JP2001034617A - Device and method for information analysis support and storage medium - Google Patents

Abstract

PROBLEM TO BE SOLVED: To enable a user to extract the information from the text character strings and to sort the texts in a common framework at need and then to obtain a processing result including a combination of the said extracting and sorting jobs. SOLUTION: A document record holding means 3 can store plural document records including the text character strings and the information on the attributes of these character strings. Each of plural information extraction means 8 can extract a character string from the text character strings to satisfy a prescribed specific condition. An information extraction instruction means 9 can select a means 8 that is applied to the designated text character strings included in one or more document records which are selected by a document record selection means 7. An attribute addition means 10 adds the character string extracted by the selected means 8 to a document record as the information on the attribute corresponding to the text character string of the destination of extraction.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

[0001] The present invention relates to an information analysis support device, an information analysis support method, and a storage medium.

[0002]

2. Description of the Related Art Conventionally, various information analysis support devices have been developed and used.

[0003] Among them, there is a function of extracting information (for example, proper nouns, numerical values, dates, etc.) contained in a text character string by using a natural language processing technique or a character string pattern matching technique. There is something.

[0004] Further, the relationship between a plurality of text strings,
Expressed by numerical values based on some similarity calculation method,
In some cases, similar text strings are classified as belonging to the same category.

[0005]

However, since the above-mentioned prior art aims to show the result of processing by a specific function to a user, the function can be switched and used as needed. Combine those results,
It is not configured to produce deeper analysis results, and is not convenient enough.

An object of the present invention is to extract information from a text string and classify texts from a common framework.
The processing can be performed as needed, so that the user can obtain a processing result obtained by combining them.

[0007]

According to a first aspect of the present invention, there is provided a document record holding means for storing a plurality of document records each including a text character string and information on the attribute of the text character string, and the document record holding means. And a document record selecting means for selecting one or more document records from the document records stored in the document record, and selecting an attribute as a reference for sorting the selected one or more document records. A sort-criterion attribute specifying unit that performs sorting on one or more document records selected by the document record selecting unit in accordance with the property of the information of the attribute specified by the sort-criterion attribute specifying unit. Sorting means, at least one kind of information extracting means for extracting a character string satisfying a predetermined specific condition from the text character string, Information extraction instructing means for enabling selection of the information extracting means to be applied to a specified text string included in one or more document records selected by the document record selecting means, and the selected information An information analysis support apparatus comprising: an attribute adding unit that adds a character string extracted by an extracting unit to the document record as information of the attribute corresponding to the text character string of the extraction destination.

Therefore, when there are a plurality of document records, a menu for extracting a character string satisfying a predetermined specific condition is selected and applied to a text character string included in the plurality of document records. Information extraction is performed, and the result can be used in the same way as the attribute information included in the document record from the beginning, and related document records can be organized.

According to a second aspect of the present invention, there is provided a document record holding means for storing a plurality of document records each comprising a text character string and attribute information of the text character string, and a document stored in the document record holding means. 1 from record
Document record selecting means for enabling selection of one or more document records; sort reference attribute designating means for enabling selection of an attribute used as a reference for sorting the selected one or more document records; Document record sorting means for sorting one or more document records selected by the document record selecting means in accordance with the property of the information of the attribute designated by the sort criterion attribute designating means;
One or more types of text classifying means for classifying the plurality of text character strings based on a predetermined method; and a specified text character string included in one or more document records selected by the document record selecting means. A classification instructing means for enabling selection of the text classification means to be applied to the document record as information of the attribute corresponding to the classified text character string. This is an information analysis support device including an attribute adding means for adding.

Therefore, when there are a plurality of document records, a menu for classifying text strings contained therein based on a predetermined method is selected and applied according to the purpose, and classification is performed. It can be used in the same way as the attribute information included in the document record from the beginning, and classification can be performed by combining a plurality of classification results. Further, the relation with other attributes is easily seen, and the efficiency of the information analysis work is improved.

According to a third aspect of the present invention, a character string satisfying a predetermined condition is extracted from the text character string included in a document record including a text character string and attribute information of the text character string. Selecting from the one or more types of menus that can be applied to a specified text string contained in one or more of the document records selected from the plurality of document records; and An information analysis including a step of extracting the character string according to a selected menu; and a step of adding the extracted character string to the document record as information of the attribute corresponding to the text string of the extraction destination. It is a support method.

Therefore, when there are a plurality of document records, a menu for extracting a character string satisfying a predetermined specific condition is selected and applied to a text character string included in the plurality of document records. Information extraction is performed, and the result can be used in the same way as the attribute information included in the document record from the beginning, and related document records can be organized.

According to a fourth aspect of the present invention, it is possible to classify the text character string included in a document record including a text character string and attribute information of the text character string based on a predetermined method. One selected from a plurality of document records from one or more menus
Selecting one to apply to a specified text string contained in one or more of the document records; classifying the text string according to the selected menu; And adding to the document record as information of the attribute corresponding to the text string thus obtained.

Therefore, when there are a plurality of document records, a menu for classifying text strings included in the document records based on a predetermined method is selected and applied according to the purpose, and classification is performed. It can be used in the same way as the attribute information included in the document record from the beginning, and classification can be performed by combining a plurality of classification results. Further, the relation with other attributes is easily seen, and the efficiency of the information analysis work is improved.

According to a fifth aspect of the present invention, there is provided a step of receiving a selection of one or more document records from a plurality of document records each including a text character string and attribute information of the text character string, and the selection. Receiving a selection of an attribute as a reference for sorting one or more of the selected document records, and providing the one or more document records selected in the document record selection receiving step in the attribute selection receiving step. A step of performing sorting according to the nature of the information of the attribute whose selection has been accepted, and one or more types capable of extracting a character string satisfying a predetermined specific condition from the text character string included in the document record From the menu of the specified text string included in one or more of the document records selected from a plurality of document records Selecting an application to be applied; extracting the character string according to the selected menu; and extracting the extracted character string into the document record as information of the attribute corresponding to the extracted text character string. A computer-readable storage medium storing a program for causing a computer to execute the adding step.

Therefore, when there are a plurality of document records, a menu for extracting a character string satisfying a predetermined specific condition is selected and applied to a text character string included in the plurality of document records. Information extraction is performed, and the result can be used in the same way as the attribute information included in the document record from the beginning, and related document records can be organized.

According to a sixth aspect of the present invention, there is provided a step of receiving a selection of one or more document records from a plurality of document records each comprising a text character string and attribute information of the text character string, Receiving a selection of an attribute as a reference for sorting one or more of the selected document records, and providing the one or more document records selected in the document record selection receiving step in the attribute selection receiving step. A step of performing sorting in accordance with the nature of the information of the attribute whose selection has been accepted; and a plurality of menus that allow the text string included in the document record to be classified based on a predetermined method. To apply to a specified text string contained in one or more of the document records selected from the document records of Selecting the text string according to the selected menu, and adding the result of the classification to the document record as information of the attribute corresponding to the classified text string. Is a computer-readable storage medium storing a program for causing a computer to execute the program.

Therefore, when there are a plurality of document records, a menu for classifying the text character strings contained therein based on a predetermined method is selected and applied according to the purpose, and the classification is performed. It can be used in the same way as the attribute information included in the document record from the beginning, and classification can be performed by combining a plurality of classification results. Further, the relation with other attributes is easily seen, and the efficiency of the information analysis work is improved.

[0019]

[First Embodiment of the Invention] FIG.
1 is a functional block diagram of an information analysis support device 1 according to a first embodiment of the present invention. The units constituting the information analysis support apparatus 1 of FIG. 1 are connected to each other, and are configured so that data of document records and control information can be exchanged between them as needed.

First, the data of the document record is received by the input means 2, and the document record holding means 3 accumulates the contents by attaching a unique document identifier to the content. here,
The document record is composed of a text character string constituting the contents of the document and zero or more pieces of attribute information. The attribute information is information on the attribute of the text character string, and includes an attribute name and a value for the attribute (hereinafter, referred to as “attribute value”). The attribute name is a label indicating what the attribute is, and may or may not be included in the document record. The attribute value is the actual content corresponding to the label. Examples of attribute information include the date the document was created and entered,
There are so-called bibliographic items such as a creator of a document, but it is not always necessary to include only bibliographic items. For example, various information that can be extracted from the document record (for example, the length of the included text character string) and the like can be attributed to, for example, “length of text character string” once extracted.

FIG. 2 is a schematic diagram of the data structure of the document record holding means 3. The “input date”, “input user”, “length of text string” and the like shown in the upper row of FIG. 2 are attribute names, and the data in each lower row is
Attribute value (other than for individual document records,
Although there is a unique document identifier, it is omitted in FIG. 2).
On an actual computer, such information is converted and stored in a general data storage format.

The document record display output means 4 displays one or more document records stored in the document record holding means 3 in a table format as shown in FIG.

The document record sorting means 5 sorts the selected one or more document records according to the value of the attribute specified by the sort reference attribute specifying means 6. The selection of a document record is performed via a document record selection means 7 which allows a user to select one or more document records from the displayed plurality of document records. This can be easily realized by a method adopted in a general user interface in which a pointer is placed on a target document record among a plurality of displayed document records and a selection is instructed there.

The selection of an attribute serving as a sorting reference is made via a sorting reference attribute designating means 6. This is done by, for example, a method adopted in a general user interface in which a pointer is placed on an attribute to be a reference among the attributes in the top line of the display in FIG. 3 and a selection is instructed there. Can be easily realized.

When a document record is selected and an attribute serving as a sorting criterion is designated, sorting is performed, and the result is output and displayed by the document record display output means 4. A detailed description of the sorting method is omitted. It should be noted that the document records in FIG. 3 are in a state of being sorted in ascending order by the input date attribute.

A plurality of information extracting means 8 (n in the example of FIG. 1 (n is an integer of 2 or more)) are means for extracting a character string satisfying a predetermined condition from a text character string. . Specifically, there is the following. --- "Unique name information extraction means" to extract unique names (company name, product name, etc.) --- "Date and time information extraction means" to extract information about date and time --- "Amount to extract information about money amount" Information Extraction Means "These can be realized by natural language processing technology or character string pattern matching processing technology. For example, the “specific name information extracting means” can be realized by morphological analysis of a character string using a company name dictionary, and the “date and time information extracting means” can perform pattern matching of a character string using a pattern dictionary of date and time notation. Can be realized by:

The information extraction instructing means 9 is for instructing which of the plurality of information extracting means 8 is to be applied to the text string contained in the selected document record. This can be easily realized by a method adopted in a general user interface, in which a list of prepared information extracting means 8 is displayed and selected from the list.

The text string of the document record selected by the document record selection means 7 is input to the information extraction instructing means 9.
When the processing is applied by the information extracting means 8 specified through the above, an extracted result is obtained for each document record. The attribute adding means 10 adds the attribute to the document record as an attribute.

For example, the application of the “unique name extracting means” to the text character string “printer of the company XX is beautiful” contained in the first document record of FIG. , "XX" is obtained as the company name. The attribute adding unit 10 adds "OO" as an attribute value for the attribute of the company name and adds this attribute information to the document record. This can be realized by a general process of adding attribute information to the end of the document record read from the document record holding unit 3 and storing it again in the document record holding unit 3.

FIG. 4 shows that the input date is from January 1, 1999.
Select document records up to 1999/1/30 (this is an easy task for the user by using the sorting result by the document record sorting means 5), and select "unique" for the text strings contained in them. FIG. 11 is a schematic diagram in which the result of adding to the target document record is displayed by the document record display output unit 4 using the application result of the “name extraction unit” as attribute information. For a document not to be processed, a display is displayed so that the document can be understood.

In the information analysis support device 1 of the present embodiment,
As described above, it is possible to add new attribute information to a document record by using the information extracting means 8. By combining this with the application of the processing by the prepared document record / sorting means 5, information analysis that has been difficult in the past becomes easier.

For example, when a new attribute obtained as a result of the information extraction, a company name, is designated as a sort reference attribute and sorting is applied, what is shown in FIG. 4 is rearranged as shown in FIG. ,Is displayed. Document records containing text strings describing each company are classified as a unit. Even when there are a large number of document records, the information can be rearranged in such a manner using the result of information extraction, so that the information analysis becomes much easier than before.

Although the effect differs depending on what kind of information extraction means 8 is provided, the sorting by attribute values is generally effective. It is important that you have what you need (technically possible) and make it available for selection.

Further, by incorporating processing (such as totaling of the number of data and cross-tabulation) as in general spreadsheet software into the information analysis support apparatus 1 of the present embodiment, further analysis support is possible. , This can be easily realized.

Next, a specific configuration example of the computer of the information analysis support apparatus 1 will be described. FIG. 6 is a block diagram illustrating an outline of a computer constituting the information analysis support device 1. As shown in FIG. 6, the information analysis support device 1
In FIG. 1, a CPU 11, a ROM 12, and a RAM 13 are connected by a bus 14. The CPU 11 centrally controls each unit of the information analysis support device 1. The ROM 12 stores a BIOS and the like. The RAM 13 stores data in a rewritable manner and serves as a work area for the CPU 11.

The bus 14 has an input device 15 composed of a keyboard and the like, and an output device 1 composed of a display.
6, a CD-ROM drive 17 and a hard disk 18 are connected. The realization of the various functions of the various means 2 to 10 shown in FIG.
-Performed based on a program read by the ROM drive 17 and stored in the hard disk 18. The input of the document record in the input means 2 is performed via the input device 15, the display of the document record by the document record display output means 4 is performed via the output device 16, and the storage of the document record by the document record holding means 3 is performed. , By storing the document record in the hard disk 18. Note that the recording medium is not limited to the CD-ROM 19, and various recording media such as a floppy disk and a magneto-optical disk can be used.

[Second Embodiment of the Invention] FIG. 7 is a functional block diagram of an information analysis support device 21 according to a second embodiment of the present invention. The units constituting the information analysis support apparatus 21 of FIG. 1 are connected to each other, and are configured to be able to exchange document record data and control information between the units as needed.

In FIG. 7, the input means 2, the document record holding means 3, the document record display and output means 4, the document record selection means 7, the sort reference attribute designating means 6, and the document record / sort means 5 are the embodiments of the present invention. Since the same means as in the first embodiment, a detailed description is omitted.

A plurality of text classification means 22 are provided (n (n is an integer of 2 or more) in the example of FIG. 7), and means for classifying a given text string into a plurality of categories or clusters. It is. Although various text classification technologies have been proposed and realized,
For example, there is the following.

--Means 1: Category Classification Based on Presence or Absence of Word First, individual text character strings are decomposed into words using a morphological analysis technique. For each category you want to classify, determine in advance which category each text string belongs to based on the criteria for the presence or absence of a word (for example, if the word XX is included, classify it into the category "XX"). To go.

-Means 2: Clustering based on similarity based on word commonality First, individual text character strings are decomposed into words using morphological analysis technology. Next, the similarity between the text strings is calculated based on the number of words commonly included in the two character strings to be compared. Calculate so that strings with common words have higher similarity than strings with only one common word, etc.), similarity between all pairs of strings After the degrees are calculated, those with similar similarities are put together in the same cluster.

-Means 3: Clustering of Text Strings by Distance Calculation Based on Spatial Configuration Using Singular Value Decomposition of Matrix Consisting of Text Strings and Words First, individual text strings are morphologically analyzed. To break it down into words. Next, a matrix expressing the relationship between the text string and the presence or absence of the word (for example, assigning the text string to a row and each word to a column, and placing the word in the text string at a position determined by the row and column) , Etc.). Using the singular value decomposition processing for the matrix, a space in which the text character string is arranged is constructed, and clustering is performed based on the distance there.

FIG. 8 is a schematic diagram of a data structure of a classification criterion based on the presence / absence of a word for each category, which is used for the “category classification based on the presence / absence of a word” of the means 1. The column on the left is the category name (value), and the column on the right is the word (or words) to be included, which is the criterion. The categorization based on this criterion is specifically performed in the following procedure, for example. First, the selected document record is read from the document record holding unit 3, and the text character string contained therein is decomposed into words by morphological analysis.
Are sequentially collated with words included in the reference data structure shown in FIG. If there is a match, return the category name. If no match is found, the category name returns the value "none".

FIG. 9 shows a display example of a document record to which the “category classification based on the presence or absence of a word” is applied.

The classification instructing means 23 is for instructing which of the plurality of text classifying means 22 is to be applied to the text character string contained in the selected document record. This can be easily realized by a method adopted in a general user interface, in which a list of prepared text classifications is displayed and selected from the list.

When all the document records shown in FIG. 9 are selected by the document record selecting means 7 and the application of "category classification based on the presence or absence of a word" is instructed by the classification instructing means 23, each of the selected document records is The category is assigned to the text string included in.
For example, the text string included in the first document record, "The printer of company XX is beautiful,"
Since “company” is included, the category name “XX” is obtained. The classification instructing means 23 enables the obtained result to be handled in the same manner as other attribute information.

That is, when the processing of the text character string of the document record selected by the document record selection means 7 by the text classification means 22 specified through the classification instruction means 23 is performed, As a result, the name or number of the category or cluster as the classification result is obtained. The attribute adding unit 24 adds the attribute to the document record as an attribute. This can be realized by a general process of adding attribute information to the end of the document record read from the document record holding unit 3 and storing it again in the document record holding unit 3.

All of the document records shown in FIG.
FIG. 10 shows a state in which the application result of the attribute adding unit 24 when the “category classification based on the presence or absence of a word” is performed is displayed via the document record display output unit 4. Note that the attribute name is given the name of the used classification standard data.

Once the attribute information has been obtained, it becomes possible for the document record sorting means 5 to sort them based on the attribute, as in the first embodiment of the invention. As a result, the classification result can be easily manipulated, and the information analysis is much easier than the conventional classification system.

In the embodiment of the present invention, by preparing a plurality of text classifying means 22 having different characteristics, their results can be obtained as separate attribute information.
First, the text is classified by a certain text classification means 22, sorted based on the attribute obtained by the classification, and only the document records having the same attribute value are selected, and another classification is performed on the document record. Processing becomes possible.

As a specific example, when it is desired to classify opinions on a product of a competitor by its contents, the classification is first performed by the company, and only the document record containing the opinion on the target company is selected from the result. , On the other hand, are classified by the similarity of the contents (means 2, means 3, etc.)
It becomes possible.

As described above, by utilizing the information analysis support apparatus 21, information analysis combining various functions becomes possible.

It should be noted that by incorporating processing (such as totaling of the number of data and cross-tabulation) as in general spreadsheet software into the present invention, further analysis support is possible, but this can be easily realized.

Further, by combining the information analysis support device 11 and the information analysis support device 21, the information that can be handled as an attribute increases, and it goes without saying that the use of the combination increases.

A specific example of the configuration of the information analysis support apparatus 21 on a computer will be described with reference to FIG.
Is the same as that described above, and a detailed description thereof will be omitted.

[0056]

An object of the present invention is to provide a menu for extracting a character string satisfying a predetermined condition from text strings included in a plurality of document records when there are a plurality of document records. The information is extracted by selecting and applying the information according to the information, and the result can be used in the same manner as the attribute information included in the document record from the beginning, and related document records can be organized.

According to a second aspect of the present invention, when there are a plurality of document records, a menu for classifying text strings included in the document records based on a predetermined method is selected and applied according to the purpose. Classification is performed, and the result can be used in the same manner as the attribute information included in the document record from the beginning, so that classification can be performed by combining a plurality of classification results. Further, the relation with other attributes is easily seen, and the efficiency of the information analysis work is improved.

According to a third aspect of the present invention, when there are a plurality of document records, a menu for extracting a character string satisfying a predetermined specific condition from a text character string included in the plurality of document records is provided according to the purpose. By selecting and applying the information, the information is extracted, and the result can be used in the same way as the attribute information included in the document record from the beginning, and related document records can be organized.

According to a fourth aspect of the present invention, when there are a plurality of document records, a menu for classifying text strings contained therein based on a predetermined method is selected and applied according to the purpose. Classification is performed, and the result can be used in the same manner as the attribute information included in the document record from the beginning, so that classification can be performed by combining a plurality of classification results. Further, the relation with other attributes is easily seen, and the efficiency of the information analysis work is improved.

According to a fifth aspect of the present invention, when there are a plurality of document records, a menu for extracting a character string which satisfies a predetermined specific condition from a text character string included in the plurality of document records is provided according to the purpose. By selecting and applying the information, the information is extracted, and the result can be used in the same way as the attribute information included in the document record from the beginning, and related document records can be organized.

According to a sixth aspect of the present invention, when there are a plurality of document records, a menu for classifying the text character strings contained therein based on a predetermined method is selected and applied according to the purpose. Classification is performed, and the result can be used in the same manner as the attribute information included in the document record from the beginning, so that classification can be performed by combining a plurality of classification results. Further, the relation with other attributes is easily seen, and the efficiency of the information analysis work is improved.

[Brief description of the drawings]

FIG. 1 is a functional block diagram of an information analysis support device according to a first embodiment of the present invention.

FIG. 2 is a schematic diagram of a data structure of a document record holding unit of the information analysis support device.

FIG. 3 is an explanatory diagram showing a display example of a document record on a document record display output unit of the information analysis support device.

FIG. 4 is an explanatory diagram showing a display example of a document record on a document record display output unit of the information analysis support device.

FIG. 5 is an explanatory diagram showing a display example of a document record on a document record display output unit of the information analysis support device.

FIG. 6 is a block diagram showing an outline of a computer constituting the information analysis support device.

FIG. 7 is a functional block diagram of an information analysis support device according to a second embodiment of the present invention.

FIG. 8 is a schematic diagram of a data structure of an example of a classification standard used by the information analysis support device.

FIG. 9 is an explanatory diagram showing a display example of a document record to which the classification criterion is applied.

FIG. 10 is an explanatory diagram showing a display example of an application result of an attribute adding unit of the information analysis support device when performing classification according to the classification criterion.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 Information analysis support means 3 Document record holding means 5 Document record / sort means 6 Sort reference attribute designation means 7 Document record selection means 8 Information extraction means 9 Information extraction instructing means 10 Attribute disable means 21 Information analysis support means 22 Text classification means 23 Classification instruction means 24 Attribute addition means

Claims (6)

[Claims] 1. A document record holding means for storing a plurality of document records each comprising a text character string and attribute information of the text character string, and at least one document from the document records stored in the document record holding means. Document record selecting means for selecting a record; sort reference attribute specifying means for enabling selection of an attribute to be a reference for sorting the selected one or more document records; and the document record selecting means A document record sorting means for sorting one or more document records selected according to the property of the information of the attribute specified by the sort reference attribute specifying means; One or more types of information extracting means for extracting a character string satisfying the condition of Information extraction instructing means for enabling selection of the information extracting means to be applied to a specified text character string included in one or more document records obtained, and a character string extracted by the selected information extracting means And an attribute adding unit for adding, as information of the attribute corresponding to the text character string of the extraction destination, to the document record. 2. A document record holding means for storing a plurality of document records each comprising a text character string and attribute information of the text character string, and at least one document from the document records stored in the document record holding means. Document record selecting means for selecting a record; sort reference attribute specifying means for enabling selection of an attribute to be a reference for sorting the selected one or more document records; and the document record selecting means Document record sorting means for performing sorting on one or more document records selected according to the property of the information of the attribute specified by the sort reference attribute specifying means, and a plurality of the text character strings are predetermined. At least one type of text classifying means for classifying based on the selected method; A classification instructing unit that enables selection of the text classification unit to be applied to a specified text string included in one or more document records; and a classification result obtained by the selected text classification unit. An information analysis support device comprising: an attribute adding unit that adds the attribute information corresponding to the text character string to the document record. 3. A type that enables a character string satisfying a predetermined condition to be extracted from the text character string included in a document record including a text character string and attribute information of the text character string. Selecting, from the menu, one to apply to a specified text string contained in one or more of the document records selected from a plurality of the document records; and selecting the character according to the selected menu. An information analysis support method, comprising: extracting a string; and adding the extracted character string to the document record as information of the attribute corresponding to the text string of the extraction destination. 4. One or more types of menus that enable the text string included in a document record including a text string and attribute information of the text string to be classified based on a predetermined method. Selecting what to apply to a specified text string contained in one or more of the document records selected from the plurality of document records; and classifying the text string according to the selected menu And a step of adding the classified result to the document record as information of the attribute corresponding to the classified text character string. 5. A step of receiving a selection of one or more document records from a plurality of document records each including a text character string and attribute information of the text string; Receiving a selection of an attribute to be a reference for sorting records; and information of the attribute received by the attribute selection receiving step for one or more document records received by the document record selection receiving step. Performing a sort in accordance with the nature of a plurality of the text strings included in the document record, from one or more types of menus capable of extracting a character string that satisfies predetermined conditions Select what to apply to a specified text string contained in one or more of the document records selected from the document record Extracting the character string according to the selected menu; and adding the extracted character string to the document record as information of the attribute corresponding to the text character string of the extraction destination. A computer-readable storage medium that stores a program to be executed by a computer. 6. A step of receiving a selection of one or more document records from a plurality of document records each including a text character string and information on attributes of the text character string; Receiving a selection of an attribute to be a reference for sorting records; and information of the attribute received by the attribute selection receiving step for one or more document records received by the document record selection receiving step. Performing a sort according to the nature of the document record, and classifying the text character string included in the document record based on a predetermined method.
Selecting from the more than one type of menu what to apply to the specified text string contained in one or more of the document records selected from the plurality of document records; and Storing a program for causing a computer to execute a step of classifying a text character string and a step of adding the result of the classification to the document record as information of the attribute corresponding to the classified text string. A readable storage medium.