document value for each of the subdocuments of the
requested document. The link/document value indicates the degree to
which the text of the selected link is present in each of the
subdocuments of the requested document.
[0083] At block 720, the ranking engine 310 compares the text of
the selected link 804 to each of the labels of the requested
document. For example, if the selected link 804 comprised the text
"XYZ merger" the ranking engine 310 would determine the degree to
which the text "XYZ merger" is present in each of the labels
assigned to the requested document and would generate a link/label
value for each of the subdocuments of the requested document. The
link/label value indicates the degree to which the text of the
selected link and the subdocuments of the requested document are
related.
[0084] The ranking engine 310 may also use additional factors in
reordering the list entries. For example, the ranking engine may
generate a size value indicative of an amount of text in each of
the plurality of subdocuments. Pursuant to this embodiment, the
size value may be larger for subdocuments comprising large amounts
of text and the size value may be smaller for subdocuments
comprising smaller amounts of text. Ranking the entries in the
table of contents, at least in part, according to the size value
tends to make entries associated with larger amount of text appear
higher on the list of entries in the table of contents.
[0085] At block 722, the ranking engine 310 reorders the list
entries according to the document/document value, the link/document
value, the link/label value or a combination of these values. In
one example embodiment, the ranking engine 310 assigns a weight to
each of the document/document, link/document, and link/label values
and then combines the weighted values to determine the reordering
of the list entries. Pursuant to another embodiment, the ranking
engine 310 reorders the list entries according to one or more of
the document/document, link/document, and link/label values and
other factors, including, for example, amount of content in the
subdocument, the size of the font used in the subdocument, whether
the text in the subdocument is boldface, the color of the text in
the subdocument, whether the text of the subdocument is flashing,
and the position of the item in the document.
[0086] After the ranking engine 310 has reordered the list entries,
execution returns to block 618 of the flowchart of FIG. 6 as
described above. Performing one or more of the blocks 716, 718, 720
together with the block 722 improves user navigation. In
particular, this functionality increases the probability that the
label listed at or near the top of the table of contents 810 will
be associated with the selected link, the text of the subdocument
including the selected link, or both.
[0087] The above-described embodiments of the present invention are
meant to be merely illustrative and not limiting. Thus, those
skilled in the art will appreciate that various changes and
modifications may be made without departing from this invention in
its broader aspects. Therefore, the appended claims encompass such
changes and modifications as fall within the scope of this
invention.
* * * * *