JPH1083404A - Full text retrieving system and recording medium recorded with program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To easily grasp where a character string corresponding to a keyword exists in a document for an operator. SOLUTION: This system comprises an image file 6 which stores document image data of each document, a database 14 which stores each character string included in each document and a character position table which stores the relation ship between each character and the character position in every document. In this case, the document designating information of a document that includes a character string corresponding to an inputted keyword is retrieved from the database 14, document image data corresponding to the retrieved document designating information is read out of an image file 6, and the read document image data and the retrieved document designating information are outputted for displaying. Furthermore, the position of each character constituting character string corresponding to a keyword in displayed document image data is read from the character position table, and the image data of a character position that is read in the displayed document image data is emphasis- displayed.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an information retrieval system for retrieving a document in which necessary information is described from a large number of documents. The present invention relates to a possible full-text search system and a recording medium on which the program is recorded.

[0002]

2. Description of the Related Art When a new document (document) is registered in a database of a general information retrieval system, a plurality of keywords included in the document are registered in the database. When a document (document) in which necessary information is described is searched using this information search system, a database is searched with a keyword related to the necessary information, and the document number and the document of the document in which the keyword is registered are searched. Name, issuer,
Document specification information that specifies the relevant document, such as the author, issue date, and holding place, is output as a search result.

Some of such information retrieval systems include:
2. Description of the Related Art A full-text search system has been put into practical use in which all characters, words, phrases, and the like included in each document are registered as keywords in a database, and information contained in each document can be searched without omission.

Further, a full-text search system has been put to practical use in which text data of each document registered in a database is also stored and held in a document file. In such a full-text search system, an operator searches a database using a keyword related to necessary information, and obtains document designation information for specifying a document including a character string of a corresponding keyword.

Then, text data of the document specified by the document specification information is called from the document file, for example, on a display screen of a CRT display device. Then, the operator determines whether or not the document includes information required by the operator.

[0006]

However, the above-described full-text search system has the following problems to be improved. Generally, in a full-text search system, all characters, words, and phrases included in a document registered in a database are searched as keywords. Therefore, characters, words, and phrases that are not directly related to the essential contents of the relevant document can also be searched.

Therefore, a large number of documents are retrieved by one retrieval operation. It becomes very complicated to check where the corresponding keyword exists in each of the text data of the large number of documents and determine whether or not the corresponding document has the desired content.

In particular, when the document is divided into paragraphs as in a normal paper format, it is relatively easy to grasp the contents. However, in general, text data does not include ruled lines, tables, and figures. Therefore, in the text data of various forms such as invoices, receipts, and slips, the position of each character on the form is completely ignored, and all the characters and words are stored together. Therefore, as far as this text data is viewed, it is impossible to grasp at which position on the form each character or each word is located.

As a result, even if a character string (word) corresponding to the keyword is found on the text data, it cannot be determined what meaning the keyword has on the document. For example, if the keyword is “total”, the position of the total determines, for example, the total sales or the total payment.

Therefore, even if a document is specified by a keyword, there is a problem that it cannot be easily determined whether or not the document is required by the user. The present invention has been made in view of such circumstances, and by storing and holding image data of each document registered in a database as an image file, document designation information as a search result and a corresponding document The image data can be displayed at the same time, and even if the document is in a form, the operator can easily grasp where the character string corresponding to the keyword exists in the document, and can obtain the required document. It is an object of the present invention to provide a full-text search system capable of easily determining whether or not the search operation has been performed and greatly improving the search operation efficiency, and a recording medium storing the program.

[0011]

According to a first aspect of the present invention, there is provided a full-text search system, comprising: an image file for storing document image data of a corresponding document for each document; A database for storing character strings corresponding to the document designation information, a character position table for storing a relationship between each character included in the document and a position of the character in the document image data for each document; Search means for searching a database for document specification information of a document including a character string corresponding to a keyword; document image data reading means for reading document image data corresponding to the searched document specification information from an image file; Search result display means for displaying and outputting the document image data and the searched document designation information, and a keyword in the displayed document image data The position of each character constituting the corresponding character string is read from the character position table, and a highlight unit for highlighting the image data of the read character position in the document image data displayed.

According to a second aspect of the present invention, when the retrieval means in the full-text retrieval system of the first aspect retrieves a plurality of pieces of document designation information, the document image data reading means reads the document image data corresponding to each document designation information, The search result display means displays and outputs a plurality of read document image data and a plurality of searched document designation information on one display screen.

[0013] In the full-text search system according to claim 3,
An image file that stores the document image data of the corresponding document for each document, a database that stores each character string included in each document corresponding to the document specification information, and a file that stores each character included in the corresponding document for each document. A character position table for storing a relationship between the position of the character in the document image data, a search means for searching a document for document specification information of a document including a character string corresponding to the input keyword, and a searched document specification Document image data reading means for reading document image data corresponding to information from an image file; and reading and reading the position of each character constituting a character string corresponding to a keyword in the read document image data from a character position table. Extraction of area image data of a predetermined area including each character position from the read document image data It comprises a stage, and a search result display means for displaying outputs the extracted region image data and the retrieved document specifying information was.

According to a fourth aspect of the present invention, in the full-text search system according to the third aspect, the image data at the character position read out in the area image data displayed by the search result display means is highlighted. Highlighting means is added.

[0015] In the full text search system of claim 5,
An image file that stores the document image data of the corresponding document for each document, a database that stores each character string included in each document corresponding to the document specification information, and a file that stores each character included in the corresponding document for each document. A character position table for storing the relationship between the position of the character in the document image data, input means for inputting a keyword and a designated area in the document image data, and a character corresponding to the keyword input via the input means Searching means for searching a document for document specification information of a document including a column; document image data reading means for reading document image data corresponding to the searched document specification information from an image file; Character position reading means for reading the position of each character constituting the character string corresponding to the keyword from the character position table; When the character position was located has been designated region input, and a retrieval result displaying means for displaying the output document image data and the retrieved document designation information was read out.

According to a sixth aspect of the present invention, there is provided the full-text search system according to the fifth aspect, further comprising image data of a read character position in a designated area in the document image data displayed by the search result display means. Is highlighted.

In the full-text search system according to the present invention, the document image data of the document registered in the database is separately stored in the image file. When the operator inputs a keyword, the database is searched for document designation information of a document including a character string corresponding to the keyword, and is displayed and output together with the document image data of the corresponding document.

Further, of the displayed document image data, the image data at the character string position corresponding to the keyword is highlighted. Therefore, the operator can see at a glance where the character string corresponding to the keyword is located in the displayed document image data, and determine whether the corresponding document is a document required by the operator. And the search operation efficiency is improved.

According to the second aspect of the present invention, when a plurality of document designation information are checked as a result of searching the database, the retrieved plurality of document designation information and the corresponding plurality of document image data are stored in one file. Displayed and output on the display screen.

Therefore, the operator can compare the displayed documents, and can more efficiently determine the documents required by the operator. In addition, by simultaneously displaying a plurality of document image data, the time required to view all search results can be reduced.

According to the third aspect of the present invention, only a predetermined area including a character position corresponding to a keyword in the document image data of the searched document designation information is extracted and displayed and output together with the document designation information.

As a result, unnecessary portions in the document image data are not displayed, so that the operator can easily find a character corresponding to the keyword from the displayed region image data. In addition, since the keyword portion is displayed in a large size, characters can be easily read.

According to the fourth aspect, since the image data of the character string corresponding to the keyword in the displayed area image data is further highlighted, the operator looks at the character string corresponding to the keyword at a glance. You can check.

In the invention according to claim 5, when the operator specifies an area in the document image data together with the keyword, a character string corresponding to the keyword in the document specification information retrieved from the database exists in the previous specified area. Only when this is done, the retrieved document specification information and the corresponding document image data are displayed and output.

That is, for example, when the position of a keyword to be searched in a document can be specified to some extent, such as in a form, the number of document specification information to be searched is limited by specifying an area. Therefore, the selection range when selecting a required document is narrowed, and as a result, the search work efficiency is improved.

In the sixth aspect, the image data of the character string portion corresponding to the keyword in the displayed document image data is further highlighted, so that the operator looks at the character string corresponding to the keyword at a glance. You can check.

According to a seventh aspect of the present invention, there is provided a computer readable recording medium having recorded thereon a program for storing document image data of a corresponding document in a first storage means for each document, The function of storing each character string corresponding to the document designation information in the second storage means, and the relationship between each character included in the corresponding document and the position of the corresponding character in the document image data for each of the documents. And a function of searching the second storage unit for the document designation information of the document including the character string corresponding to the input keyword, and a function corresponding to the document designation information obtained by the search. A function of reading the document image data from the first storage unit, a function of displaying the read document image data and the retrieved document designation information, and a function of displaying the displayed document image data. Reading the position of each character constituting the character string corresponding to the keyword from the third storage means, and highlighting the image data at the read sentence position in the displayed document image data. The program that realizes the function is recorded.

[0028] In a computer readable recording medium on which the program according to claim 8 is recorded, a mechanism for storing document image data of the document in the first storage means for each document; A function of storing the character string in the second storage means in correspondence with the document designation information, and a third relation between each character included in the corresponding document and a position of the corresponding character in the document image data for each of the documents. A function of storing the document in the storage unit, a function of causing the second storage unit to search for the document designation information of the document including the character string corresponding to the input keyword, and a function of the document corresponding to the searched document designation information. The function of reading image data from the first storage means, and the position of each character constituting the character string corresponding to the keyword in the read document image data is stored in the third storage means. A function of reading and extracting area image data of a predetermined area including the read character positions from the read document image data, and displaying the extracted area image data and the searched document designation image. A program for realizing the output function is recorded.

Further, in a computer-readable recording medium on which the program according to claim 9 is recorded, a mechanism for storing document image data of the document in the first storage means for each document; The function of storing each character string corresponding to the document designation information in the second storage means, and the relationship between each character included in the corresponding document and the position of the corresponding character in the document image data for each of the documents. 3, a function for receiving an input of a keyword and a designated area in the document image data, and a document designation information of a document including the character string corresponding to the inputted keyword in a second unit. A function to search from storage means,
A function for reading the document image data corresponding to the searched document designation information from the first storage means, and a case where the read document image data is located in the input designated area, A program for realizing a function of displaying and outputting the read document image data and the searched document designation information.

[0030]

DESCRIPTION OF THE PREFERRED EMBODIMENTS One embodiment of the present invention will be described below with reference to the drawings. FIG. 1 is a block diagram showing the overall configuration of the full-text search system according to the embodiment. The administrator of the full-text search system inserts the form 1 on which the document 1a having the print format shown in FIG. The image reading device 2 reads the image data of the document 1a described in the format shown in FIG. 2 of the inserted form 1, and sends it to the character recognition device 3 and the image management unit 4 as document image data. In addition, the administrator inserts the form 1 into the image reading device 2 and inputs a document number as document specifying information for specifying the form 1 into the document number input unit 5.

The image management unit 4 writes the document image data of the document 1 a of the form 1 input from the image reading device 2 into the image file 6 with the document number input from the document number input unit 5.

The character recognizing device 3 reads each character from the document image data of the document 1a of the form 1 input from the image reading device 2, converts it to a character code, and sends it to the text management unit 11 as text data. Naturally, the text data does not include information on ruled lines, tables, and description positions described in the form 1. The text management unit 11 writes the input text data into the document file 12 with the document number input from the document number input unit 5.

The character position detecting section 7 determines that each character included in the text data input from the character
a in the document image data of document a, and a character position table 8a shown in FIG.
Create Each of the created character position tables 8a is written to the character position file 8 with the document number input from the document number input unit 5.

More specifically, as shown in FIG.
Are represented by (x, y) coordinates on the document image data. The correspondence setting unit 9 creates the management table 10a shown in FIG. 4 based on each information from the image management unit 4, the character position detection unit 7, and the document number input unit 5. In this management table 10a,
For each document number that specifies each document 1a, the number of the document image data of the corresponding document 1a and the number of the character position table 8a are registered. The correspondence setting unit 9 writes the created management table 10a into the management file 10.

The database registration unit 13 registers each character and each word included in the text data of each document 1a stored in the document file 12 in the full-text search database 14 having the format shown in FIG.

More specifically, as shown in FIG. 5, for each character such as hiragana, katakana, kanji, and alphanumeric characters,
For each two-character word composed of two characters, for each three-character word composed of three characters, and for each four-character word composed of four characters, the corresponding character string is included in the text data of each document. Is registered as information as to whether or not it exists.
If it exists, [1] is set, and if it does not exist, [0] is set.

For example, a two-character word consisting of two characters of [today] is a document 1a of document number 1 and a document 1 of document number 3
a. Also, [Invoice] 3
The three-letter word composed of letters indicates that the word is described only in the document 1a of document number 2.

The search input unit 15 is composed of, for example, a keyboard, and the operator inputs a keyword. Further, the operator inputs a designated area as needed. The database search unit 16 searches the full-text search database 14 using the keyword input via the search input unit 15, and sets the document designation information set to [1] including a character string (word) corresponding to the keyword. The document number as is read. The database search unit 16 sends one or more document numbers searched in the full-text search database 14 to the next search result editing unit 17.

The search result editing unit 17 reads the number of the document image data of the document 1a corresponding to the searched document number and the number of the character position table 8a from the management table 10a of the management file 10. Then, the document image data of the corresponding number is read from the image file 6. Also, from the character position table 8a of the corresponding number in the character position file 8,
The (x, y) coordinates indicating the position of the character string (word) corresponding to the input keyword in the corresponding document image data are read.

Then, the search result editing unit 17 displays and outputs the searched document number, the keyword, and the document image data of the corresponding document number on the display screen 18a of the display unit 18 as a search result.

Further, the search result editing unit 17 includes the display unit 1
8, image data of a character string (word) corresponding to the read keyword in the document image data displayed on the display screen 18a is emphasized in a color different from other parts such as [yellow] or [green]. indicate.

The search result is displayed on the display screen 18 of the display unit 18.
A plurality of [display modes] are prepared for the display method to be displayed in a, and the operator can arbitrarily select this [display mode].

Specifically, as [display mode], FIG.
A total of four [reduced display mode] shown in FIG. 0, [highlighted display mode] shown in FIG. 11, [region extraction display mode] shown in FIG. 12, and [region designation display mode] not shown are prepared.

Next, the specific operations of the database search unit 16 and the search and edit unit 17 when each [display mode] is designated will be described with reference to FIGS. FIG. 6 is a flowchart showing the search processing operation when the display mode is set to [reduced display mode].

When a keyword is input from the search input unit 15 at P1, the database search / search unit 16 reads one or more documents 1a containing the character string (word) of the keyword from the full-text search database 14. The document number is searched (P2). Next, the document image data of the document 1a corresponding to the document number is read from the image file 6 (P3). The read document image data is reduced to, for example, about ８ of the area of the display screen 18a of the display unit 18 (P4).

As shown in FIG. 10 (a), the reduced document image data 19, the keyword 20, and the document number 21 are used as search results to display the display screen 18 of the display unit 18.
Display output to a.

At P6, if there is a document number that has been searched and has not been displayed yet, the process returns to P3, and display processing of the search result for the document number is started. Then, for example, when two document numbers are searched, FIG.
As shown in FIG. 5, two reduced document image data 19, one keyword 20, and two document numbers 21 are displayed on the display screen 18a.

As described above, the reduced document image data 19 of the document 1 a corresponding to the document number 21 is displayed on the display screen 18 a together with the searched document number 21. The content of the document 1a including the corresponding character string (word) can be confirmed on the display screen 18a. Therefore, the operator can easily determine in a short time whether or not the displayed document 1a is the document 1a required by the operator. As a result, search operation efficiency is greatly improved.

FIG. 7 is a flowchart showing a search processing operation when the display mode is set to [highlighted display mode]. Q
When a keyword is input from the search input unit 15 in step 1,
The database search / search unit 16 searches the full-text search database 14 for one or more document numbers of the document 1a including the character string (word) of the corresponding keyword (Q
2). Next, the document image data of the document 1a corresponding to the document number is read from the image file 6 (Q3).

Next, at Q4, the character position of the character string (word) corresponding to the keyword is indicated from the character position table 8a of the corresponding document number of the character position file 8 (x, y).
Read the coordinates. Then, at Q5, the previously read document image data 22, keyword 20, and document number 21 are displayed on the display screen 18a of the display unit 18 as shown in FIG.

Further, the image data 23 at the read character position in the displayed document image data 19 is highlighted in a different color (Q6). In Q7, if there is a document number that has been searched and has not been displayed yet, the process returns to Q3, and display processing of the search result for the document number is started.

Then, for example, when two document numbers are retrieved, as shown in FIG.
In a, two document image data 22, one keyword 20, and two document numbers 21 are displayed.

As described above, the document image data 22 of the document 1 a corresponding to the document number 21 is displayed on the display screen 18 a together with the retrieved document number 21, and corresponds to the keyword 20 in the document image data 22. The character string (word) image data 23 is highlighted.

Therefore, the operator can check the contents of the document 1a including the character string (word) corresponding to the keyword 20 on the display screen 18a, and immediately determine the character position corresponding to the keyword in the document 1a. You can check.
Therefore, the operator can easily and quickly determine whether the displayed document 1a is the document 1a required by the operator. As a result, search operation efficiency is greatly improved.

Further, in addition to the keyword 20 and the document number 21, the display screen 18 a of the display 18
The document image data 22 of the document 1a is also displayed at the same time. Therefore, even if each character string (word) described in the document 1a is separated by a ruled line or a table as in the form 1, etc., the corresponding document 1a of the character string (word) corresponding to the keyword You can see at a glance the location described above.

Therefore, it is possible to grasp in a short time whether or not the character string corresponding to the keyword described in the document 1a relates to the information required by the user. FIG. 8 is a flowchart showing a search processing operation when the display mode is set to [region extraction display mode].

When a keyword is input from the search input unit 15 at R1, the database search / search unit 16 reads one or more documents 1a containing the character string (word) of the keyword from the full-text search database 14. The document number is searched (R2). Next, the document image data of the document 1a corresponding to the document number is read from the image file 6 (R3).

Next, at R4, the character position of the character string (word) corresponding to the keyword is indicated from the character position table 8a of the corresponding document number of the character position file 8 (x, y).
Read the coordinates. Then, in R5, a predetermined area surrounding the read character position is set, and image data of the set predetermined area is extracted as area image data from the read document image data (R6). Then, at R7, FIG.
As shown in (a), the region image data 2 extracted this time
4. Display the keyword 20 and the document number 21 on the display screen 18a of the display unit 18.

At R8, if there is a document number that has been searched and has not been displayed yet, the process returns to R3, and display processing of the search result for the document number is started. Then, for example, when two document numbers are searched, the display screen 18
In a, two area image data 24, one keyword 20 and two document numbers 21 are displayed.

As described above, only the area image data 24 of the predetermined area surrounding the character string (word) corresponding to the keyword is displayed on the display screen 18a, and the character string (word) corresponding to the keyword is not described. Since the partial image data is not displayed, the operator can easily confirm the character string (word) corresponding to the keyword from the displayed area image data 24. As a result, it is possible to quickly determine whether the document 1a is necessary.

Further, as shown in FIG. 12B, a character string corresponding to a keyword in the area image data 24 displayed on the display screen 18a by using the method adopted in the above-mentioned [highlighted display mode]. (Word) image data 23 can be highlighted.

In this case, since the operator can immediately grasp the character string (word) corresponding to the keyword in the area image data 24, the above-described search work efficiency can be further improved.

FIG. 9 is a flowchart showing a search processing operation when the display mode is set to [area designation display mode]. In S1, a keyword is input from the search input unit 15, and further in S2, a designated area in the document 1a is input from the search input unit 15, and in S3, the database search / search unit 16 Then, one or more document numbers of the document 1a including the character string (word) of the corresponding keyword are searched (S3).

The specified area is, for example, information indicating the approximate position of the upper half, the lower half, or the right half, the left half, or the center of the document 1a, and the specific area such as the first line and the last line. There may be information indicating a position.

Next, in S4, the document image data of the document 1a corresponding to the document number is read from the image file 6. Next, in S5, the (x, y) coordinates indicating the character position of the character string (word) corresponding to the keyword are read from the character position table 8a of the corresponding document number in the character position file 8.

Then, in S6, the read character position is compared with the previously input designated area, and if the read character position falls within the designated area (S7), the flow advances to S8 to read first. The document image data, the keyword, and the document number are displayed on the display screen 18a of the display unit 18 (S8).

Next, in step S9, the image data of the character string (word) corresponding to the keyword in the document image data displayed on the display screen 8a is obtained by using the method adopted in the above-mentioned [highlighted display mode]. Is highlighted.

At S10, if there is a document number that has been searched and has not been displayed yet, the process returns to S4, and display processing of the search result for the document number is started. Then, for example, when two document numbers are searched, the display screen 1
8a includes two document image data and one keyword 2
0 and two document numbers 21 are displayed.

For example, when the position of a keyword to be searched in a document can be specified to some extent, such as in the form 1, the number of document numbers to be searched is limited by specifying an area. The selection range when selecting the required document is narrowed, and as a result, the search work efficiency is improved.

In the full-text search system configured as described above, as a display result displayed on the display screen 18a of the display unit 18, in addition to the ordinary searched document number 21 and the keyword 20, the corresponding document number The document image data 19 and 22 of the document 1a to be displayed are displayed at the same time.

Therefore, as an operator, the contents of the document 1a including the character string (word) corresponding to the keyword 20 can be confirmed on the display screen 18a, and the operator needs to display the displayed document 1a. It is possible to easily determine in a short time whether or not the document is a document 1a. As a result, search operation efficiency is greatly improved.

Further, as the [display mode] for displaying and outputting the search result on the display screen 18a of the display unit 18, the [reduced display mode] shown in FIG. 10, the [highlighted display mode] shown in FIG. 11, and the [display mode] shown in FIG. An area extraction display mode] and an unillustrated [area designation display mode] are prepared, and the operator selects an optimum [display mode] from these as necessary.
By performing the search operation by designating, the search process can be efficiently performed by the best method suitable for the user.

The present invention is not limited to the above embodiment. For example, the functions of each unit (image management unit, character position detection unit, correspondence setting unit, text management unit, database registration unit, search result editing unit, database search unit) constituting the full text search system shown in FIG. The CD-ROM is written in advance on a recording medium such as a CD-ROM, and the CD-ROM is mounted on a computer equipped with a CD-ROM drive. Can be realized.

As a recording medium, the above-mentioned CD-RO
In addition to M, a magnetic tape, DVD-ROM, floppy disk, MO, MD, CD-R, memory card, etc. may be used.

[0075]

As described above, in the recording medium storing the full-text search system and the program of the present invention, the image data of each document registered in the database is stored and held as an image file. Then, the document designation information as the search result and the image data of the corresponding document are simultaneously displayed on the display screen.

Therefore, even if the document is in a form, it is easy for the operator to know where in the document the character string corresponding to the keyword is located. It is possible to easily determine whether or not a document is to be searched for, and the search work efficiency can be greatly improved.

Further, image data of a character string corresponding to a keyword in the document image data displayed on the display screen is highlighted in, for example, a different color. Therefore, the operator can confirm the character string corresponding to the keyword at a glance at a glance.

Further, as a result, a search area for a keyword in a document can be designated, the number of document numbers to be searched is limited, and a selection range for selecting a required document for an operator is narrowed. As a result, the search operation efficiency is improved.

Further, only the image data of a predetermined area including the character string corresponding to the keyword among the document image data of the searched document number is displayed. Therefore, a character string (word) corresponding to the keyword on the displayed image data can be confirmed for the operator in a shorter time, and the search work efficiency can be greatly improved.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a schematic configuration of a full-text search system according to an embodiment of the present invention.

FIG. 2 is a diagram showing a form in which a document registered in the full-text search system is described.

FIG. 3 is a diagram showing storage contents of a character position table formed in the full-text search system.

FIG. 4 is a diagram showing storage contents of a management table formed in the full-text search system.

FIG. 5 is a diagram showing storage contents of a full-text search database formed in the full-text search system.

FIG. 6 is a flowchart showing search processing when the reduced display mode of the full-text search system is set.

FIG. 7 is a flowchart showing a search process when the highlight mode is set in the full-text search system.

FIG. 8 is a flowchart showing search processing when the area extraction display mode is set in the full-text search system.

FIG. 9 is a flowchart showing a search process when an area designation display mode is set in the full-text search system.

FIG. 10 is a view showing search results displayed when the reduced text display mode of the full-text search system is set.

FIG. 11 is a view showing a displayed search result when a highlight mode is set in the full-text search system.

FIG. 12 is a diagram showing search results displayed when the area extraction display mode is set in the full-text search system.

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1 ... Form, 1a ... Document, 2 ... Image reading device, 3 ... Character recognition device, 5 ... Document number input part, 6 ... Image file,
8 ... character position file, 8a ... character position table, 10
... management file, 10a ... management table, 12 ... document file, 14 ... full text search database, 15 ... search input section, 16 ... database search section, 17 ... search result editing section, 18 ... display section, 18a ... display screen, 19: Reduced document image data, 20: Keyword, 21: Document number, 22: Document image data, 24: Area image data

 ──────────────────────────────────────────────────続 き Continued on the front page (72) Yoshiaki Suzuki, Inventor: 1 Toshiba-cho, Fuchu-shi, Tokyo, Japan Inside the Toshiba Fuchu Plant, Inc. 72) Inventor Tsunatsuki Hamamoto 1 Toshiba-cho, Fuchu-shi, Tokyo Tokyo, Japan Inside the Fuchu Plant, Toshiba Corporation (72) Inventor Naoya 1 Toshiba-cho, Fuchu-shi, Tokyo Inside the Fuchu Plant, Toshiba Corporation (72) Inventor Tomohide Sugaya Tokyo 1 Toshiba-cho, Fuchu-shi Toshiba Corporation Fuchu factory

Claims (9)

[Claims] An image file storing document image data of the document for each document; a database storing character strings included in each document in correspondence with document designation information; A character position table that stores a relationship between each character included in the document and a position of the character in the document image data; and a search that searches the database for document specification information of a document including a character string corresponding to an input keyword. Means, document image data reading means for reading document image data corresponding to the retrieved document designation information from the image file, and retrieval for displaying and outputting the read document image data and the retrieved document designation information. A result display unit, and a position of each character constituting a character string corresponding to the keyword in the displayed document image data. A full-text search system comprising: a highlighting unit that reads from the character position table and highlights the image data of the read character position in the displayed document image data. 2. When the search means searches for a plurality of document designation information, the document image data reading means reads document image data corresponding to each document designation information, and the search result display means reads the document image data. 2. The full-text search system according to claim 1, wherein a plurality of document image data and the searched plurality of document designation information are displayed and output on one display screen. 3. An image file for storing document image data of a corresponding document for each document; a database for storing character strings included in each of the documents in correspondence with document designation information; A character position table that stores a relationship between each character included in the document and a position of the character in the document image data; and a search that searches the database for document specification information of a document including a character string corresponding to an input keyword. Means, document image data reading means for reading the document image data corresponding to the searched document designation information from the image file, and a character string corresponding to the keyword in the read document image data. The position of a character is read from the character position table, and area image data of a predetermined area including the read character positions is read out from the character position table. Issued and the area image data extracting means for extracting from the document image data, full-text search system and a search result display means for displaying output the extracted region image data and the retrieved document specifying information was. 4. The full-text search system according to claim 3, further comprising highlighting means for highlighting the image data at the read character position in the area image data displayed by the search result display means. 5. An image file for storing document image data of a corresponding document for each document; a database for storing character strings included in each of the documents in correspondence with document designation information; A character position table for storing a relationship between each character included in the document and a position of the character in the document image data; input means for inputting a keyword and a designated area in the document image data; Search means for searching the database for document designation information of a document including a character string corresponding to the keyword input by the user, and document image data for reading document image data corresponding to the searched document designation information from the image file Reading means; and positions of respective characters constituting a character string corresponding to the keyword in the read document image data. Character position reading means for reading out from the character position table, when the read character position is located in the input designated area, the read document image data and the searched document designation information A full-text search system including a search result display unit for displaying and outputting. 6. The full text according to claim 5, further comprising highlighting means for highlighting image data at the read character position in the designated area in the document image data displayed by the search result display means. Search system. 7. A function for storing document image data of a corresponding document in a first storage means for each document, and a second storage means for storing each character string included in each document in correspondence with document designation information. A function of storing the relationship between each character included in the document and the position of the character in the document image data in the third storage unit for each document, and a function corresponding to the input keyword. A function of retrieving document specification information of a document including the character string from the second storage means, a function of reading the document image data corresponding to the detected document specification information from the first storage means, A function of displaying and outputting the read document image data and the retrieved document designation information; and a third position of each character constituting the character string corresponding to the keyword in the displayed document image data.
A computer-readable recording medium which stores a program read from the storage unit of claim 1 and realizing a function of highlighting the read image data at the sentence position in the displayed document image data. 8. A storage device for storing document image data of a corresponding document in a first storage device for each document, and a second storage device for storing each character string included in each document in correspondence with document designation information. A function of storing the relationship between each character included in the document and the position of the character in the document image data for each document in the third storage unit, and a function corresponding to the input keyword. A function for causing the second storage unit to search for the document designation information of the document including the character string; a function for reading out the document image data corresponding to the searched document designation information from the first storage unit; The position of each character constituting the character string corresponding to the keyword in the read document image data is read from the third storage unit, and the area image data of the predetermined area including the read character position is stored in the previous storage unit. A function of extracting from the read document image data, the extracted region image data and the retrieved document specified image recording a program for realizing a function of displaying and outputting the computer-readable recording medium. 9. A storage device for storing document image data of a corresponding document in a first storage device for each document, and a second storage device for storing each character string included in each document in correspondence with document designation information. A function of storing, in a third storage means, a relationship between each character included in the document and a position of the character in the document image data for each document, a keyword and the document image data A function of receiving an input of a designated area in the document, a function of searching the second storage means for document specification information of a document including the character string corresponding to the input keyword, and a function of responding to the searched document specification information. A function of reading the document image data from the first storage means, and reading the document image data when the read document image data is located in the input designated area. And a computer-readable recording medium recording a program for realizing a function of displaying and outputting the searched document designation information.