JP2006252167A - Document processing device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a mechanism capable of achieving favorable reading comprehension of a document without making a reader refer to annoying dictionary reference. <P>SOLUTION: When an annotation is added to a word appearing in a Chinese document, and the document is scanned, a description indicating either one of pronunciation, meaning, and grammar/syntax of the word added with the annotation is presented. Each pattern of adding annotation is preliminarily associated with pronunciation, meaning, and grammar/syntax, and the type of description to be presented is uniquely specified in response to a pattern of adding annotation obtained by analyzing a document. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to a technique for presenting explanations of words included in a document.

Various techniques have been proposed so far that support efficient reading of documents written in Japanese or foreign languages. For example, Patent Literature 1 discloses a reading support device that can quickly support reading a foreign language document without forcing the user to perform complicated operations. In Patent Document 2, a paper document is converted into a digital document in which a predetermined hypertext is embedded, thereby constructing continuity / relevance of information between a digital world such as a computer and a paper document. There is a technical disclosure. In Patent Document 3, when a user reads a sentence on a computer monitor, a word or phrase that the user may not know about the sentence is automatically dictionary-drawn according to the reading level of the user and presented. There is a disclosure.
JP-A-8-095978 JP-A-10-283364 JP 2000-089882 A

By the way, in reading a document, especially a document written in a foreign language, the meaning of the word / phrase that has been marked after reading through the document with appropriate marks on the words / phrases that need dictionary lookup appearing in the document. Generally, this is done by a procedure such as searching for a dictionary and then rereading it. And when reading in accordance with such a procedure, you should check later, such as adding a mark with a straight line where you do not know the pronunciation of a word, but adding a mark with a wavy line if you do not understand the meaning. It is more efficient to use different mark appending modes depending on the type of content.
However, when the document is read and read in such a procedure, there is a problem that the dictionary lookup of the marked words becomes troublesome.
The present invention has been devised under such a background, and an object of the present invention is to provide a mechanism that can realize a good reading of a document without forcing the reader to perform a troublesome dictionary lookup operation.

  A document processing apparatus according to a preferred aspect of the present invention, a word-by-word commentary information group in which a pair of several kinds of commentary information relating to one word and a pattern of an annotation addition to a document is collected appears in the document. Comment information storage means stored for each word to be obtained, input means for inputting a bitmap representing a document including one or more words with annotations added thereto, and when a bitmap is input from the input means, Extraction means for extracting an annotation from a document represented by a bitmap, target phrase specifying means for specifying a phrase to which the extracted annotation is added from the bitmap, and the explanation information storage in association with the specified phrase Comment information group specifying means for specifying the comment information group for each phrase stored in the means, and analyzing the extracted annotation. Commentary acquisition means for identifying the pattern of the additional writing mode to which the annotation corresponds, and acquiring commentary information paired with the identified pattern from the specified commentary group by phrase, and commentary for outputting the acquired commentary information Output means.

  In this aspect, the target word specifying unit cuts out an image on the bitmap around the extracted annotation, and performs character recognition on the cut-out image so that the word or phrase to which the annotation is additionally written It may be specified as a character string.

  An annotation erasing unit for erasing the annotation from the input bitmap may be further provided.

  A commentary legend image representing the relationship between the pattern of the additional writing mode associated with various commentary information in the commentary information storage means and the type of commentary corresponding to each pattern is generated, and the generated commentary legend image is the annotation Legend generation means for overwriting a predetermined drawing position in a new bitmap obtained by erasing the image may be further provided.

  In another preferred aspect of the present invention, a document processing apparatus includes, in a document, a description information group by phrase that summarizes each pair of several types of description information related to one word and a pattern of an annotation addition form to the document. Explanation information storage means stored for each word that may appear, input means for inputting a character code string of a document, display means for displaying the input character code string as a character string representing a document, and the display Annotation adding means for adding an annotation to a part of the character string, extracting means for extracting the added annotation, and target phrase specification for specifying a character or a group of characters to which the extracted annotation is added as a phrase A comment information group specifying means for specifying a phrase-specific comment information group stored in the comment information storage means in association with the specified word; Analyzing the extracted annotation to identify the pattern of the additional writing mode to which the annotation corresponds, commentary acquisition means for acquiring commentary information paired with the identified pattern from the specified commentary group by phrase, and Comment output means for outputting the acquired comment information.

  According to the present invention, it is possible to realize good reading of a document without forcing the reader to perform a troublesome dictionary lookup operation.

(First embodiment)
A first embodiment of the present invention will be described.
This embodiment has the following two features. The first feature is that when an annotation is added to a word or phrase appearing in a Chinese document to be read and the document is scanned, “reading”, “meaning”, and “grammar / syntax” The explanation that expresses any of the above is to be presented immediately. The second feature is that each pattern of how annotations are added is associated with “reading”, “meaning”, and “grammar / syntax” in advance, and annotations obtained by analyzing the document are added. The type of commentary to be presented according to how it is done is uniquely specified.
In the following description, “explanatory information” means each piece of information representing any one of “reading”, “meaning”, or “grammar / syntax” regarding a certain phrase.

FIG. 1 is a block diagram showing a hardware configuration of the document processing apparatus according to the present embodiment. As shown in the figure, the apparatus includes a comment information storage unit 10, a document image input unit 11, an annotation extraction unit 12, a target phrase specifying unit 13, a character recognition unit 14, a comment acquisition unit 15, and a comment output unit 16. .
The comment information storage unit 10 has two databases, a comment database 10a and a write-once pattern database 10b.
FIG. 2 is a data structure diagram of the explanation database 10a. This database is a collection of a plurality of records each having four fields of “word / phrase”, “reading”, “meaning”, and “grammar / syntax”. In the “word / phrase” field, a character code representing each word / phrase that can appear in the document is stored. In the three fields of “reading”, “meaning”, and “grammar / syntax”, various explanatory information is stored. Specifically, the commentary information indicating the reading of the phrase is in the “reading” field, the commentary information indicating the meaning of the phrase is in the “meaning” field, and the description information indicating the grammar and syntax is in the “grammar / syntax” field. Each is remembered.

  FIG. 3 is a data structure diagram of the postscript pattern database 10b. This database is a collection of three records each having two fields of “explanation type” and “annotation additional writing mode”. An explanation type identifier is stored in the “explanation type” field of each record. The commentary type identifier is an identifier that represents three types of commentary: “reading”, “meaning”, and “grammar / syntax”. In the “additional writing mode” field, an additional writing mode identifier is stored. The additional recording mode identifier is an identifier that represents a pattern of how an annotation is added. The pattern of additional writing is as follows: “triangle”, which is the method of adding the target word with a triangle, “circle”, which is the way of adding the target word, and a line under the target word. There are three "underline" that is the method of additional writing. Referring to FIG. 3, in this database, the explanation of “reading” is a pattern of “triangle”, the explanation of “meaning” is a pattern of “circle”, and the explanation of “grammar / syntax” is a pattern of “underline”. It can be seen that they are associated with each other.

  Next, an overview of the function of each unit shown in FIG. 1 will be described. Document image data is input from the document image input unit 11 of the apparatus. This document image data is bitmap data representing a document including one or more words / phrases to which annotations are added. The annotation extraction unit 12 extracts an annotation from the document image data. The target phrase specifying unit 13 specifies, from the document image data, an image of the phrase to which the annotation extracted by the annotation extracting unit 12 is added. The character recognition unit 14 converts the phrase specified by the target phrase specifying unit 13 into a character code. The comment acquisition unit 15 acquires, from the comment information storage unit 10, the comment information corresponding to the pattern of the additional writing mode to which the annotation extracted by the annotation extraction unit 12 corresponds and the word / phrase encoded by the character recognition unit 14. The acquired comment information is printed on a predetermined sheet by the comment output unit 16.

FIG. 4 is a flowchart showing the operation of the present apparatus.
In step 100 shown in the figure, document image data is input from the document image input unit 11. The input document image data is stored in a bitmap memory (not shown).
When the document image data is stored in the bitmap memory, the annotation extraction unit 12 extracts one of the annotations described in the document data (S110).

When the annotation is extracted, the target phrase specifying unit 13 cuts out the image of the phrase in which the annotation has been added from the document image data in the bitmap memory (S120). Subsequently, the character recognizing unit 14 attempts character recognition on the image cut out in step 110, and stores a character code representing a word obtained by the recognition in a recognition result memory (not shown) (S130).
When the character code of the phrase is stored in the recognition result memory, the comment acquisition unit 15 analyzes the annotation extracted in step 110, so that the annotation is added as “triangle” or “circle”. , “Underline” pattern is identified (S140).

Subsequently, the comment acquisition unit 15 reads the type identifier associated with the pattern identified in step 140 from the additional write pattern database 10b (S150).
Further, the comment acquisition unit 15 identifies a record corresponding to the phrase represented by the character code stored in the recognition result memory from the comment database 10a, and among the three types of comment information stored in the identified record, The comment information corresponding to the type identifier read in step 150 is acquired (S160).

The comment acquisition unit 15 stores the comment information acquired in step 160 in a page memory (not shown) (S170).
When step 170 is executed, it is determined whether annotations that have not yet been extracted remain in the document image data in the bitmap memory. If annotations that have not been extracted remain, the process returns to step 110 and the subsequent processing is repeated. On the other hand, when all the annotations have been extracted, the process proceeds to step 180, and the comment output unit 16 adds the comment information obtained by superimposing the comment information stored in the page memory on the document image data in the bitmap memory. Print the document on paper.

Here, the relationship between a Chinese document image to be read and a document with explanation will be described with a specific document example.
FIG. 5A shows a Chinese document image to be read, and FIG. 5B shows a document with a comment obtained through processing by this apparatus.
Referring to FIG. 5A, the annotation 21 is added to the words in the top sentence, the annotation 22 is added to the words in the second sentence from the top, and the annotation 23 is added to the words in the sentence on the third line. Has been. On the other hand, referring to FIG. 5 (B), commentary information indicating the reading “nan (second voice)” is added to the phrase with the annotation 21 added, and “˜da” is added to the phrase with the annotation 22 added. Commentary information representing the meaning of “I think” is associated, and the phrase “annotation 23” is added to the phrase with “23 [A] Yue [B]”. ] "To indicate. It can be seen that commentary information representing the syntax “is associated. This means that the annotation 21 is added to the “triangle” pattern, the annotation 22 is added to the “circle” pattern, and the annotation 23 is added to the “underline” pattern. It means that it was judged.

  In the present embodiment described above, when an annotation is added to a word or phrase appearing in a Chinese document to be read and the document is scanned, “reading” and “meaning” are determined according to the pattern of how the annotation is added. , And explanations representing either "grammar / syntax" are immediately presented. Therefore, a user who reads and understands a document can proceed with reading and understanding a desired explanation as appropriate.

(Second Embodiment)
A second embodiment of the present invention will be described.
In the above embodiment, a Chinese document to be read is input as a bitmap, and a phrase is specified by recognizing a character on the bitmap. On the other hand, this embodiment has a configuration in which character recognition is not performed by inputting a document to be read in a character code string data format.

FIG. 6 is a block diagram showing a hardware configuration of the document processing apparatus according to the present embodiment. As shown in the figure, the present apparatus includes an explanation information storage unit 10, a document data input unit 17, a display unit 18, an annotation addition unit 19, an annotation extraction unit 12, a target phrase specifying unit 13, an explanation acquisition unit 15, and an explanation output. The unit 16 is provided. Unlike the first embodiment, the character recognition unit 14 is not provided.
The comment information storage unit 10 has a comment database 10a and a write-once pattern database 10b, and the data structures of both databases are the same as those in the first embodiment.

  Next, an overview of the function of each unit shown in FIG. 6 will be described. Document data is input from the document data input unit 17 of the apparatus. This document data represents a Chinese document as a character code string. The display unit 18 is a touch display that serves as both a display device and an input device, and displays document data input via the document data input unit 17 as a character string representing a document. The annotation appending unit 19 is a stylus pen, and appends an annotation to an arbitrary character or a group of characters on the display device. The annotation extraction unit 12 extracts the annotation added on the display unit 18. The target word / phrase specifying unit 13 specifies a character code to which an annotation is added or a group of characters as a word. The comment acquisition unit 15 acquires, from the comment information storage unit 10, comment information corresponding to the pattern of the additional writing mode to which the annotation extracted by the annotation extraction unit 12 corresponds and the word specified by the target word specifying unit 13. The comment output unit 16 causes the display unit 18 to display the comment information acquired by the comment acquisition unit 15 as a pop-up.

FIG. 7 is a flowchart showing the operation of the present apparatus.
In step 200 shown in the figure, document data is input from the document data input unit 17. The input document data is stored in a document data memory (not shown).
The document data stored in the document data memory is displayed on the display unit 18 as a Chinese character string representing the document (S210).

The user of this apparatus reads the Chinese character string displayed on the display unit 18 and, when a phrase with unknown reading, meaning, or grammar appears, adds an annotation to the phrase using the annotation appending unit 19. .
When the annotation is added, the annotation extraction unit 12 extracts the added annotation (S220).

The target word / phrase specifying unit 13 specifies a character or a character string to which an annotation is added as a word / phrase (S230).
Next, when the comment acquisition unit 15 analyzes the annotation extracted in step 220, whether the annotation is added to any of “triangle”, “circle”, or “underline” pattern. Identify (S240).

The comment acquisition unit 15 reads the type identifier associated with the pattern identified in step 240 from the additional write pattern database 10b (S250).
Further, the comment acquisition unit 15 identifies a record corresponding to the phrase represented by the character code stored in the recognition result memory from the comment database 10a, and among the three types of comment information stored in the identified record, The comment information corresponding to the type identifier read in step 250 is acquired (S260).

  The comment acquisition unit 15 displays the comment information acquired in step 260 on the display unit 18 as a pop-up (S270). As will be described later, the comment information pop-up is displayed on the display unit 18 together with a button labeled “Close comment”. When the user selects this button, the comment information pop-up is erased. After that, the state returns to step 210. Then, when the annotation is added again by the user, the processing after step 220 is repeated.

Here, the transition of the state from when the Chinese document data to be read is displayed on the display unit 18 until the commentary information is displayed as a pop-up will be described with a specific document example.
FIG. 8A shows the display state of the display unit 18 at the stage where step 210 is executed. FIG. 8B shows a display state at a stage where annotations are additionally written. FIG. 8C shows the display state of the display unit 18 at the stage where step 270 is executed.
Referring to FIG. 8 (A), in the upper part of the screen, buttons indicating “close document” and “close commentary” are displayed, and below that, the Chinese language to be read is displayed. The document is displayed in several columns. In FIG. 8B, annotations are added to the words in the second sentence from the top. Referring to FIG. 8C, it can be seen that the commentary added with the annotation is associated with commentary information indicating the meaning of “I think”. This means that it is determined that the annotation is added to the “circle” pattern.
According to the present embodiment described above, it is possible to refer to commentary information about a word as a popup simply by adding an annotation to a predetermined word on the displayed document using a stylus pen.

(Third embodiment)
A third embodiment of the present invention will be described.
In this embodiment, after inputting the Chinese document image data in a state in which an annotation unrelated to “reading”, “meaning”, or “grammar / syntax” is added and deleting the annotation, “reading”, A function for overwriting the correspondence between "meaning" and "grammar / syntax" and the pattern of additional writing mode, and then printing it on paper and outputting it.

  FIG. 9 is a block diagram showing a hardware configuration of the document processing apparatus according to the present embodiment. As shown in the figure, this apparatus includes a comment information storage unit 10, a document image input unit 11, an annotation extraction unit 12, a target phrase specifying unit 13, a character recognition unit 14, a comment acquisition unit 15, a comment output unit 16, An annotation erasing unit 20, a commentary legend generating unit 21, and a document output unit 22 with a legend are provided.

  Data structure of both databases included in the comment information storage unit 10, comment information storage unit 10, document image input unit 11, annotation extraction unit 12, target phrase specifying unit 13, character recognition unit 14, comment acquisition unit 15, and comment output unit The function of 16 is the same as that of the first embodiment.

  The annotation erasing unit 20 erases only the annotation from the bitmap input from the document image input unit 11. The commentary legend generating unit 21 generates a commentary legend image representing each aspect relationship between the commentary type and the additional writing form in a new bitmap obtained by deleting the annotation by the annotation deleting unit 20. The legend-added document output unit 22 overwrites the commentary legend image on the bitmap from which the annotation has been deleted, and then prints it on the sheet.

FIG. 10 is a flowchart showing the operation of this embodiment.
In step 10 shown in the figure, document image data is input from the document image input unit 11. This document image data is a bitmap obtained by scanning a document in which annotations unrelated to “reading”, “meaning”, or “grammar / syntax” are added. The input document image data is stored in a bitmap memory (not shown).
When document image data is stored in the bitmap memory, the annotation extraction unit 12 extracts one of the annotations described in the document data (S20).
Next, the annotation erasing unit 20 erases the annotation extracted in step 20 from the document image data in the bitmap memory (S30).

When step 30 is executed, it is determined whether annotations that have not yet been extracted remain in the document image data in the bitmap memory. If annotations that have not been extracted remain, the process returns to step 20 and the subsequent processing is repeated. On the other hand, when all the annotations have been extracted, the process proceeds to step 40, where the commentary legend generating unit 21 generates a commentary legend image representing the correspondence between the commentary type identifier and the recordable form identifier in each record of the write-once pattern database 10b. .
When the explanatory legend image is generated, the legend-added document output unit 22 overwrites the explanatory legend image on the document image data from which all annotations have been deleted by the annotation deleting unit 20, and the document image with the legend obtained by the overwriting. Is printed on paper and output (S50).

  The user advances the precedent overwritten document output in step 50 while adding a predetermined annotation to a word whose “reading”, “meaning”, or “grammar / syntax” is unknown. Then, the document image data obtained by scanning the document for which the annotation has been added is input from the document image input unit 11 again. When the document image data is input, the operations after step 100 shown in FIG. 4 are executed.

Here, the relationship between a document in which an annotation irrelevant to “reading”, “meaning”, or “grammar / syntax” is added and a document with a legend will be described with a specific example of the document.
FIG. 11A is a document in which annotations unrelated to “reading”, “meaning”, or “grammar / syntax” are added, and FIG. 11B is obtained through processing by this apparatus. A document with a legend. Referring to FIG. 11A, it is understood that annotations unrelated to “reading”, “meaning”, or “grammar / syntax” are added to the sentences in the second and third rows from the top. . On the other hand, referring to FIG. 11 (B), the annotation added in FIG. 11 (A) is deleted, and instead of “Legend”, “△ Reading”, “○ Meaning”, “Underline Grammar / Syntax It can be seen that the explanatory legend image in which the note “” is surrounded by a rectangle is overwritten.

  According to the present embodiment described above, the annotation is temporarily deleted from the document in which the annotations unrelated to “reading”, “meaning”, or “grammar / syntax” are written, and “reading”, “meaning” Or, an annotation indicating that “grammar / syntax” is unknown can be added again.

(Other embodiments)
The invention according to the present application can be variously modified.
The above embodiment has been described on the assumption that the document to be read is written in Chinese, but the present invention may of course be applied to a document written in another language.

  Further, in the postscript pattern database 10b of the above embodiment, the explanation of “reading” is a pattern of “triangle”, the explanation of “meaning” is a pattern of “circle”, and the explanation of “grammar / syntax” is “underlined”. Was associated with each of the patterns. On the other hand, in accordance with the type of language describing the document to be read, the correspondence relationship between the commentary type and the additional writing mode pattern in the database may be different. For example, as shown in FIG. 12A, “reading” is used when only one character is added with an annotation, “meaning” when there are two characters, and “character” when there are three or more characters. It may be associated with “grammar / syntax”. Such correspondence is particularly suitable for reading Japanese and Chinese documents. Also, as shown in FIG. 12 (B), when the character to which the annotation is added is 2 characters or less, it can be associated with “meaning” and when it is 3 characters or more, it can be associated with “grammar / syntax”. Good. Such a correspondence relationship is particularly suitable for reading a Korean document constructed by phonograms that do not require reading. In addition, as shown in FIG. 12C, when the character with the annotation is one word, “reading and meaning”, and when it is two or three words, “phrase meaning”, When there are four or more words, they may be associated with “grammar / syntax”. Such a correspondence relationship is suitable for reading a document in an Indo-European language such as English.

It is a hardware block diagram of a document processing apparatus. It is a data structure figure of an explanation database. It is a data structure figure of a postscript pattern database. It is a flowchart which shows operation | movement of a document processing apparatus. Chinese document image and document with explanation. It is a hardware block diagram of a document processing apparatus. It is a flowchart which shows operation | movement of a document processing apparatus. It is a figure which shows the transition of a document image. It is a hardware block diagram of a document processing apparatus. It is a flowchart which shows operation | movement of a document processing apparatus. It is a figure which shows the transition of a document image. It is a data structure figure of a postscript pattern database (modification).

Explanation of symbols

DESCRIPTION OF SYMBOLS 11 ... Document image input part, 12 ... Annotation extraction part, 13 ... Target word specific | specification part, 14 ... Character recognition part, 15 ... Explanation acquisition part, 16 ... Explanation output part, 17 ... Document data input part, 18 ... Display part, DESCRIPTION OF SYMBOLS 19 ... Annotation appending part, 20 ... Annotation deletion part, 20 ... S, 21 ... Annotation, 21 ... Explanation legend generation part, 22 ... Document output part

Claims (5)

Commentary information storage means for storing a commentary group by phrase that summarizes each pair of several kinds of commentary information related to one word and a pattern of an annotation addition to the document for each word that can appear in the document;
An input means for inputting a bitmap representing a document including one or more words to which annotations are added;
When a bitmap is input from the input unit, an extraction unit that extracts an annotation from a document represented by the bitmap;
A target phrase specifying means for specifying from the bitmap the phrase to which the extracted annotation is added;
Commentary information group specifying means for specifying a word-specific commentary information group stored in the commentary information storage means in association with the specified word;
Analyzing the extracted annotation to identify the pattern of the additional writing mode to which the annotation corresponds, commentary acquisition means for acquiring commentary information paired with the identified pattern from the specified commentary group by phrase,
A document processing apparatus comprising: comment output means for outputting the acquired comment information. The document processing apparatus according to claim 1,
The target phrase specifying unit extracts an image on the bitmap around the extracted annotation, and performs character recognition on the extracted image, thereby specifying the phrase in which the annotation is added as a character or a character string. Yes Document processing device. The document processing apparatus according to claim 1 or 2,
A document processing apparatus further comprising annotation erasing means for erasing an annotation from the input bitmap. The document processing apparatus according to claim 3.
A commentary legend image representing the relationship between the pattern of the additional writing mode associated with various commentary information in the commentary information storage means and the type of commentary corresponding to each pattern is generated, and the generated commentary legend image is the annotation A document processing apparatus further comprising a legend generating means for overwriting a predetermined drawing position in a new bitmap obtained by erasing data. Commentary information storage means for storing a commentary group by phrase that summarizes each pair of several kinds of commentary information related to one word and a pattern of an annotation addition to the document for each word that can appear in the document;
An input means for inputting a character code string of a document;
Display means for displaying the inputted character code string as a character string representing a document;
Annotation adding means for adding an annotation to a part of the displayed character string;
Extracting means for extracting the appended annotation;
A target phrase specifying means for specifying, as a phrase, a character or a group of characters to which the extracted annotation is added;
Commentary information group specifying means for specifying a word-specific commentary information group stored in the commentary information storage means in association with the specified word;
Analyzing the extracted annotation to identify the pattern of the additional writing mode to which the annotation corresponds, commentary acquisition means for acquiring commentary information paired with the identified pattern from the specified commentary group by phrase,
A document processing apparatus comprising: comment output means for outputting the acquired comment information.

Images