JP2001155017A - Tagged document preparing device and recording medium recorded with program therefor - Google Patents

Abstract

PROBLEM TO BE SOLVED: To efficiently prepare a tagged document by supporting the knowledge of a worker related to tagging and substituting the operation of preparation or editing. SOLUTION: This device is composed of a tag information extracting part 102 for extracting and storing tag information containing the use frequency of an actually used tag from a document, a sentence analytic part 105 for extracting a sentence expression pattern from the document and storing the extracted expression pattern in relation to the said tag information, a use tag inferring part 107 for inferring a tag to be used on the basis of the said stored tag information and expression pattern information, and a document input editing part 108 for inputting the document and applying a tag to the inputted document on the basis of the inferred result of the said inferring part.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a tagged document creating apparatus, and more particularly to a tagged document creating apparatus capable of newly creating or editing a tagged document, and a recording medium recording the program thereof.

[0002]

2. Description of the Related Art In recent years, various needs have arisen for documents, such as handling documents as data and using one document in common in various forms of content, and existing documents and new documents are created as tagged documents. It is desired.

[0003] When creating a tagged document, it is necessary to have knowledge of tags to be added to the document. In addition, since a procedure different from that of a word processor is required to create a document, it is difficult to create or edit the document, and the work load is large.

Japanese Patent Application Laid-Open No. Hei 10-63649 discloses a method of extracting character string information and appearance information from a document created by a word processor or the like, assigning tags based on the extracted information, and creating a tagged document. It is shown.
In Japanese Patent Application Laid-Open No. 9-160917, the coordinates of a character input area in a fixed format are associated with a tag to be added to a character string at the coordinate position, and a tagged document is created based on the information. It has been shown.

[0005]

SUMMARY OF THE INVENTION The above-mentioned Japanese Patent Application Laid-Open No. 10-63 is disclosed.
649 and Japanese Patent Application Laid-Open No. 9-160917 each obtain information required for tagging from a document in a format different from the tagged document, and create a tagged document through a process of conversion. are doing. Therefore, it cannot be applied to the task of directly creating or editing a tagged document.

That is, in the case of directly creating or editing a tagged document, any of the above techniques has a high degree of difficulty in document creation, and the time and effort for tagging is still large.

SUMMARY OF THE INVENTION The present invention has been made in view of the above-mentioned problems. In directly creating or editing a tagged document, a tag which is expected to be used is guessed and displayed to facilitate creation or editing of the tagged document. Is what you do.

[0008]

The present invention employs the following means in order to solve the above-mentioned problems.

A tag information extracting unit for extracting and accumulating tag information including the frequency of use of the actually used tag from the document;
A sentence analysis unit that extracts a sentence expression pattern from a document and stores the extracted expression pattern in association with the tag information, a used tag estimation unit that estimates a used tag based on the stored tag information and expression pattern information, The document input / editing unit includes a document input / editing unit. The document input / editing unit inputs a document, and attaches a tag to the input document based on the estimation result of the estimation unit.

A tag information extraction process for extracting and storing tag information including a frequency of use of a tag actually used from a document, extracting a text expression pattern from a document, and extracting the extracted expression pattern with the tag information A sentence analysis process for associating and accumulating, a used tag inferring process for estimating a used tag based on the accumulated tag information and expression pattern information, and inputting a document. This is a computer-readable recording medium on which a program for causing a computer to execute a document input / edit process for adding a tag is recorded.

[0011]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described below with reference to FIGS. FIG. 1 is a diagram showing a tagged document creation apparatus according to an embodiment of the present invention. In the figure, reference numeral 101 denotes a DTD that defines a document structure by a tag set.
(Document Type Definition) storage unit, 102 is DTD
Alternatively, a tag information extraction unit for extracting tag information from the target tag-added document storage unit 104; a tag information storage unit 103 for storing the tag information extracted by the tag information extraction unit; Documents, 10
5 is a sentence analysis unit that extracts a sentence expression pattern from the read sentence, 106 is a sentence expression pattern storage unit that stores the extracted sentence expression pattern in association with the tag information,
107 is a used tag estimating unit for estimating a used tag by referring to the tag information accumulated in the tag information accumulating unit 103 and the sentence expression pattern accumulated in the sentence expression pattern accumulating unit;
Is a document input / editing unit that creates or edits a document with a tag, and 109 is an operation recording unit that classifies and records the contents of operations performed by the input editing unit 108 such as adding or modifying tags.

The document input / editing unit 108 reads the target tag-added document 104, and performs an operation related to correction or addition of a tag in the read target tag-added document storage unit 104. The tag estimating unit 107, in accordance with an instruction from the document input editing unit 108, the tag information 103 stored in the tag information storage unit, the sentence expression pattern stored in the sentence expression pattern storage unit 106,
The tag which is expected to be used is guessed based on the tag, and the guessed tag is directly added to the tagged document being created / edited by the document input / editing unit 108, or is displayed as a candidate for a tag to be used.

The document input / editing unit 108 classifies operation contents such as addition and correction of a tag, and stores the classification in the operation recording unit 109. Further, the document input editing unit 108 sends the tag added to the edited document, the edited text, and the position information of the tag including the editing target position to the tag information extraction unit 102 and the text analysis unit 105. These pieces of information are stored in the tag information storage unit and the sentence expression pattern storage unit via the tag information extraction unit and the sentence analysis unit. The operation recording unit 109 supports an input operation by reproducing a repeatedly performed operation based on the stored operation record.

When the document input / editing unit 108 is instructed to write back the created or edited tagged documents, the document input / editing unit 108 stores these documents in the target tagged document storing unit 10.
Write back to 4. When an instruction to discard a created or edited document with a tag is issued, the tag information storage unit 1
03, the contents stored in the sentence expression pattern storage unit 106 and the operation recording unit 109 are returned to the state before the target tag-added document 104 is read.

Note that the tag information extracting unit 102, the sentence analyzing unit 105, and the document input / editing unit 108 can operate according to the standard for creating a tagged document without a DTD and a target tagged document. Further, the document input / editing unit 108 includes a tag information storage unit 103 and a sentence expression pattern storage unit 106 corresponding to functions for processing such as determination, change, and cancellation of the applied tag used, and the selected function. And a function of recovering the information stored in the operation record storage unit 109. The recovery is performed when the function of changing or canceling the use tag is executed. When the change is executed, the tag information before the change is deleted and the information on the tag after the change is added. When the application stop function is executed, the tag information before the change is deleted, and the editing target position and the like are returned.

FIG. 2 is a diagram showing tag information stored in the tag information storage section 103. In the figure, 201 is the top node, 202 is the document node, 203 is the data node, 204 is the tag stored in the document node 202,
205 and 206 are tags accumulated in the data node unit 203. These tags are managed in a tree structure starting from the highest node 201. This makes it possible to correlate with the document with the tag being edited or an existing tagged document, or the document structure defined by the DTD, and use the tag information as information that complements the document structure.

The document node unit 202 stores a top-level tag indicating a tag set for creating a tagged document, and enables creation or editing of a tagged document created with a different tag set. For example, in the case of the document node unit 202 in FIG. 2, it indicates that information of a tagged document having the tag A and the tag H as the highest-order tags is stored. The data node unit 203 indicates that information of a tag actually used under the target highest-order tag is stored. For example, in the case of the tag information of FIG.
Regarding the information of the tagged document having the tag 204 as the top tag, the tag B and the tag G under the tag A 204
It indicates that the tag C, the tag D, the tag E, and the tag I were used under 05. Further, the data node unit 20
3 indicate that the tags B205, C206, and the like stored in node 3 are given the number of times the corresponding tag was actually used at that position and a unique node number. For example, in the case of tag B205, 50
It is used once and N11 is assigned as a node number.

The node number given to each node in the tag information is used to grasp the relation with the information stored in the sentence expression pattern.

By accumulating tag information in this manner, the structure of chapters, sections, and sections of a book or the edition, class, and
The fact that the frequency of occurrence of a sentence element in a document structure in a chapter or section structure or the like is characterized can be utilized in the tagged document creation apparatus of the present embodiment. In other words, in a document structure that is expanded from a higher-order structure to a lower-order structure, by accumulating information on the position where the tag is used and the frequency of use of the tag, an overview of the features and rules related to the document expansion can be obtained. You can figure out.

FIG. 3 is a diagram showing a sentence expression pattern stored in the expression pattern storage unit 106. In the figure, 301 is a node number part, 302 is a tag name part, 303
Is a sentence information section. Reference numeral 304 denotes a node number stored in the node number part 301, and reference numerals 305 to 309 denote tags stored in the tag name part.

Each piece of information is managed by the node number stored in the node number section 301. Node number part 301
Are stored in correspondence with the node numbers of the nodes given in the tag information of FIG. 2, so that the relation between the tag information and the sentence expression pattern can be grasped. For example, FIG. 3 shows that the information is information on the tag B of the node number N11 and the tag G of the node number N12 in the data node unit 203 shown in FIG.

In the tag name of the tag name part 302, the name of the tag used immediately below the tag of the corresponding node number is stored. In storing a tag name, the same tag name may be redundantly stored depending on whether the tag alone makes sense or not.

The sentence information unit 303 analyzes one or a plurality of sentences or character strings input to the corresponding tag by the sentence analyzing unit 105, and extracts and stores a character string which is a feature from the sentence or the character string. .

When analyzing a sentence, the sentence analysis unit 105 focuses on a character string such as a particle, a verb, an auxiliary verb, a character string indicating reference or association, a noun indicating a sentence element, and the like, and indicates a specific phenomenon or object. Proper nouns are not included. For example, tags C305 and D306 shown in FIG. 3 indicate that only tag, verb, auxiliary verb, and the like are stored as sentence information, so that this tag is added to information that alone has meaning. The tag D307 includes, as document information, a character string indicating a reference or association such as “shown next” or “street” and a character including a noun indicating a text element such as “feature” or “task”. The storage of the column indicates that the document element is composed of a plurality of elements to make sense.

If there is a character string indicating a reference or association,
You can determine if the tag alone makes sense. A noun indicating a sentence element is used to associate a character string in a sentence with a tag of each document element, and it is possible to further increase the guessing accuracy of a tag expected to be used.
For example, the character string “in the table” of the tag D307 indicates that the table is used in the following description and that the tag for the table is the tag I309 by association. In other words, when there is a sentence including a character string such as “shown in the following table”, a tag I30 associated with “in the table” follows the tag including the character string.
9 will be inserted. Naturally, if the tags indicated by the nouns indicating the sentence elements are different, different tags are associated with each other. If the tags are the same, nouns indicating the plurality of sentence elements are associated with one tag. For example, FIG. 3 shows that the tag E308 is a tag used to indicate a feature and a task, and that the tag I309 is a tag for a table.

By accumulating the sentence expression patterns in this manner, in a tagged document, the relationship between the document elements is clearly indicated by one or more sentences or character strings included in the document elements such as headings and paragraphs. That
This can be utilized in the tagged document creation apparatus of the present embodiment. In other words, based on the sentence included in the document element or the character string included in the character string, determine whether the document element is information that makes sense alone or that is made up of multiple document elements. Can be. This processing is possible because the sentences or character strings included in the edited or existing tagged document are written in accordance with the Japanese grammar.

In the present invention, the sentence analysis unit 1 shown in FIG.
There is no limitation on the processing method or the type of the character string to be focused on in the analysis of the text or the character string performed by 05, and various text analysis methods can be used. For example, a sentence can be analyzed using a natural language processing technique, or a proper noun can be generalized using a thesaurus technique.

FIG. 4 is a diagram showing a record of operations stored in the operation recording unit 109. In the figure, reference numeral 401 denotes a classifying unit, and 402, an operation recording unit, in which the content of an operation associated with an operation type 403 constituting the classifying unit 401 is recorded. Reference numerals 403 to 405 denote operation types constituting the classification unit 401, and one operation is stored for each specific operation or tag.

Operation types 403 constituting the classification unit 401
405 are prepared corresponding to a specific operation or tag, and are stored one by one for each specific operation or tag.
Also, operations such as cursor movement and key input are not included. Thus, unnecessary operations among operations performed in the past are not reproduced. Further, the amount of data to be stored does not increase.

The operation recording unit 402 records the contents of the actual operation in association with the types of operations constituting the classification unit 401. For example, in the recording of the operation shown in FIG.
The operation types “character string selection” 403, “selection of tag B” 404, and “selection of tag B” 404 include “enclose with tag A”, “untagging”, and “clipboard”, respectively. The information indicating the actual operation of “copy to” is stored.

As described above, the operation is recorded by the document input / editing unit 108 shown in FIG. 1 by classifying the operations actually performed when creating or editing the tagged text.
It was saved in. By using the record of this operation, it is possible to reproduce an operation that is repeatedly performed when creating or editing a tagged text. Therefore, by using the record of this operation, the trouble of creating or editing a tagged document can be reduced.

FIG. 5 is a flowchart showing a procedure in which the used tag estimating unit 107 estimates a used tag.

When the position of the document to be edited is changed in the document input / editing section 108 shown in FIG. 1 for editing the tagged document being created or edited, the used tag estimating section 107 starts to guess the used tag. An instruction is issued.

Here, the position to be edited is a position where a tag or a character string can be input in the document structure of the tagged document, and specific position information is defined by an edited or existing tagged document or DTD. The document input and edit unit 108 shown in FIG. Further, the document input / editing unit 108 detects a change in the position to be edited, which is an interface for creating or editing a tagged document displayed on the display of the computer, and is located on the tagged document being created or edited. This is when the existing editing position is changed. For example, when the existing editing position is designated by the cursor, the detection is performed by moving the cursor, and when a specific character string is selected, the detection is performed by changing the state of the selection.

Next, in step S01, a process of acquiring a position to be edited is performed. The process of obtaining the edit target position is performed by obtaining the information of the edit target position after the change from the document input editing unit 108. By this processing, it is possible to acquire information that is the basis for the used tag estimation unit 107 to access the tag information storage unit 103 or the sentence expression pattern storage unit 106 shown in FIG.

Next, in step S02, a list of tags that can be input or specified at the edit target position is acquired from the document input / editing unit 108. By this processing, the used tag estimating unit 107 has grasped the basic information such as the presence / absence, range, and type of the tag that can be specified at the editing target position, and the standard for creating the tagged document, The tag used can be inferred according to the existing tagged document or the document structure defined in the DTD. Further, the list of specifiable tags displays the possibility of use according to the use frequency information stored in the tag information storage unit 103 in a priority order.

Next, in step S03, a tag expected to be used is estimated. The used tag inferring unit 107 includes information accumulated in the tag information accumulating unit 103, the sentence expression pattern accumulating unit 106, and the operation recording unit 109 illustrated in FIG. 1, the editing target position information acquired in the acquisition of the editing target position S01, and The tag to be used is estimated from information obtained by referring to the information of the list of specifiable tags acquired in step S02 as needed.

More specifically, based on the edit target position acquired in step S01, the tag of the tag information shown in FIG. 2 and the information stored in the sentence expression pattern shown in FIG. Is determined on the basis of the position of the tag in the text expression pattern based on the tag and its content character string, and the frequency of use of the tag stored in the associated tag information shown in FIG. Further, referring to the record of the operations shown in FIG. 4, it is checked whether the corresponding operation has been performed, and if the relevant operation has been performed, it is determined whether or not the operation can be reproduced. Check by validity.

When referring to the sentence expression pattern shown in FIG. 3, even if the corresponding information cannot be found,
It is possible to guess the tag expected to be used based on the tag information shown in FIG. In this case, information on the used tags is derived from the information on the frequency of use of the tags. Further, there may be a situation where tag information shown in FIG. 2 and information corresponding to the sentence expression pattern shown in FIG. 3 cannot be found. In this case, based on the information of the specifiable tag list acquired in step S02,
The tag used is estimated based on the document structure being edited or an existing tagged document or the document structure defined in the DTD. Therefore, the processing is not interrupted.

FIG. 6 shows an example of a user interface displayed on the document input / editing unit 108. In the figure, reference numeral 601 denotes a document with a tag, which is a document with a target tag to be edited and displayed so as to be editable. The tagged document 60
1 is not limited to display with a tree structure,
It can be displayed in another display format such as WYSIWYG (What You See Is What You Get) display. What is important here is that the document structure of the tagged document is held by the document input / editing unit 108 shown in FIG. 1 and that the tagged document is displayed in an editable format.

Reference numeral 602 denotes a menu, which displays a list of tags that can be used at the position to be edited in the order in which they are more likely to be used. The menu 602 also displays commands for enabling processing such as confirmation and cancellation. More specifically, the tag that is most likely to be used based on the estimation result of the used tag in step S03 is placed at the top, and other tags may be used based on the priority ranking result in step S02. Display in ascending order. As a result, the time and effort of selecting a tag is reduced, and the usability is improved.

The contents of the menu 602 are the document 6 with the tag.
01 can be changed as appropriate in accordance with the change of the editing target position or the editing operation. For example, FIG. 6 shows that tag C is selected and displayed at the top of menu 602 as a tag that is expected to be used.
This shows that a temporary addition is made at a certain position of the cursor 603 of FIG.

By displaying the tags expected to be used in the menu 602 as described above, it is possible to grasp the tags that can be specified at the edit target position even if the document structure is not correctly grasped. Further, since the tag is automatically and temporarily added to the position where the cursor 603 is located, the trouble of selecting the tag can be greatly reduced. Further, even if an unintended tag is specified and temporarily added, the change can be made only by selecting an item in the menu 602, so that a mistake in tagging can be reduced.

The document input / editing unit 108 shown in FIG.
Can display a use tag estimated by the use tag estimating unit 107 as a candidate in the menu 602, and can provide a function that is not applied to the tagged document being edited. By providing this function, it is possible to create or edit a document with a tag more smoothly.

FIG. 7 is a flowchart showing a processing procedure when the used tag candidate displayed by the used tag estimation unit 107 is selected and used.

First, in step S11, the document input / editing unit 108 receives the estimation result of the used tag from the used tag estimating unit 107 shown in FIG. 1, and determines the use of the used tag. Judgment of the usage of the used tag is based on whether the tag is a tag for modifying the character string in the text, that is, whether the tag can be inserted between the character strings in the tag in which the character string can be input. It can be determined from the definition on the document structure.

In step S12, it is determined whether a tag can be added. That is, it is determined whether or not the used tag is a tag for modifying a character string. If it is determined that the used tag is not a tag for modifying a character string, step S1 is performed.
Then, the process is terminated without adding a tag. In step S13, a tag is added to the tagged document being edited.

When it is determined in step S12 that the tag to be used is a tag for modifying a character string, step S13 is skipped because the position to be edited is changed by moving the cursor during input or editing of a character. To prevent a tag for character string modification from being inserted each time. In general, string modification is less frequent and should not be done unconditionally.
If a tag for modifying a character string is added unconditionally, the time and effort required to release the tag increases.

Whether or not the process of selecting and using the use tag candidate as shown in step 12 is actually necessary is determined by the tag document being created or edited, or the DT.
Determined by the document structure defined in D. For this reason, it is preferable to enable or disable this function. As a result, it is easy to change the interface due to the difference in the document structure, and the work efficiency is improved.

Further, it is possible to create a tagged document in the form of batch processing without using an interface for sequentially editing the tagged document shown in FIG. For example, by giving information of a tag set to be used and text data before tagging, based on the tag information shown in FIG. 2 and the information accumulated in the sentence expression pattern shown in FIG.
A tagged document can be created by the document input editing unit 108 and the used tag estimation unit 107 shown in FIG.

[0051]

As described above, according to the present invention, the operator's knowledge of tagging is supplemented, and the creation or editing operation is performed on behalf of the operator. Can be created well.

[Brief description of the drawings]

FIG. 1 is a diagram showing a tagged document creating apparatus according to an embodiment of the present invention.

FIG. 2 is a diagram illustrating tag information stored in a tag information storage unit.

FIG. 3 is a diagram showing a sentence expression pattern stored in an expression pattern storage unit.

FIG. 4 is a diagram illustrating recording of an operation recorded in an operation recording unit.

FIG. 5 is a flowchart illustrating a procedure in which a used tag estimating unit estimates a used tag.

FIG. 6 is a diagram illustrating an example of a user interface displayed on an input editing unit.

FIG. 7 is a flowchart illustrating a processing procedure when a used tag candidate displayed by a used tag estimation unit is selected and used;

[Explanation of symbols]

 Reference Signs List 101 DTD 102 Tag information extraction unit 103 Tag information storage unit 104 Document with target tag 105 Sentence analysis unit 106 Sentence expression pattern storage unit 107 Tag estimation unit used 108 Document input editing unit 109 Operation recording unit

Claims (5)

[Claims] A tag information extraction unit for extracting and storing tag information including a frequency of use of an actually used tag from a document; extracting a text expression pattern from a document; A sentence analysis unit for associating and accumulating, a used tag estimating unit for estimating a used tag based on the accumulated tag information and expression pattern information, and a document input, and a document input based on the estimation result of the estimating unit. A tagged document creating apparatus comprising a document input / editing section for assigning a tag. 2. The tagged document creating apparatus according to claim 1, wherein the tag information is information on a tag use frequency and information on a hierarchical position in a document structure where the tag is used. 3. The apparatus according to claim 1, wherein the expression pattern information is information indicating a grammatical relationship between a plurality of document elements. 4. The method according to claim 1, wherein the use tag estimating unit includes an operation recording unit, and the storage unit stores the operation tag for each specific operation directly related to the tag selection. A tagged document creation device for recording an operation. 5. A tag information extraction process for extracting and storing tag information including a frequency of use of a tag actually used from a document, extracting a text expression pattern from the document, and extracting the extracted expression pattern with the tag information. Sentence analysis processing for associating and accumulating; use tag estimation processing for estimating a use tag based on the accumulated tag information and expression pattern information; inputting a document; and inputting a document based on the estimation result of the estimating unit. A computer-readable recording medium on which a program for causing a computer to execute a document input / edit process for adding a tag is recorded.