JP2000010981A - Link identifying method and electronic document system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To obtain a method and a system for passively showing a document relating to a reader without interfering with a readout process by estimating user's interest from notes that the user attaches to a source document when reading the source document. SOLUTION: The user's interest is estimated from the notes that the user attaches to the source document when reading the source document. By this identifying method, two kinds of target parts relating to the passage of the source document 16, i.e., a target part 22 relating to the passage with specific notes of the source document and a target part 22 relating to the whole source document are discriminated. Once the relation between the passage and target part 22 of the source document 16 is established, the target part 22 is displayed by clicking a selectable link on the display of the source document 16.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

FIELD OF THE INVENTION The present invention relates generally to electronic document reading systems. In particular, the present invention relates to an electronic document reading system that suggests other related documents when displaying a first document.

[0002]

2. Description of the Related Art Searching for documents similar to documents deemed relevant by a user is known as relevant feedback. Related feedback can be found in "Introduction t" by G. Salton et al.
o Modern Information Retrieval "(McGraw Hill, 1983)
Which is incorporated herein by reference in its entirety. Interfaces that support related feedback still require explicit action on the part of the reader and do not voluntarily suggest related documents. Information search interfaces designed for window-based computing environments typically filter the search results of other related documents through a list of separate windows or by replacing visible documents with search results. Represent. These systems are very interruptive and hinder the reading process.

[0003] Hypertext interfaces can be provided by providing a margin containing links, or by "Hyper
by embedding the link in the text of the source document in a way pioneered by "ties"
Displays a document link related to the source document. This system is based on Schneideman
rman) "User Interface Design for the Hyperti
es Electronic Encyclopedia "(Proceedings of Hyperte
xt '87, November 1987, Chapel Hill, NC), which is incorporated herein by reference in its entirety. However, these links are static and are formed by the hypertext author along with the source document. Some systems, such as Trellis
s) displays links dynamically, but only from a fixed set of links defined earlier. The trellis is described by R. "Programmable Browsing Sema" by R. Furuta et al.
ntics and Trellis "(Proceedings of Hypertext '89, N
ovember 1989, Pittsburgh, PA, ACM Press), which is incorporated herein by reference in its entirety.

[0004] The HieNet System is
The internode similarity measure is used to form a hypertext link based on a link already formed by the hypertext author. This system is described in T. "Hienet: A Us by DTChang
er-Centered Approach for Automatic Link Generatio
n "(Proceedings of Hypertext '93, November 1993, Se
attle, WA, ACM Press), which is incorporated herein by reference in its entirety. When the author creates a link from document A to document B, the system automatically adds links from all documents similar to document A to all documents similar to document B. The anchors of these automatically generated links are indicated by icons in the margins of the various documents. Clicking on the icon will display a pop-up menu containing a list of potential documents ranked for the query. This system also relies on links already created by the author.

[0005] Another conventional system relates to a hypertext-like method of displaying search results. Hinet is
Automatically display the link in the margin, but the margin anchor is not related to the content of the passage adjacent to the anchor. Hinet does not distinguish between links between documents and between passages and links between documents. further,
Hi-Net does not indicate the number and nature of documents that can be searched over margin links.

[0006] Visualization of an information retrieval system (Visualizatio)
n of Information Retrieval System: hereinafter referred to as VOIR). "Que by G. Golovinsky
ries? Links? Is There a Difference? "(Proceedings o
f CHI '97 "(March 1997, Atlanta, GA, ACM Press)," W
hat the Query Told the Link: The Integration of Hyp
ertext and Information Retrieval "(Proceedings of H
ypertext '97, April1997, Southhampton, UK, ACM Pre
ss), each of which is incorporated herein by reference in its entirety. VOIR is a mechanism for dynamically creating and resolving hypertext links using queries computed from text surrounding selected anchors. The VOIR uses a query to retrieve the set of documents associated with the passage containing the selected anchor. VOIR does not show the user links that have a pre-established relationship. Conversely, to submit a query and establish a relationship, the user must wait and select an anchor. VOIR was specifically designed to support interaction information exploration, rather than to facilitate the reading process. Therefore, the focus of VOIR is to support navigation between documents. Therefore, users are expected to make cognitive efforts by browsing. Furthermore, V
OIR does not allow users to annotate or tag documents. The VOIR also does not indicate which link was selected to create a particular display.

[0007] Remembranc agent (Remembranc
e Agent: hereinafter referred to as RA). J. "A Conti" by Lord (BJRhodes) et al.
nuously Running Automated Information Retrieval Sy
stem "(Proceedings of The First International Confe
rence on the Practical Application of Intelligent
Agents in Multi-Agent Technology, PAAM '96, April,
1997, London, UK), which is incorporated herein by reference in its entirety. RA is EMACS
Runs in a window and suggests documentation for the last few lines of text typed by the user. R
A is designed to search the user's private data to suggest documents related to the text being typed. However, these proposals are temporary and relate only to the text currently being written. RA does not support the read task because it constantly replaces suggestions when a user edits a document.

[0008] QRL is a query-based information search interface that uses ink-like marks on text to identify Boolean queries. This system is described in G. "Queries-R-Links by G. Golovchinsky et al .:
Graphical Markup for TextNavigation "(Proceedings
of INTERCHI '93, April 1993, Amsterdam, The Nether
lands, ACM Press), which is incorporated herein by reference in its entirety. Query terms are selected in rectangles. Straight lines connect the rectangles to represent the Boolean AND operator.

[0009]

All of these systems require extensive user interaction or support only writing to generate links to related documents. There is a need for an electronic document reading system that supports reading by generating passive and unobtrusive links to related documents.

[0010]

SUMMARY OF THE INVENTION The present invention provides a method and system for passively indicating a document associated with a reader without interfering with the reading process.

The present invention automatically detects other document parts that the reader may be interested in or other parts of the same document based on the reader's interaction with the part of the source document being read. Intuitively supports reading by When people read the text, they / they often annotate to highlight passages and terms that are of interest or controversy. The presence of such marks and scribbles, or their relative density, can be used as an indicator of relative interest shown by the reader in a particular passage. If multiple documents related to the document being read are utilized, the reader may be interested in finding relevant parts of other documents or other parts of the same document as part of the reading process.

References to other document parts related to a particular passage of interest to the user or to other parts of the same document are placed in the margin of the source document and refer to other documents that are generally similar to the source document. Is inserted as an end note. The systems and methods of the present invention maintain links once identified to facilitate non-linear reading and browsing.

The system and method of the present invention infers a user's interest from annotations made on the source document by the user as the user is reading the source document. Thus, the systems and methods of the present invention minimize cognitive overhead in at least two ways. The method is: 1) No explicit query is required to identify other document parts related to the source document part annotated by the user or other parts of the same document. 2) Selectable links to other related document parts or other parts of the same document are provided out of the way at the margins and at the end of the source document. This example is shown in FIGS.

Also, the system and method of the present invention present suggestions to the reader in a manner that adapts to other interactions, rather than bothering the user with formal interactions. Portions of other documents proposed by the methods and systems of the present invention or other portions of the same document are accessible by the following selectable links. However, the systems and methods of the present invention do not force the user to follow when a proposal is made. Rather, the suggestion can be followed when (or if so) the user understands that the suggestion will be followed. The systems and methods of the present invention display the type of proposed and referenced portion with an icon and provide a level of text to the icon to give the user a better understanding of the target of the link.

Specific embodiments of the present invention will be described below. A first aspect of the present invention is a method for identifying, for a source document, at least one link to a target portion, wherein each target portion is associated with the source document and is annotated by at least one user of the source document. A link identification method for identifying an annotated passage and identifying at least one target portion associated with at least one annotated passage of a source document. A second aspect of the present invention, in the first aspect, further comprises displaying a selectable link of each of the identified target portions on a display of the source document. A third aspect of the invention, in the second aspect, comprises that the selectable link is displayed as an end note of the source document. A fourth aspect of the present invention includes, in the second aspect, the selectable link is displayed near at least one identified annotated passage. According to a fifth aspect of the present invention, in the fourth aspect, the identifying step comprises responding to an annotation of at least one passage of the source document. In a sixth aspect of the present invention, in the fifth aspect, the step of identifying at least one target portion comprises responding to an annotation of at least one annotated passage of the source document. A seventh aspect of the present invention comprises, in the fifth aspect, the selectable link is displayed in a margin of a display of the source document adjacent to at least one annotated passage. An eighth aspect of the present invention, in the second aspect, includes displaying an icon representing the type of the at least one identified target portion. A ninth aspect of the present invention includes, in the second aspect, displaying a title of a document in which the identified at least one target portion has occurred. According to a tenth aspect of the present invention, in the second aspect,
Including displaying a summary of the document in which the at least one identified target portion occurred. An eleventh aspect of the invention is a method according to the first aspect, wherein the step of identifying an annotated passage by at least one user segments the source document into passages and identifies at least one of the passages as having an annotation. Including. According to a twelfth aspect of the present invention, in the first aspect, the step of identifying at least one target portion determines a degree of relevancy based on the user-identified term and the term identified using a related feedback technique. Including. According to a thirteenth aspect of the present invention,
In a twelfth aspect, the step of identifying includes using a weighted sum query. According to a fourteenth aspect of the present invention,
In a first aspect, the method further includes determining whether a selectable link has been selected and displaying at least one target portion identified in response to selection of the selectable link. A fifteenth aspect of the present invention is the method according to the first aspect, wherein the identifying step comprises: identifying the plurality of target portions as being associated with the source document; clustering the identified plurality of target portions; Selecting at least one of the identified plurality of target portions of each cluster representing all of the identified plurality of target documents in each cluster, wherein the selectable link is selected of the identified plurality of target portions. Referencing at least one of the two. According to a sixteenth aspect of the present invention, in the first aspect, the relevance is determined based on a similarity between the at least one target portion and the source document. According to a seventeenth aspect of the present invention,
In a thirteenth aspect, the identifying step further comprises identifying at least one target portion that exceeds a predetermined similarity threshold.

An eighteenth aspect of the present invention is an electronic document system for proposing at least one target portion associated with a source document to a display of the source document, wherein the system identifies an annotated passage of at least one user of the source document. An electronic document system including a processor for identifying at least one target portion as being associated with an annotated passage of a source document. The nineteenth aspect of the present invention further comprises, in the eighteenth aspect, a display displaying a selectable link that references the identified at least one target portion of the display of the source document. According to a twentieth aspect of the present invention, in the nineteenth aspect, the processor identifies at least one target portion based on a degree of association between the at least one target portion and the source document;
The selectable link includes referencing at least one identified target portion. According to a twentieth aspect of the present invention, in the twentieth aspect, the processor identifies at least one annotated passage of the source document and associates at least one target portion with the identified at least one annotated passage. Identifying that the selectable link is displayed near at least one identified annotated passage.
A twenty-second aspect of the present invention includes, in the nineteenth aspect, wherein the selectable link is displayed as an end note of the source document. A twenty-third aspect of the present invention includes, in the nineteenth aspect, wherein the selectable link is displayed in a margin adjacent to at least one annotated passage. A twenty-fourth aspect of the present invention is the nineteenth aspect, further comprising a user interface, wherein the display displays at least one target portion identified in response to a user selecting a selectable link. Including. A twenty-fifth aspect of the present invention is the method according to the nineteenth aspect, wherein the processor identifies a plurality of target portions based on the degree of relevance to the source document, clusters the identified plurality of target portions, Selecting at least one of the plurality of identified target portions of the cluster, representing all of the identified target portions, wherein the selectable link references the selected at least one of the second plurality of portions; including. According to a twenty-sixth aspect of the present invention,
In a nineteenth aspect, a display includes displaying an icon representing the type of at least one identified target portion. A twenty-seventh aspect of the present invention is directed to the first aspect.
In a ninth aspect, a display includes displaying a title of a document in which the at least one identified target portion occurred. A twenty-eighth aspect of the present invention provides:
In a nineteenth aspect, a display includes displaying a summary of a document in which the identified at least one target portion has occurred. A twenty-ninth aspect of the present invention provides:
In an eighteenth aspect, the system further comprises a user input interface, wherein the processor comprises identifying at least one target portion in response to an annotation of the passage of the source document by the user. A thirtieth aspect of the present invention comprises, in the eighteenth aspect, the processor identifying at least one target portion based on the user-identified term and the term identified based on the associated feedback technique. A thirty-first aspect of the present invention comprises, in the thirtieth aspect, the processor identifying at least one target portion based on the weighted sum query. A thirty-second aspect of the present invention provides the method according to the eighteenth aspect, wherein the processor determines a degree of association between the source document and the at least one target part based on a similarity between the context of the target part and the source document. . A thirty-third aspect of the present invention comprises, in the thirty-second aspect, the processor identifying at least one of the plurality of target portions as exceeding a predetermined similarity threshold.

[0017] These and other features and advantages of the present invention are set forth in the following detailed description of the preferred embodiments,
It is clear from this.

[0018]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Preferred embodiments of the present invention will be described below in detail with reference to the drawings.

FIG. 1 shows a block diagram of one embodiment of a document reading system 10 of the present invention. The document reading system 10 includes a display 18
It includes a processor 12 in communication with a first memory 14 for storing a source document 16 currently being read by a user above. The processor 12 also communicates with a second memory 20 that stores the possibly related target portion 22. The target portion may optionally include a portion of the source document 16. Target portion 22 may include the entire document.

The user interacts with and controls the document reading system 10 via any number of conventional input / output devices 24, such as a mouse 26, keyboard 28, or pen-based interface 30. . The input / output device 24 also communicates with an input / output interface (I / O) 31
1 communicates with the processor 12.

As shown in FIG. 1, system 10 is preferably implemented on a programmed general purpose computer. However, the system 10 may comprise a dedicated computer, a programmed microprocessor or microcontroller and any necessary peripheral integrated circuit elements, ASICs or other integrated circuits, hardwired electronic or logic circuits, such as discrete element circuits, programmable logic devices, For example, PLD, PLA, FPGA, or P
It can also be implemented using AL or the like. In general,
Any device on a finite state machine that can implement the flowcharts shown in FIGS. 4 and 5 can be used to implement system 10.

Further, as shown in FIG. 1, the storage devices or memories 14 and 20 are preferably implemented using static or dynamic RAM. However, devices 14 and 20 can also be implemented using floppy disks and disk drives, writable optical disks, disk drives, hard drives, flash memory, and the like. It is also understood that devices 14 and 20 can be either separate parts of a single memory or physically separate memories.

It is further understood that the links 15 and 17 connecting devices 14 and 20 and processor 12 are wired or wireless links to a network (not shown). The network can be a local area network, a wide area network, an intranet, the Internet or any other distributed processing and storage network. In this case, the electronic document 16 is linked from the physically remote memory device 14 to link 1
Withdrawn via 5 and processed by processor 12, target portion 22 is remotely accessed via link 17 in the manner outlined below. In this case, the electronic document 16 and the target portion 22 may be stored locally on some of the other memory devices (not shown) of the system 10.

The method of the present invention is applied to each source document 1
Identify the two types of target parts associated with the six passages. The two types are: 1) the target part 2 associated with a particular annotated passage in the source document.
2 and 2) Target part 22 relating to the entire source document. Once the relationship between the passage of the source document and the target portion 22 has been established, the target portion 22 is displayed by clicking on a selectable link on the display of the source document 16. The system and method of the present invention may show on a single display all target portions 22 identified as being associated with one or more passages of the source document 16.

Examples of references to two types of target portions 22 are shown in FIGS. Source document 16
Target portion 22 associated with a particular passage 32 of
Is the source document 16 near the related passage 32
Is identified by the margin display 34 arranged in the margin of the image. As shown in FIG.
The target portion 22 generally associated with 6 is annotated and shown as an end note 36 in the source document. End note 36 may include the type, title, and summary information of the document in which the target portion occurred.

The present invention is not limited by a method or system for determining the type of annotation or the context of the annotation. Annotations may be structured or unstructured.

FIGS. 4 and 5 are flowcharts outlining the control routine of one embodiment of the method of the present invention. The control routine starts from step S100 and continues to step S110. In step S110, the control routine receives the source document 16 as input and continues to step S120. In step S120,
The control routine segments the source document into a series of passages, and the control routine continues to step S130. The segmentation of step S120 may be performed by any conventional segmentation system or method. In step S130, the control routine stores the segment of the source document as additional target portion 22, and continues to step S140. In this way, the segmented passages of the source document are made available as potentially relevant parts to be referenced.

In step S140, the control routine determines whether an annotation has been entered into the system. Annotations may be entered in any number of different ways using any number of different systems. One preferred method of entering annotations is to use a stylus on a pen-based input device to mark directly on the display of the source document. In step S140, if the control routine determines that no annotation has been entered into the system, the control routine returns to step S140. Step S140
When the control routine determines that a comment has been input in step S150, the control routine continues to step S150.

In step S150, the control routine determines the context of the annotation. The context may be determined at step S150 by any conventional system and method. The context of the annotation is preferably determined using a bounding box. The system creates a bounding box surrounding the annotation mark and extends the bounds of the box to define the context. Horizontal boundaries extend horizontally to the edges of the text, and vertical boundaries extend vertically to include a complete word, sentence, or paragraph. The amount of text surrounded by the bounding box is initially established using a predetermined preference that can be adjusted to suit the user. Further, the user may directly manipulate the bounding box to determine the context of the annotation. After the control routine determines the context of the annotation in step S150, the control routine continues to step S160.

In step S160, the control routine analyzes the annotation context and generates a query that is used to search the target portion 22 database. The system and method of the present invention derive a query from the text of the annotated passage and the nature of the annotation. The control routine then continues to step S170, where a query is used to identify the relevant target portion 22. The relevant target portion 22 is identified using the query generated in step S160 by determining the best matching target portion that exceeds a predetermined threshold. It will be appreciated that any number of conventional methods or systems may be used to determine the best matching target portion, and the present invention is not limited to a particular search method or system. Further, the predetermined threshold may be adjusted by the user according to the user's preference.

Next, the control routine continues to step S180, where the links of each of the identified target portions are displayed. If the link corresponds to a target portion associated with the context of a particular annotation, the link is displayed as a margin display 34 in the margin adjacent to the annotation, as shown in FIG. Alternatively, if the control routine identifies in step S170 a target portion that is entirely associated with the source document, the control routine displays the target portion link as an end note 36, as shown in FIG. Next, the control routine continues to step S190.

In step S190, the control routine includes:
Determine whether the user has selected the displayed link. If the control routine determines in step S190 that the link has been selected, the control routine continues to step S200. In step S200, the control routine displays a target portion corresponding to the selected link. Next, the control routine continues to step S210, where the control routine stops.

In another embodiment, not shown, the control routine processes the displayed target portion as an input source document and returns to step S110, where the target portion may be processed. In another embodiment, if a link is selected in step S190, the entire document with the identified target portion associated with the selected link is displayed, and the entire document with the target portion is sourced in step S110. Entered as a document. However, step S19
If 0 and no link is selected, the control routine continues to step S220.

In step S220, the control routine determines whether the user has entered an end routine command. If the user has not entered the end routine command in step S220, the control routine proceeds to step S220.
Return to 140. Alternatively, if the control routine determines that the user has entered an end routine command in step S220, the control routine continues to step S210, where the control routine stops.

The control routines of the method and system of the present invention may be set to operate when the document reading system 10 is turned on or upon receiving a user input command (eg, an annotation).

Further, the number of target portions identified and / or displayed may be adjusted using a predetermined threshold. This threshold may prevent the display of too many links, which may prevent the system from being useful.

The target portion can be identified using one or more conventional methods. For example, the relevance of the target portion may be determined using a text-based statistical similarity method, or may use a full-text Boolean or stochastic method. Other examples include linguistic or logical analysis methods, speaker identification or recognition and image similarity algorithms. The invention is not limited to any known or future method or system for determining the relevance of a target portion to a source document.

The method and system of the present invention use a passage or portion of a document as a unit of search, rather than the entire document. The target portion 22 is indexed by segmenting the document into tiles that include statistically similar ranges of text. Long documents that talk about many topics tend not to match queries very well,
On the other hand, because certain parts of these documents match so closely, tiles are used as a unit of search rather than documents.

The method and system of the present invention recognizes several distinct ink patterns from which a query is computed. Ink patterns or marks include underlined words, circled words, circled passages, and margin annotations. Preferably, each type of annotation results in a different search engine query. For example, marks that select certain words translate into queries that emphasize these words from other words in the same sentence, while marks that select long passages generate queries that search for similar phrases. Alternatively, a separate query may be computed for each stroke.

The ink strokes may be grouped by time and type of annotation. For example, annotation types may be identified by color, or annotations having the same or similar colors may be grouped to generate a query.

It will be understood that the term annotation as used herein is intended to include text, digital ink, audio, video or any other input associated with a document. It will also be understood that the term document is intended to include text, video, audio and any other media or any combination of media. Furthermore, it will be understood that the term text is intended to include text, strokes or digital ink in bitmap format, audio, images, video or any other structure or content of a document. It will also be appreciated that the term "annotation" is intended to include text, digital ink, audio, images, video or any other input associated with a document.

The term “similarity” or “relevance”
It will be understood that it is intended to include any measure of a document or a related portion of a document or portion of another document or a portion of a document that includes another portion of the same document.

Although described herein with reference to the specific embodiments described above, many alternatives, modifications and variations will be apparent to those skilled in the art. Accordingly, the above described preferred embodiments are illustrative and not restrictive. Various changes may be made without departing from the spirit and scope herein.

[Brief description of the drawings]

FIG. 1 is a block diagram of one embodiment of an electronic document reading system according to the present invention.

FIG. 2 illustrates a source document having an icon in a margin adjacent to an annotated passage.

FIG. 3 illustrates another source document having an end note.

FIG. 4 is a flowchart showing an outline of a control routine according to one embodiment of the present invention.

FIG. 5 is a flowchart showing an outline of a control routine according to one embodiment of the present invention.

[Explanation of symbols]

 10 Document Reading System 12 Processor 14 First Memory 16 Source Document 18 Display 20 Second Memory 22 Target Portion 24 Input / Output Device 32 Associated Passage 34 Margin Display

 ────────────────────────────────────────────────── ─── Continued on the front page (72) Inventor Gene Gorobchinsky United States 94306 California Palo Alto El Camino Real 4250 Number Sea 327 (72) Inventor Mark D. Weiser United States 94301 Palo Alto Greenwood Avenue California 1144

Claims (19)

[Claims] 1. A method for identifying at least one link to a target portion for a source document, wherein each target portion is associated with the source document and identifies an annotated passage by at least one user of the source document. And identifying at least one target portion associated with at least one annotated passage of the source document. 2. The method of claim 1, further comprising displaying a selectable link for each of the identified target portions on a display of the source document. 3. The method of claim 2, wherein an icon representing the type of at least one identified target portion is displayed. 4. The method of claim 2, wherein the title of the document in which the identified at least one target portion occurred is displayed. 5. The method of claim 2, wherein a summary of the document in which the at least one identified target portion occurred is displayed. 6. The step of identifying an annotated passage by at least one user includes segmenting the source document into passages and identifying at least one of the passages as having an annotation.
The method of claim 1. 7. The method of claim 1, wherein identifying at least one target portion comprises determining a relevance based on the user-identified term and the term identified using related feedback techniques. . 8. The method of claim 1, further comprising: determining whether a selectable link has been selected; and displaying at least one target portion identified in response to the selection of the selectable link. the method of. 9. An identification step comprising: identifying a plurality of target portions as being associated with a source document; clustering the identified plurality of target portions; and identifying a plurality of identified targets within each cluster. Selecting at least one of the identified plurality of target portions of each cluster, representative of all of the documents, wherein the selectable link references the selected at least one of the identified plurality of target portions; The method of claim 1, further comprising: 10. The method of claim 1, wherein the relevance is determined based on a similarity between the at least one target portion and the source document. 11. An electronic document system for suggesting at least one target portion associated with a source document on a display of the source document, wherein the electronic document system identifies at least one user-annotated passage of the source document and includes at least one target portion. An electronic document system that includes a processor that identifies the source document as being associated with the annotated passage of the source document. 12. The system of claim 11, further comprising a display displaying a selectable link that references the at least one identified target portion of the display of the source document. 13. The claim, wherein the processor identifies at least one target portion based on a relevance of the at least one target portion to the source document, and wherein the selectable link references the identified at least one target portion. Item 13. The system according to Item 12. 14. A processor identifying at least one annotated passage of the source document, identifying at least one target portion as being associated with the identified at least one annotated passage,
14. The system of claim 13, wherein the selectable link is displayed near at least one identified annotated passage. 15. The system further comprising a user interface,
2. The user interface of claim 1, wherein the display displays at least one target portion identified in response to a selection of the selectable link by the user.
2. The system according to 2. 16. The processor identifies a plurality of target portions based on relevance to the source document, clusters the identified plurality of target portions, and represents all of the identified target portions within each cluster. ,
13. The method of claim 12, wherein at least one of the plurality of identified target portions of the cluster is selected, and the selectable link references the selected at least one of the plurality of second portions.
The described system. 17. The system of claim 11, further comprising a user input interface, wherein the processor identifies at least one target portion in response to an annotation of a passage of the source document by the user. 18. The system of claim 11, wherein the processor identifies at least one target portion based on the user-identified term and the term identified based on related feedback techniques. 19. The system of claim 11, wherein the processor determines a relevancy between the source document and the at least one target portion based on a similarity between the context of the target portion and the source document.