JP2004070376A - Document display device and method therefor - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To display a key phrase in a document by a display attribute according to a category. <P>SOLUTION: A document storage 11 stores a document of reading object. A text extraction part 12 extracts a content text from each page of the document of the reading object. The extracted text is stored in a text storage 13 by page as a file. A unique expression extraction part 14 receives a text file for each page from the text storage part 13; extracts a unique expression included in the text file; outputs unique expression information including the category, a unique expression character string, positional information, or the like of the unique expression; and stores the unique expression information in a unique expression storage part 15. A display control data storage part 17 stores a color for displaying for each category. A display data generation part 16 under the control of a display control part 18 generates display data accompanying classification-by-color information to be displayed on a display unit 19. The display unit 19 displays the document by page and distinguishes a unique expression appearance part by color according to the category thereof. <P>COPYRIGHT: (C)2004,JPO

Description

[0001]
TECHNICAL FIELD OF THE INVENTION
The present invention relates to a technique for assisting in reading a sentence, in which particularly important words and phrases are classified into categories and displayed with different display attributes for each category.
[0002]
[Background Art]
As a large number of documents are digitized, an enormous amount of documents have become available. The user needs to determine whether such an enormous amount of documents is useful, meets the purpose, or the like. Regardless of whether you read the document itself or create a summary and read it, ultimately humans need to read the document and its summary to determine its value. It is desired that such reading work can be easily performed.
[0003]
The present inventors have conducted intensive research and have developed a support technology for reading a large amount of electronic documents relatively easily and at high speed.
[0004]
In other words, when a user reads a document, if a word of 5W1H (who, what, when, where, why, how) or an occurrence of a similarly important word is highlighted, a large number of documents are required. Despite the fact, it has been found that it is relatively easy to determine whether a document is relevant. In the following, proper nouns such as person names, place names, and organization names, and words equivalent to 5W1H such as date and time and price (currency) may be referred to as proper expressions. The present inventors have arrived at the knowledge that it is effective to divide and display such named expressions into categories. Specific names of the category corresponding to Who include person names and organization names (company names), specific names of the category corresponding to Where include place names, and specific names of the category corresponding to When include date and time expressions. is there.
[0005]
When displaying a sentence, it is desirable to color-code the occurrence of an important word (specific expression) according to the category, and display the hit location from the category list display.
[0006]
The following is known as a prior art related to the present invention.
(1) A search system (Namazu, a full-text search system developed by the Namaz project) that displays the appearance of the search key in the same color, and a search system (Google trademark) that displays the search key in different colors for each search key. Are known.
(2) Japanese Patent Laying-Open No. 2000-99526 discloses an apparatus for displaying information extracted from document data in a limited display area at a scroll speed that is easy for a user to read. It refers to documents that contain proper nouns as the subject of action.
[0007]
In the case where the search key is displayed in a different color as in the prior art (1), the words in which category cannot be identified at a glance, and must be carefully read and determined by a human. .
[0008]
Further, even in the case of extracting and displaying a document including proper nouns as in the related art (2), the identification of the display according to the category is not mentioned. I needed to judge.
[0009]
It is not efficient to read through the entire document when determining whether the document is related. Documents are searched by a conventional search engine, and when a document is composed of a plurality of pages, it is necessary to read all pages. It is also desirable to be able to read only a more limited range.
[0010]
[Problems to be solved by the invention]
The present invention has been made in view of the above circumstances, and has as its object to provide a technique for displaying an occurrence of an important word (specific expression) included in a document with a display attribute corresponding to a category. Another object of the present invention is to provide a technique for limiting the range of reading. In addition, this invention is not limited only to the above-mentioned object. Other objects of the present invention will be understood from the following description.
[0011]
[Means for Solving the Problems]
According to the present invention, in order to achieve the above object, a configuration as described in the claims is adopted. Here, before describing the invention in detail, the description of the claims will be supplementarily described.
[0012]
That is, according to one aspect of the present invention, in order to achieve the above object, a document display device includes: means for extracting a phrase having an arbitrary attribute included in a plurality of predetermined attribute groups from a document; Means for associating a display attribute corresponding to each attribute with each of the words; displaying at least a part of the document, and displaying the word included in at least a part of the document with a corresponding display attribute Document display means is provided.
[0013]
In this configuration, the appearance of the phrase can be highlighted with the display attribute corresponding to the attribute of the phrase. Therefore, does the user need to immediately understand only that part, understand the general contents, and further peruse it? It is possible to determine whether the document is not worth reading and skip reading comprehension.
[0014]
The predetermined attribute group is a word (5W1H category) that is important in understanding the context of a proper noun or the like or a word that is equally important. Which words are applicable depends on the contents and purpose of document processing and the like, and is not limited.
[0015]
According to another aspect of the present invention, a document display device includes: means for extracting, from a document, a phrase having an arbitrary attribute included in a plurality of predetermined attribute groups; Means for displaying words that are categorized and displayed; means for selecting an extracted word or an attribute of the extracted words based on a user operation; and displaying at least a part of the document and the document And document display means for displaying the selected phrase or the phrase having the selected attribute with a corresponding display attribute.
[0016]
In this configuration, the extracted words are classified and displayed according to their attributes, so that the general skeleton of the document can be grasped from the displayed contents, and the words and the like in the skeleton can be grasped as necessary. To highlight it in the document to see the content of the phrase in a detailed context.
[0017]
According to still another aspect of the present invention, a document search device includes: means for storing a plurality of documents; means for dividing a document into document segments; means for extracting an index for each of the document segments. Index storage means for storing the association between the document segment and the index; search means for performing a keyword search by referring to the index storage means for each document segment; and means for displaying a search result by the search means. Provided.
[0018]
In this configuration, the search can be performed not by the document unit but by the document segment unit, so that the context matching the keyword can be narrowed down and reviewed, and all pages of the hit document can be read as in the past. No more hassle.
[0019]
The document segment is, for example, a page. In addition, any document component such as a chapter or a section may be used.
[0020]
A phrase having an arbitrary attribute included in a plurality of predetermined attribute groups may be extracted from a document segment and used as an index. A phrase extracted when displaying a hit document segment may be highlighted and displayed with a display attribute corresponding to the attribute.
[0021]
The present invention can be realized not only as a device or a system but also as a method. In addition, it goes without saying that a part of such an invention can be configured as software. Also, it goes without saying that a software product used for causing a computer to execute such software is also included in the technical scope of the present invention.
[0022]
The above aspects of the present invention and other aspects of the present invention are set forth in the following claims, and will be described in detail below with reference to embodiments.
[0023]
BEST MODE FOR CARRYING OUT THE INVENTION
Hereinafter, embodiments of the present invention will be described.
[0024]
[First Embodiment]
FIG. 1 shows a first embodiment in which the present invention is applied to a document processing apparatus suitable for browsing and reading a large number of documents stored in a storage unit such as a file server. In the document processing device, the document storage unit 11, the text extraction unit 12, the text storage unit 13, the named expression extraction unit 14, the named expression storage unit 15, the display data generation unit 16, the display control data storage unit 17, the display control unit 18 and a display unit 19.
[0025]
The document storage unit 11 stores a document to be read. The document storage unit 11 only needs to specify and retrieve one document one by one, and may store a plurality of documents. The text extracting unit 12 extracts a content text from each page in a document to be read. The extracted text is stored in the text storage unit 13 as one file for each page. As described later, it is preferable to attach a suffix indicating a page number to a document name so that the user can know which page of which document (for example, “foo — 01”).
[0026]
The named entity extraction unit 14 receives a text file for each page from the text storage unit 13 and extracts named entities included therein. In this example, morphological analysis and a predetermined extraction rule are used. This will be described later with an example. The named entity extraction unit 14 extracts named entities and outputs named entity information including the category, named entity character string, position information, and the like. The named entity information is stored in the named entity storage unit 15.
[0027]
The display control data storage unit 17 stores colors to be displayed for each category. The display control unit 18 performs various controls related to display. For example, it controls which category is enabled for color coding and in what display form. The display data generation unit 16 generates display data to be displayed on the display unit 19 under the control of the display control unit 18. For example, data in an HTML (or XML) format including color control, page selection, and frames is generated.
[0028]
The display unit 19 displays a page of a document or the like. At this time, the appearance of the unique expression is classified by color according to the category. The display unit 19 is, for example, a so-called web browser for rasterizing an HTML file or the like.
[0029]
Next, the operation of this embodiment will be described with a specific example.
[0030]
The text for one page extracted by the text extraction unit 12 is, for example, as shown in FIG. The named entity extracting unit 14 extracts named entity information from the text as shown in FIG. The extraction result is described in, for example, XML, but is not limited to this. Each entity expression information is described using a tag of “entity” and includes a character string, a category, and position information (offset from the beginning of the text and length of a phrase). The character string is, for example, “Fuji Xerox” (trademark), “Yamada”, “$ 100”, or the like. The categories are "COMPANY", "PERSON", "CURRENCY", and the like. The position information is represented by the number of bytes.
[0031]
The display data generation unit 16 generates an HTML file using the text, and attaches a tag for color-coding the appearance position of the named entity using the named entity information. For example, “<FONT COLOR =“ color code ”>” and “</ FONT>” enclose a specific expression portion that is color-coded. The coloring of the unique expression is determined for each category, and the information is stored in the display control data storage unit 17. The display control data stored in the display control data storage 17 is, for example, as shown in FIG. The display data generation unit 16 adds a tag for color classification to the HTML file with reference to the display control data in the display control data storage unit 17 under the control of the display control unit 18.
[0032]
The display as shown in FIG. 5 is performed on the display unit 19 based on the display data thus generated. In the example of FIG. 5, the entity of the category of “COMPANY” (company name) is displayed in “red” (indicated by an arrow A), and the entity of the category of “PERSON” (person name) is displayed in “blue” (arrow B). ), The entity of the category of “CURRENCY” (currency) is displayed in “purple” (indicated by arrow C), and the entity of the category of “DATE” (date and time) is “green” (arrow D) ). Buttons for each category are displayed at the top of the display screen (the main frame on the right side), so that the on / off of color classification for each category may be controlled.
[0033]
Note that the "tree display" button in the left frame of FIG. 5 is for expressing the hierarchy of the unique expression as described later in detail, and the "summary display" is for displaying the summary. . A summary may be automatically generated from the text. The “download” button is for downloading a document or text to a user terminal or the like.
[0034]
When the "tree display" button on the left side is operated in the state of FIG. 5, a tree of the unique expression is displayed as shown in FIG. In this tree structure, the top hierarchy (below the root; the root is not displayed) is the name of each category such as “COMPANY”. The next layer is an individual entity expression belonging to each category, for example, “Fuji Xerox” (trademark). The next level (last level) is the page number where the unique expression appears. In the example of FIG. 6, the page number does not appear because the document has only one page. However, if the document has a plurality of pages, the user can operate the unique expression “+” to change the corresponding page number to the next page. Appear as a hierarchy of
[0035]
The tree structure information can be generated from the entity expression information as shown in FIG. For example, page number information may be added to each entity.
[0036]
In the example of FIG. 6, the color classification target can be specified by operating each category name and each unique expression in the tree structure in the left frame. For example, a symbol for selectively enabling the color coding is embedded in the category name or each unique expression, and the symbol is sent to the CGI or the servlet on the web server side by an operation such as a click to enable the color coding. May be switched. Objects may be embedded in web pages.
[0037]
FIG. 7 shows a display example when a document has a plurality of pages. In this case, the page can be switched by clicking the page number in the page selection section of the lower right frame.
[0038]
Next, a detailed operation of the named entity extraction unit 14 will be described. In this example, a morphological analysis is performed on an input character string that is a text, and a named entity is extracted by applying a named entity extraction rule to the result.
[0039]
For example, consider an input example as shown in FIG. The dictionary used for the morphological analysis is, for example, as shown in FIG. In the morphological analysis, continuous words are cut out from the part of speech information using a connection availability table that defines the availability of words. Here, the description of the connection availability table is omitted. FIG. 10 shows the result of the morphological analysis. “/” Indicates a break between morphemes. Part of speech information obtained from morphological analysis is described in “<>”. The starting position and length of each morpheme are also described, but are omitted in the figure.
[0040]
The named entity extraction rules are as shown in FIG. By applying this rule, unique expression information as shown in FIG. 12 is obtained.
[0041]
In this embodiment, the unique expressions are displayed in different colors for each category, so that the user can grasp at a glance the general contents of the document and a part of the document, and read the document as necessary. In addition, since the appearing specific expressions are displayed in a hierarchy for each category, the appearing specific expressions can be grasped only by the tree display, and the items to be color-coded (nodes in the tree structure, individual categories and individual specific expressions) are displayed. ) Can be displayed in different colors by clicking or the like, which is very convenient. Also, pages are displayed in a hierarchy, and pages can be selected. If there is an item that you want to pay attention to, you can immediately display the page for that item, and you can color-code only the item of interest, so Arrive immediately and understand the content immediately.
[0042]
[Second embodiment]
Next, a second embodiment will be described. In this embodiment, a related document search function is added to the first embodiment.
[0043]
FIG. 13 shows a document processing apparatus according to the second embodiment, in which parts corresponding to those in FIG. 1 are denoted by corresponding reference numerals. In FIG. 13, the index generation unit 20 generates index data for related document search with respect to the text of each page in the text storage unit 13. The text for each page has a file name obtained by concatenating the document name and the suffix indicating the page as described above. The index is composed of, for example, the named entities extracted by the named entity extraction unit 14. The index name is associated with the page file. The index data is stored in the index storage unit 21. The search unit 22 receives the keyword from the user, performs a search by referring to the index storage unit 21, and passes the search result to the display control unit 18. As a result, the display data of the search result is changed as shown in FIG. It is generated by the display data generation unit 16 and displayed on the display unit 19. Search results are page-by-page. Therefore, when the corresponding page is selected, the display as shown in FIGS. 5 and 6 is performed. Search results are displayed on a page-by-page basis, so you can immediately grasp the target location.
[0044]
The other configuration is the same as that of the first embodiment, and the description is omitted.
[0045]
It should be noted that the present invention is not limited to the above-described embodiment, and various changes can be made without departing from the gist of the present invention. For example, what kind of entity expression is selected depends on the content of the document processing, and is not limited to the one described in this embodiment. In short, any words or phrases that are important in grasping the document may be used. The named entity extraction is not limited to the example of the above embodiment. Although color coding is performed for highlighting the unique expression and the like, various display attributes such as a blink attribute, an underline display, and the like can be used.
[0046]
【The invention's effect】
As described above, according to the present invention, important words and phrases in a document or a document segment can be immediately grasped by, for example, color coding for each category, so that the approximate document or document segment can be easily confirmed. If the search is performed in document segment units, the range to be confirmed can be narrowed.
[Brief description of the drawings]
FIG. 1 is a block diagram showing a configuration of a first embodiment of the present invention.
FIG. 2 is a diagram illustrating an example of an input text according to the embodiment.
FIG. 3 is a diagram illustrating an example of a named entity extraction result according to the embodiment.
FIG. 4 is a diagram illustrating an example of display control data according to the embodiment.
FIG. 5 is a diagram illustrating a display example of a document according to the above embodiment.
FIG. 6 is a diagram illustrating an example of a tree display according to the above embodiment.
FIG. 7 is a diagram illustrating an example of displaying a document of a plurality of pages according to the embodiment.
FIG. 8 is a diagram illustrating an operation of extracting a named entity in the above embodiment.
FIG. 9 is a diagram illustrating an operation of extracting a named entity according to the embodiment.
FIG. 10 is a diagram illustrating an operation of extracting a named entity in the embodiment.
FIG. 11 is a diagram illustrating an operation of extracting a named entity in the embodiment.
FIG. 12 is a diagram illustrating an operation of extracting a named entity according to the embodiment.
FIG. 13 is a block diagram showing a configuration of a second exemplary embodiment of the present invention.
FIG. 14 is a diagram illustrating a display example of a search result according to the second embodiment.
[Explanation of symbols]
11 Document storage unit 12 Text extraction unit 13 Text storage unit 14 Named expression extraction unit 15 Named expression storage unit 16 Display data generation unit 17 Display control data storage unit 18 Display control unit 19 Display unit 20 Index generation unit 21 Index storage unit 22 Search Department

Claims (12)

Means for extracting, from a document, a phrase having an arbitrary attribute included in a plurality of predetermined attribute groups;
Means for associating the extracted words with corresponding display attributes according to each attribute;
Document display means for displaying at least a part of the document and displaying the words included in at least a part of the document with corresponding display attributes. Means for dividing the document into document segments;
Means for extracting, from each of the document segments, a phrase having an arbitrary attribute included in a plurality of predetermined attribute groups;
A document segment display means for displaying at least a part of the document segment and displaying the words included in at least a part of the document segment with corresponding display attributes. 3. The document display device according to claim 2, wherein said document segment is a page constituting a document. Extracted phrase means for classifying and displaying the extracted phrases for each of the attributes,
Means for selecting an extracted phrase or an attribute of the extracted phrase based on a user operation,
The display means displays at least a part of the document and displays the selected phrase or the phrase having the selected attribute included in at least a part of the document with a corresponding display attribute. The document display device according to claim 1. Means for extracting, from a document, a phrase having an arbitrary attribute included in a plurality of predetermined attribute groups;
Term display means for classifying and displaying the extracted terms for each of the attributes,
Means for selecting an extracted phrase or an attribute of the extracted phrase based on a user operation;
Document display means for displaying at least a part of the document, and displaying the selected phrase or the phrase having the selected attribute included in at least a part of the document with a corresponding display attribute. A document display device comprising: Means for dividing the document into document segments;
Means for extracting, from each of the document segments, a phrase having an arbitrary attribute included in a plurality of predetermined attribute groups;
Term display means for classifying and displaying the extracted terms for each of the attributes,
Means for selecting an extracted phrase or an attribute of the extracted phrase based on a user operation;
A document segment display for displaying at least a part of the document segment and displaying the selected word or the word having the selected attribute included in at least a part of the document segment with a corresponding display attribute And a document display device. Means for storing a plurality of documents;
Means for dividing the document into document segments;
Means for extracting an index for each of said document segments;
Index storage means for storing an association between the document segment and the index;
Search means for performing a keyword search by referring to the index storage means in document segment units;
Means for displaying a search result by the search means. Means for extracting, from each of the document segments, a phrase having an arbitrary attribute included in a plurality of predetermined attribute groups;
8. The retrieval apparatus according to claim 7, further comprising: document segment display means for displaying at least a part of the document segment and displaying the words included in at least a part of the document segment with corresponding display attributes. Extracting from the document a phrase having any attribute included in the plurality of predetermined attribute groups;
Associating each extracted attribute with a corresponding display attribute according to each attribute;
A display step of displaying at least a part of the document and displaying the words included in at least a part of the document with corresponding display attributes. Extracting from the document a phrase having any attribute included in the plurality of predetermined attribute groups;
Associating each extracted attribute with a corresponding display attribute according to each attribute;
Displaying at least a part of the document and displaying the words included in at least a part of the document with corresponding display attributes by a computer. Computer program. Extracting from the document a phrase having any attribute included in the plurality of predetermined attribute groups;
A phrase display step of classifying and displaying the extracted phrases for each attribute,
Selecting an extracted phrase or an attribute of the extracted phrase based on a user operation;
A document display step of displaying at least a part of the document, and displaying the selected phrase or the phrase having the selected attribute included in at least a part of the document with a corresponding display attribute. A document display method comprising: Extracting from the document a phrase having any attribute included in the plurality of predetermined attribute groups;
A phrase display step of classifying and displaying the extracted phrases for each attribute,
Selecting an extracted phrase or an attribute of the extracted phrase based on a user operation;
A document display step of displaying at least a part of the document, and displaying the selected phrase or the phrase having the selected attribute included in at least a part of the document with a corresponding display attribute. A computer program for displaying a document, which is used to cause a computer to execute the following.

Images