GB2362734A - Accessing text information using a concept map - Google Patents

Abstract

A text information database accessing and searching method comprising generating an associative dictionary by associative analysis from primary text information, performing an associative retrieving process depending on a user-input keyword 1, visualising associative information relevant to the keyword as a concept map 2, and retrieving the primary information based on a second keyword 3 designated by the user on the concept map and presenting the primary information relevant to the second keyword 4. The first keyword includes a keyword indicating a key concept, a keyword indicating a focus which is an attribute of the concept, and a keyword indicating an aspect, which limits the type of keyword to a specific area, and may be selected by a pull-down menu.

Description

2362734 1 TEXT INFORMATION ACCESS APPARATUS, METHOD, AND STORAGE MEDIUM

THEREFOR The present invention relates to a system of accessing text information, and more specifically to a text information access apparatus, method, and storage medium for the apparatus which enable target information to be accessed from text information such as a document database, etc.

With the recent - progress of information technology (IT), office automation (OA) systems have become more and more popular. In addition, personal ccffs have becene more popular for domestic use year by year. Furthermore, with the spread of the Internet the network age has ccue upon us, and access to Web pages of the World Wide Web (WW) and electric commerce (EC) have been put to practical use. In addition, storage devices such as hard disk devices, optical disk devices, etc. have been marketed with a larger capacity at a more reasonable piice. As a resdlt, an increasing number of documents have been managed in 2 electronic files.

The abovementioned progress of - IT has largely increased the amount of information, u4iich can be referred to as a flood of information, in the world.

Normally, documents are stored as text in databases, f iles, Web pages, etc. However, under circumstances described above, it is very difficult to immediately obtain useful and desired information from a large and increasing amount of information.

A growing amount of text information increases the volume of knowledge available to persons and organizations. However, unless important knowledge can be accessed and effectively utilized, most of the knowledge is uselessly stored. Therefore, it is important to be able to utilize an increasing amount of text information, and an effective utilization system is strongly demanded.

To utilize text information, it is necessary to successfully access desired or useful information.

However, there are the following problems.

1. How can the desired or useful information be successfully accessed? 2. How can the existence of the information be osignalled? [Retrieval using a key] 3 Retrieval using a key is performed by a user using a key (keyword or keyword-based queries containing the keyword representing user-desired information). It can involve, for exmple, a retrieval system in which the author of the contents preliminarily assigns a key, and an entire text retrieval system in which any partial character string in the contents can be used as a key.

[Browsing through a link] This is a system of retrieving hypertext represented by WM Contents of hypertext can be linked with the contents of other pieces of hypertext using link information embedded in the hypertext. The user can access various information while browsing through the hypertext by tracing the links preliminarily assigned by the author of the hypertext.

In the above mentioned retrieval system using a key, whether or not desired information can be successfully obtained depends on the key. Therefore, it is important to prepare an appropriate key. However, to prepare the appropriate key, the purpose of the retrieval has to be clarified. Furthermore, it is hard for a common user to prepare a key correctly representing desired information-. In addition, with an increasing amount of information, 4 it becomes more and more difficult to specify information by appropriately combining keys.

On the other hand, the problem to be considered bef ore preparing a key is that no keys can be prepared f or information of which a user is not aware. In this case, however useful for the user the information may be, he or she cannot access the information without noticing the existence of it.

In addition, for example, a person in charge of business planning of a company, is required to search for the situation of the business of competitors. In response to such a requirement, the conventional system has difficulty preparing an appropriate key (keyword- based queries) although a newspaper database is available.

Additionally, in the range of the currently provided keyword-based queries, which depends on a retrieval engine, it may be impossible to prepare an appropriate key.

In the above mentioned browsing system using a link, desired information can be accessed even when the purpose of the retrieval is not clarified. Additionally, when information is linked with intentionally or incidentally accessed hypertext, the user can access the information even if the user is not aware of the information. However, since a link destination of the contents of hypertext is determined by a link prepared by an author of the contents, an accessible range is limited to closed information determined by a range prepared by the author.

Bmbodiments of the present invention may provide a database access system capable of accessing desired information without using a strict key in accessing a document database, and presenting information in the document database of which a user has not been aware.

A text information access apparatus according first asnect of the Dresent invention includes to a the following three units; An associative dictionary generation unit generating an associative dictionary by performing an associative analysis according to - primary information about text.

A map generation unit performing an associative retrieval process according to a first keyword input by a user, and visualizing associative information relevant to the first keyword as a concept map.

A retrieval unit retrieving the primary information according to a second keyword sp6-cif ied by the user on the concept map, and presenting the 6 primary information relevant to the second keyword The first keyword may be used by a user as the focus of a feature or a tendency of information. The map generation unit may present a concept map on which a keyword input as a focus, and keywords highly relevant to the keyword are arranged (first embodiment).

In addition,. the first keyword can further contain a keyword functioning as an aspect limiting the type of keyword highly relevant to the keyword as a focus (second embodiment).

The first keyword can further contain a keyword which is a key concept indicating an item that the user demands to know, a keyword which is a focus indicating the attribute of the item, and a keyword which is an aspect limiting the type of keyword highly relevant to the keyword which is a focus (third embodiment).

The text information access apparatus with the above mentioned configuration according to the second aspect of the present invention may present a user with the secondary information representing the feature and the tendency of the contents (primary information) of the text information in the form of a conceiDt map according to the aspect of the user, may clarify the 7 user's intention to retrieve information, and may inform the user about the existence of information of which the user has not been aware. Then, it may automatically retrieve the primary information relevant to the keyword by specifying a keyword in the concept map in which the user takes interest in the first, second, or third embodiment. Therefore, the user can access the target primary information with associations. In addition, the above nentioned concept map may be generated based on the contents of the text information, that is, the primary information.

However, unlike the summary of individual contents, ---itmay represent the entire contents, or the feature and the tendency of a subset of the contents. Therefore, even when (in a step in which) the user cannot suggest the optimum keyword, an appropriate keyword can be input by using the concept map as a retrieval interface, thereby successfully accessing the target primary information.

A text information access apparatus according to the sec aspect of the present invention may be applied when the above mentioned first keyword is based on the above mentioned third embodiment, and, in addition to each of the units forming th text information access apparatus according to the first 8 aspect of the present invention, it may include the following units.

A storage unit preliminarily storing a set of the keyword which is a focus and the keyword which is an aspect.

A menu display unit displaying as a menu the set of the keyword which is a focus and the keyword which is an aspect stored in the storage unit.

A selection unit selecting a set of a keyword which is a user-desired focus and a keyword which is an aspect from the menu.

A text information access apparatus according to a third aspect of the present invention further include an automatic classification unit for automatically classifying the primary information, and generating a database for each category in addition to each unit included in the text information access apparatus according to the first aspect mentioned above. The associative dictionary generation unit nay generate the associative dictionary based on the database for each category.

A text information access apparatus according to a fourth aspect of the present invention may further include an information extraction/thesaurus generation unit for extracting information from the 9 primary information, generating a thesaurus, and generating a dictionary f or each category in addition to each unit included in the text information access apparatus according to the f irst aspect mentioned above. Then, the map generation unit way generate the concept map by referring to the dictionary for each category.

A text information access apparatus according to a f if th aspect of the present invention may further include an automatic translation unit for automatically translating a keyword stored in the associative dictionary in cooperation with the associative dictionary generation unit in addition to each unit included in the text information access apparatus according to the f irst aspect mentioned above., the map generation unit may visualize and display a concept map on which a keyword translated by the automatic translation unit is arranged.

A text information access apparatus according to a sixth aspect of the present invention way include, in addition to each unit included in the text information access apparatus according to the first aspect of the present invention, an automatic classification unit for automatically classifying the primary information, and generating a database for each category, an information extraction/ thesaurus generation unit for extracting information from the primary inf6rmation, generating a thesaurus, and generating a dictionary f or each category, and an automatic translation unit for automatically translating a keyword stored in the associative dictionary in cooperation with the associative dictionary generation unit. The map generation unit may generate a concept map by referring to the dictionary for each category.

With a configuration of text information access apparatus according to the first aspect of the present invention, a retrieval unit of a text information access apparatus according to a seventh aspect of the present invention may present a list of the primary information before presenting the entire text of the primary information relevant to the second keyword. In this case, the user may select desired primary information from the list.

A text information access apparatus according to a eighth aspect of the present invention way further include an automatic summary unit for automatically summarizing primary information selected by the user in addition to each unit included in a- text information access apparatus according to a seventh

11 aspect mentioned above, and may present smnery text generated by the automatic summary unit before presenting the entire text about the primary information.

A text information access apparatus according to a ninth aspect of the present invention may further include an automatic translation. unit for automatically translating the summary text generated by the automatic summary unit in addition to the configuration of the text information access apparatus according to the eighth aspect, and may present the summary text as translated text.

With a text information access apparatus according to a ninth aspect of the present invention, an automatic translation unit of a text information access apparatus according to a tenth aspect of the present invention may autematically translate the list of the primary information selected by the user, and may translate and present the list of the primary information.

In addition to the configuration of the text information access apparatus according to the seventh aspect of the present invention, a text information access apparatus according to an eleventh aspect of the present invention may further include an autematic 12 translation unit for automatically translating the primary information selected by the user, and an automatic summary unit for generating summary text of the primary information generated by the automatic translation unit, and may present the summary text generated by the automatic summary unit before presenting the entire text about the primary information.

With the configuration of the text information access apparatus according to the eleventh aspect of the present invention, an automatic translation unit of a text information access apparatus according to a twelfth aspect of the present invention may automatically translate the list of the primary information selected by the user, and translate and present the list of the primary information.

In addition to the configuration of the text information access apparatus according to the first aspect of the present invention, a text information access apparatus according to a thirteenth aspect of the present invention may further include a first automatic translation unit for automatically translating the primary information presented by the retrieval unit, and may present the translated texb of the primary information generated by the first 13 automatic translation unit.

In addition to the configuration of the text information access. apparatus according to the thirteenth aspect of the present invention, a text information access apparatus according to a fourteenth aspect of the present invention may further includ a second automatic translation unit for automatically translating a keyword stored in the associative dictionary in cooperation with the associative dictionary generation unit. The map generation unit may visualize and display a concept nap on which keywords translated by the second automatic translation unit are arranged.

With the. configuration of the text information access apparatus according to the thirteenth aspect of the present invention, -an automatic translation unit of a text information access apparatus according to a fifteenth aspect of the present invention may automatically translate a list of the primary information selected by the user, and translate and present the list of the primary information.

In addition to the configuration of the text information access apparatus according to thd sixth aspect of the present invention, a text information 14 access apparatus according to a sixteenth aspect of the present invention may further include an automatic summary unit for automatically summarizing the primary information selected by the user from the concept map, and an automatic translation unit for translating the summary text of the primary information generated by the automatic summary unit, and nay present the summary text generated by the automatic translation unit before presenting the entire text of the primary information.

In addition to the configuration according to the sixth aspect of the present invention, a text information access apparatus according to a seventeenth aspect of the present invention way further include.. an automatic translation unit for translating the primary information selected by the user from the concept map, and an automatic summary unit for generating summary text of the primary information generated by the automatic translation unit, and may present the summary text generated by the automatic summary unit bef ore presenting the entire text of the primary information.

In addition to the configuration of the text information access apparatus according to thd first aspect of the present invention, a text information access apparatus according to an eighteenth aspect of the present invention may further include a first designation unit designating at least one second keyword on the concept map.

A text information access apparatus according to a nineteenth aspect of the present invention is based on a configuration in.which the concept map is formed by second keywords and relation lines connecting the second keywords. In addition to the configuration of the text information access apparatus according to the first aspect of the present invention, or the configuration of the text information access apparatus according to the eighteenth aspect of the present invention, a text information access apparatus according to the nineteenth aspect of the present invention may include a second designation unit for designating at least one relation line on the concept map, and a determination unit for detecting the relation line designated by the second designation unit, and determining the second keyword used when the retrieval unit retrieves the primary information from the relation line..

In a text information access apparatus according to a twentieth aspect of the Oresent invention with the configuration of the text 16 information access apparatus according to the nineteenth aspect of the present invention, the determination unit way detect the designation position of the relation line, and determine the weight of the second keyword designated by the second designation unit, the retrieval unit retrieving the primary information based on the weight for the second keyword determined by the determination unit.

A text information accessing method according to the present invention includes a step (a) of generating an associative dictionary by an associative analysis from primary information of text, a step (b) of performing an associative retrieving process depending on a user-input first keyword, and visualizing associative information relevant to the f irst keyword as a concept map, and a step (c) of retrieving the primary information based on the second keyword designated by the user on the concept map, and presenting the primary information relevant to the second keyword.

A computer-readable storage medium according to the present invention stores a program for directing a computer to perform a process including: a step (a) of generating an associative dictionary by an associative analysis from primary information of 17 text; a step (b) of performing an associative retrieving process depending on a user-input f irst keyword, and visualizing associative information relevant to the first keyword as a concept map; and a step (c) of retrieving the primary information based on the second keyword designated by the user on the concept map, and presenting the primary information relevant to the second keyword.

Brief Description of the Drawings

FIG. 1 shows the configuration of the hardware according to an embodiment of the present invention; FIG. 2 is a general flowchart of the entireoperation of an embodiment of the present invention; FIG. 3 shows an example of the process performed according to the general flowchart shown in FIG. 2; FIG. 4 is a block diagram of the configuration of the system according to the first embodiment of the present invention; FIG. 5 is a flowchart of the entire operation of the first embodiment of the present invention; FIG. 6 is a flowchart of a practical example of the procedure of the process of a retrieval- phase shown in FIG. 5; 18 FIG. 7 shows an example of the user input system using an input of a focus; FIG. 8 shows an example of the user input system using a focus and an aspect; FIG. 9 is a flowchart of the retrieval procedure using f our types of user input systems based on a focus and an observation point; FIG. 10 shows an example of the user input system in which a key concept, a focus, and an aspect are input; FIG. 11 shows an example of the user input system in which a set of a key concept, a focus, and an aspect is FIG. realizing the FIG. 13 storing menu FIG. 14 of realizing shown in FIG FIG. 15 of realizing shown in FIG.

FIG. 16 configuration input; 12 shows an example of a menu screen f or user input system shown in FIG. 11; shows the data structure of a file data for the menu screen display; is a flowchart of the process procedure the retrieval by the user input system 10; is a flowchart of the process procedure the retrieval by the user input system 11; is a block diagram showing the system of the database access apparatus 19 according to the invention; FIG. 17 is a flowchart of the entire operation of the database access apparatus according to the second embodiment; FIG. 18 is a block diagram of the system configuration of the database access apparatus according to the third embodiment of the present invention; FIG. 19 shows an example of a concept map displayed when a dictionary of company names is used as a dictionary for each category in the database access apparatus according to the third embodiment of the present invention; FIG. 20 is a flowchart of the entire operation of the database access apparatus according to the third embodiment; FIG. 21 is a block diagram of the system configuration of the database access apparatus according to the fourth embodiment of the present invention; FIG. 22 is a flowchart of the entire operation of the database access apparatus according to the fourth embodiment; FIG. 23 is a block diagram of the system second embodiment of the present configuration of the database access apparatus according to the fifth embodiment of the present invention; FIG. 24 is a flowchart of the entire operation of the database access apparatus according to the fifth embodiment; FIG. 25 is configuration of according to the invention; FIG. 26 is of the database sixth embodiment; FIG. 27 is configuration of according to the invention; FIG. 28 is a flowchart of the entire operation of the database access apparatus according to the seventh embodiment; FIG. 29 is a block diagram of the system configuration of the database access apparatus according to the eighth embodiment of the present invention; FIG. 30 is a flowchart of the entire operation a block diagram of the system the database access apparatus sixth embodiment of the present a flowchart of the entire operation access apparatus according to the a block diagram of the system the database access apparatus seventh embodiment of the present 21 of the database access apparatus according to the eighth embodiment; FIG. 31 is a block diagram of the system configuration of the database access apparatus according to the ninth embodiment of the present invention; F1P. 32 is a flowchart of the entire operation of the database access apparatus according to the ninth embodiment; FIG. 33 is a block diagram of the system configuration of the database access apparatus according to the tenth embodiment of the present invention; FIG. 34 is a flowchart of the entire operation of the database access apparatus according to the tenth embodiment; FIG. 35 is a block diagram of the system configuration of the database access apparatus according to the eleventh embodiment of the present invention; FIG. 36 is a flowchart of the entire operation of the database access apparatus according to the eleventh embodiment; FIG. 37 is a block diagram ofthe -system configuration of the database access apparatus 22 according to the twelf th embodiment of the present invention; FIG. 38 is a flowchart of the entire operation of the database access apparatus according to the twelfth embodiment; FIG. 39 is a block diagram of the system configuration of the database access apparatus according to the thirteenth embodiment of the present invention; FIG. 40 is a flowchart of the entire operation of the database access apparatus according to the thirteenth embodiment; FIG. 41 is a block diagram of the system configuration of the database access apparatus according to the fourteenth embodiment of the present invention; FIG. 42 shows the system of retrieving primary information by designating secondary information by clicking a button on a node on a concept map; FIG. 43 is a flowchart of the process procedure for realizing the system of retrieving primary information by designating secondary information by specifying a node as shown in FIG. 42; FIG. 44 shows the system of retrieving primary information by designating secondary information by 23 clicking a button on a relation line on a concept map; FIG. 45 is a flowchart of the process procedure for realizing the system of retrieving primary information shown in FIG. 44; FIG. 46 shows the system of performing a weighted retrieval process by designating primary information by a click position of a relation line on a concept map; FIG. 47 shows an example in which a retrieval result depends on the click position when a button is clicked on the same relation line in the weighted retrieval.system shown in FIG. 46; FIG. 48 is a flowchart of the process procedure for realizing the system of performing the weighted retrieval process on the primary information shown in FIG. 46;.

FIG. 49 is a flowchart of the process procedure for realizing the system of retrieving primary information by clicking a mouse on a node as shown in FIG. 42,. and realizing the system of retrieving primary information 40 by clicking a mouse on a relation line as shown in FIG. 44; FIG. 50 is a flowchart of the process procedure for realizing the system of retrieving primary information by clicking a mouse on a node as shown in 24 FIG. 42, and realizing the system of retrieving primary information 40 by a click position of a relation line as shown in FIG. 46; FIG. 51 is a flowchart of the process procedure for retrieving primary information by selecting plural pieces of secondary information by clicking a mouse on a plurality of nodes and a plurality of relations; FIG. 52 is a flowchart of the process procedure for retrieving primary information by selecting plural pieces of secondary information by detecting the click on a plurality of nodes and the click positions of a plurality of relation lines; and FIG. 53 shows an example of an interface for retrieving primary information by selecting plural pieces of secondary information by detecting the click on a plurality of nodes and the click positions of aplurality of relation lines.

FIG. 54 is a block diagram of the hardware configuration of a computer realizing each embodiment of the present invention.

Description of the Preferred Embodiments

Embodiments of the present invention are described below by referring to the attached drawings.

According to the embodiments of the present invention, a concept map is information about a database used as secondary The method of generating a concept map is described in the documents [1] and [2) published by the inventor of the present invention.

[1] Isamu Watabe and Kazuo Misue,'Text Mining by associative relation of words' published in Information Basic Study Material 99-Fl-55 (1999) in the 55th Meeting of Information Processing Society of Japan.

[2] Isamu Watabe and Kazuo Misue, 'Technology of Visualizing Associative Relation for Text Mining' published in Information Basic Study Material 99-FI-55 (1999) in the 55th Meeting of Information Processing Society of Japan.

Described below is the outline of the method of generating a concept map.

1. Computing relation level The relation level between keywords represented on a map is computed. First, a document is processed by a morphological analysis, etc. to extract a keyword. Then, the frequency and the cooccurrence level of the extracted keyword in each document are obtained, and the relation level between keywords is computed.

26 2. Selecting keyword A keyword to be used on a map is selected. For example, it is selected on a user specified selection condition.Otherwise, keywords are ranked in a statistic process, etc., and the keywords having a higher order are selected 3. Layout A map is generated by laying out keywords on a two-dimensional plane. The layout of keywords is obtained by considering the distance between keywords, etc. so that the relation level can be easily recognized. on a two-dimensional plane.

Described below is the database access apparatus according to an embodiment of the present invention using the concept map.

FIG. 1 shows the configuration of the hardware according to an embodiment of the present invention.

The system shown in FIG. 1 is a client/server system in which a database server 10 and a plurality of clients 20 (20-1, 20-N) are connected through a network 30.

The database server 10 comprises a CPU 11, an external storage device 12, an input device 13, and a. display 14. The external storage device 12 stores a database. Information requested by the 27 client 20 to the database server 10 is read by the CPU 11 from the database, and transmitted to the client 20.

The database is formed by inputting a command and data from the input device 13. The contents of the database are displayed by the CPU 11 on the display 14, and the database can be edited on the screen of the display 14.

The client 20 comprises a CPU 21, an input device 22, and a display 23. On the client side 20, a user accesses the database provided in the database server 10 through a GUI (graphical user interface) displayed on the display 23. At this time, the input device 22 is used to input data to the GUI. Although not shown in FIG. 1, the client 20 comprises a storage device storing a program to direct the CPU 11 to perform the function of accessing the database according to an embodiment of the present invention described later.

The storage device can be a storage medium drive device for driving a removable and portable storage medium, an external storage device, etc. using a storage medium such as a magnetic disk, an optical disk, etc. As a portable storage medium, a CD (CDROM, CD-R, CD-RW, etc.) of various standards, an optical disk (magneto - optical disk, phase variable 28 optical disk, etc.) of various standards, a DVD (DVDROM, DVD-RAM, DVDR/W, EVE + RW, etc.) of various standards, a PC card, a floppy disk, etc.

The network 30 is a LAN, a MAN, a WAN, Internet, Intranet, Extranet, etc. The database server 10 and the client 20 are connected to the network 30 through a network connection appliance such as a NIC (network interface card), a router, a modem, etc. depending on the type of the network 30.

FIG. 2 is a general flowchart of the entire operation of the embodiment with the above mentioned configuration.

First, when a user inputs data on the client side 20 (in step Sll), the database server 10 is accessed, the secondary information is generated by the search for the database stored in the external storage device 12, and the information is displayed on the display 23 (in the form of a concept map) (step S12).

Then, the user specifies and inputs a portion in which he or she takes interest in the secondary information displayed on the display 23 (step S13).

In response to the input, the client 20 accesses the above mentioned database on the database server side 10 again, and retrieves the primary information which is the factor of the feature of the above 29 is mentioned specified portion from the database, and display on the display 23 a candidate list of the obtained primary information)for example, a list of only titles) (step S14). Then, the user selects desired primary information from the candidate list through the input device 22 (step S15). Through the selection, the client 20 accesses the database of the database server 10 again, reads the selected primary information from the database, and displays (presents) the contents on the display 23 (step S16).

When there is only one candidate for the primary information in step S14, control can be immediately passed to step so that the primary information can be presented. The process procedure can be applied to a similar process flowchart described later.

FIG. 3 shows an example of the process performed according to the general flowchart shown in FIG. 2.

The process is described below by referring to the process sequence 1 through 6 shown in FIG. 3.

1. Input (step S11) As shown in FIG. 3, two input fields 'key concept' and 'aspect/focus' are displayed on the display 23. Assume that the user inputs IXYZ Corporation' as a key concept and 'strategy/business' as an aspect/focus in response to the display as shown in FIG. 3. Also assume that the user inputs immediately after requesting to know the information about 'XYZ Corporation', focusing the information on business', and requesting-to obtain the information from the aspect of 'strategy'.

Thus, the 'key concept 1 is a keyword of the information about what a user wants to know, the 'focus' is a keyword indicating what is made much of about the information, and the 'aspect' is a keyword indicating from what aspect the user wants to know the information input as the 'focus'.

2. Presenting secondary information (step S12) Based on the input described in 1 above, For example, the map (concept map) as shown in FIG. 3 is displayed on the display 23. In this example, some (three in this example) business ('Zl business', 'Yl business', and 'Zl business') highly relevant to XYZ Corporation are retrieved, and are fixedly displayed at the vertex of a polygon (a triangle in this example). In addition, the strategy name (a word followed by 'strategy', for example, 1... strategy') highly relevant to the business is arranged and displayed on the plane depending on the relation level. In this example, 'aaa strategy' and 'bbb strategy' are displayed. Arrangements are made such 31 that the 'aaa strategy 1 is highly relevant to the 'X1 business ' while the ' bbb strategy 1 is highly relevant to the 1Y1 business'. In addition, in FIG. 3, the line connecting each business and each strategy name, or between strategy names is referred to as a 'relation line'.

In addition, although not shown in FIG. 3, keywords highly relevant to the strategy name arranged on the screen are retrieved from the database as necessary, and the process of arranging them around the strategy.name is. recursively repeated.

3. Input on secondary information (step S13) As shown in FIG. 3, the secondary information is selected by clicking a mouse button, etc. on the keyword retrieved as an 'aspect' or a relation line retrieved as an 1 aspect. In the example shown in FIG. 3, the 'aaa strategy' or the relation line connecting the 'aaa strategy' and the 'Z1 business' is selected.

When the mouse button is clicked on the keyword aaa strategy, the primary information about the ' aaa strategy' is retrieved and displayed (presented). When the relation line connecting the 'aaa strategy' and the 'Zl business' is selected, the primary 25- information (primary information containing the two 32 keywords) as the factor connecting them is retrieved and displayed (presented).

4. Displaying a list of primary information (step S14) In response to the secondary information described in 3 above, the primary information relevant to the secondary information is retrieved f rom the database, and is displayed (presented) on the display 23. The number of pieces of corresponding primary information depends on the size of a database, but is actually large. Therefore, it. is not practical to immediately display all primary information on the display 23 (for example, the number of pieces of primary information that can be displayed is limited by a screen size, etc.). Accordingly, the list of the primary information is displayed on the display 23, and is presented to the user. At this time, f or example, the title of the primary information is listed as shown in FIG. 3. In this case, it is displayed in order of, for example, ranking.

5. Selecting primary information (step S15) User-desired primary information is selected from the list of primary information presented as candidates.

6. Presenting primary information (step S16) 33 The contents of the primary information selected by the user in 5. above are displayed on the display 23. FIG. 3 shows the contents of '3. aaa strategy of XYZ Corporation' selected from the list shown in FIG. 3-4.

FIG. 4 is a block diagram of the system configuration of the database access apparatus according to the first embodiment of the present invention.

In FIG. 4, an associative dictionary generation unit 50 generates an associative dictionary 60 from primary information 40 such as a newspaper article 41, a patent document 42, a Web document 43, etc.

A map generation unit 70 represents the associative information stored in the associative dictionary 60 as a concept map 80, and presents it to a user 110.

A retrieval engine 90 retrieves original text 100 of corresponding primary information 40 by the operation of the user 110 on the concept map 80, and presents the original text 100 to the user 110.

According to the first embodiment with the above mentioned configuration, the user can obtain the concept map 80 representing the feature afid the tendency of the primary information 40. In addition, 34 the user can access the primary information 40 through the concept map 80.

FIG. 5 is a flowchart of the entire operation according to the f irst embodiment with the above mentioned configuration. The flowchart can be roughly classified into a generation phase of a dictionary, etc. and a database (M) access phase (retrieval phase).

In the generation phase of a dictionary, etc., the primary information 40 is collected and stored (step S21), and then the associative dictionary generation unit 50 generates the associative dictionary 60 (step S22).

When the user 110 inputs a keyword (step S23), the map generation unit 70 refers to the associative dictionary 60, and presents secondary information from the concept map 80 (step S24). Then, when the user 110 selects any secondary information from the concept map 80 (step S25), the retrieval engine 90 is informed of the selected secondary information, retrieves the primary information 40 containing the secondary information, and lists the primary information 40 as a candidate (step S26). When the user 110 selects any primary information 40 from the list, (step S27), the retrieval engine 90 displays (presents) the contents of the primary information 40 on the screen of the display 23 (step S28).

Described below in detail is the configuration according to the first embodiment of the present invention.

When the concept map 80 is generated, it is necessary for the user 110 to input a 'focus' to specify about what the feature and the tendency are to be checked.

FIG. 6 is a flowchart of an embodiment of the process procedure of the above mentioned retrieval phase. In FIG. 6, a step also shown in the flowchart of the retrieval phase shown in FIG. 5 is assigned the same step number assigned in FIG. 5. The retrieval phase shown in FIG. 6 is characterized by the listing of the titles of the prospective primary information 40 as the listing of the primary information 40 (step S26A)., and by the presentation of the original text as the presentation of the primary information 40 (step S28A). The processes of other steps are the same as those shown in the flowchart in FIG. 5.

FIG. 7 shows an example of the user input system using the above mentioned focus input.

As shown in FIG. 7A, for example, thred input fields 130 of 'focus' are displayed on the screen of

36 the display 23. When the keywords 'Xl business', 'Yl business', and 'Zl business' are input to the three input f ields 130 as shown in FIG. 7, three keywords 81 are fixedly arranged at the vertexes of the triangle on the concept map 80 as shown in FIG. 7B, and keywords 82 (in this example, 1 aaa ' and ' bbb 1) highly relevant to these three keywords are arranged at appropriate positions on the concept map 80. In addition, the input keyword is displayed as being connected to the automatically displayed keyword 82 through a relation line 83.

In the above mentioned display system, the relationship between the automatically retrieved and displayed keyword 82 and the keyword 81 input as a 'focus' can be obtained from the relative positions on the concept map 80. Then, for example, a user 110 is made to take much interest in the arrangements of keywords, that is, to consider the reason why a keyword and another keyword are positioned close to each other, the reason why a keyword is highly relevant to the keywords 81 positioned at all vertexes, etc. As a result, the user 110 is interested in the primary information.

FIG. 8 shows another example of a user- input system.

37 41 In this example, an input field 140 of an aspect 1 is provided. That is, an 'aspect 1 can be input in addition to a 'focus'. In the example shown in FIG. 8A, a 'strategy' is displayed as an aspect. An aspect is input as in the example shown in FIG. 7. Thus, by inputting a 'strategy' as an aspect, the keyword (word) 81 automatically displayed on the concept map 80.is limited to a strategy name (a word ended with 'strategy' as 1... strategy' as shown in

FIG. 8B.

When various types of words are arranged on the concept map 80, information sometimes becomes scattered, and it is difficult to know what the concept map 80 means. Therefore, as an example shown in FIG. 8, the type of the relation line 83 automatically arranged on the concept map 80 can be limited by the 'aspect' by allowing the 'strategy' to be input in addition to the 'focus', thereby clarifying the meaning of the concept map 80.

FIG. 9 is a process flow of the retrieval procedure using the input system shown in FIGS. 7 and In FIG. 9, a step also shown in the flowchart in FIG. 5 is assigned the same step number. Sihce the processes of the steps have been described above, the 38 explanation is omitted here.

The f lowchart shown in FIG. 9 shows four types of user input systems. In the f irst input system, only a 'focus' is input in step S31. In the second input system, only an 'aspect' is input in step S41. In the third input system, a ' f ocus 1 is input in step S52, and then an 'aspect' is input in step S52. In the fourth input system, an 'aspect' is input in step S61, and then a ' focus' is input in step S62. The difference between the third input system and the fourth input system is that the 'focus' is input first in the third input system whole the 'aspect' is input first in the fourth input system.

When a map generation button displayed on the input screen of the display 23 is clicked in step S71 after only a 'focus', only an 'aspect', or a 'focus' and an 'aspect' are input in any of the above mentioned f irst through fourth input systems, the processes in steps S24 through S28 are performed, and the primary information is f inally presented to the user 110.

From a viewpoint of a user, a user sometimes does not practically understand a focus. For example, there is not a small number or requests to know the 'main business of XYZ Corporation' not practically 39 about 'business of... In this case, as shown by the example in FIGS. 7 and 8, the system of practically inputting a 'focus' requires the user to preliminarily check the main business of XYZ Corporation.

FIG. 10 shows an example of an input for solving the above mentioned problems.

In this example, as shown in FIG. 10A, the user 110 inputs a 'key concept' 150, a 'focus' 160, and an 'aspect' 140. The focus 160 is not a practical focus as the ' aspect' 140 shown in FIGS. 5 and 6, but a type of focus. Since 'business' is input as the focus 160 in this case,a business name (for example, a word ended with 'business' as 1... business') highly relevant to the IXYZ Corporation' input as a key concept is retrieved from the associative dictionary 6 0, and 3 higher order business names (in this example, 'Xl business', 1Y1 business', and 'Z1 business') are focuses. Then, as shown in FIG. 10B, these businesses are arranged at the vertexes of a triangle on the concept map 80.

In the input system shown in FIG. 10, the user 110 inputs the types of focus and aspect. However, when the type of primary information and the uJage of database is appropriately limited, several combinations of a focus and an aspect can be prepared.

FIG. 11 shows a system of allowing the user 110 to select a combination of a 'focus' and a 'aspect' (aspect/focus) from a menu.

In FIG. 11A, a 'strategy/business' is presented as a menu of an aspect/focus 165. As the I key concept' 150, IXYZ Corporation' is input. When the user 110 selects the menu displayed as an example, the concept map 80 similar to the map shown in FIG. 9 is displayed as shown in FIG. 11B.

As a combination of a focus and an aspect, in addition to the above mentioned example, and in consideration of the usage for business in a database such as newspaper articles, etc., 'market/business', strategy/market', 'company/business', strategy/area', etc. can be suggested in addition to strategy/business' (strategic relation is checked depending on the business).

FIG. 12 shows an example of the menu screen which realizes the input system shown in FIG. 11.

In this example, as shown in FIG. 12A, a combo box 170 is used on the menu display of the aspect/focus 165. When a lower scroll button 170a of the combo-box 170 is clicked, a list 170b is-opened as shown in FIG. 12B. The list 170b displays a list 41 of 'measure/business', 'measure/market', Y strategy/business I strategy/market ',... When the button is clicked on the 'strategy/business' and selected from the list 170b, the 'strategy/business' is displayed in the box portion 170c of the combo-box 170 as shown in FIG. 12B. In addition, a map generation button 173 is provided for the menu screen. The map generation button 173 is used by the user 110 specifying the generation of the concept map 80.

Provided below the map generation button 173 is a concept map display area 174 in which the concept map 80 is displayed.

In FIG. 12B, a part of the screen is hidden by the combo-box 170 as a result of selecting the 'strategy/business', but the concept map 80 as shown in FIG. 11 is displayed.

FIG. 13 show the data structure of a f ile 175 above mentioned menu data. As shown in f records 0, 1, 2,... of the file 175 data such as 'measure/business', measure/market', 'strategy/business',..., etc. By referring to the file 175, the list 170b of the combo box 170 is displayed.

In addition to the system shown in FIG. - 12, a system of inputting an aspect and a f ocus using a menu storing the FIG. 13 thE store menu 42 can also be realized by presenting 1 aspects ' and 'focuses' as an individual menu to the user 110 who individually selects an 1 aspect' and a ' focus ' f rom the menu. In this case, a file storing menu data for 'aspects' and a file storing menu data for 'focuses' can be individually prepared, or the f ile 175 shown in FIG. 13 can be used.

FIGS. 14 and 15 are flowcharts of the method of retrieving the primary information in which the system of inputting an 'aspect' and a 'focus' using the above mentioned two menus is adopted.

FIG. 14 is a flowchart of the retrieving method using the system of inputting an 'aspect' and a 'focus' from individual menus.

In this system, the user 110 first inputs a key concept (step S81). Then, focuses are displayed on a menu, and the user 110 selects a desired focus from the menu (step S82). Then, aspects are displayed on a menu, and the user 110 selects a desired aspect from the menu (step S83).

The steps S81, S82, and S83 can be skipped.

As described above, after the user 110 has input a key concept, a f ocus, and an aspect, he or she clicks the map generation button 173 on the screen (step S84). When the input of the click is conf irmed, 43 it is determined whether or not all input items (a key concept, a focus, and an aspect) have been correctly entered (step S85). The determining process is performed because the user 110 may click the map generation button after skipping the steps S81 through S82 without completely inputting a key concept, a focus, and an aspect.

If it is determined in step S85 that the input items have not been completely input (NO in step S85), then control is returned to step S81 after displaying a warning message, etc. to allow the user 110 to input an item not yet input. If it is determined in step S85 that the input items have all been correctly input (YES in step S85), then the processes in steps S24 through S28 are performed.

FIG. 15 is a flowchart of the retrieving method using the system (refer to FIG. FIG. 12) of inputting an 'aspect' and a 'focus' on the same menu.

In this system, the user 110 first inputs a key concept (step S91). Then, a menu of combinations of a focus and an aspect is displayed, and the user 110 selects a desired combination of a focus and an aspect (step S92).

As described above, when the user 110 ihputs a keyword and a combination of a focus and an aspect, 44 he or she clicks the map generation button on the screen (step S93). When the input of the click is confirmed, it is determined whether or not all input items (a key concept, and a combination of a f ocus and an aspect) have been correctly entered (step S94).

The determining process is performed because the user may click the map generation button after skipping the steps S91 and S92 without completely inputting a key concept, and a combination of a f ocus and an aspect.

If it is determined in step S94 that the input items have not been completely input (NO in step- S94), then control is returned to step S91 after displaying a warning message, etc. to allow the user 110 to input an item not yet input. If it is determined in step S94 that the input items have all been correctly input (YES in step S94), then the processes in steps S24 through S28 are performed.

Described below is the database access apparatus according to the second embodiment of the present invention.

When an associative computation is performed based on various documents, and when a user wants to check the associative relation to the businegs of a company, the information about the sports sponsored by the company exists as noise, thereby suppressing an expected associative relation. In this case, characteristics from noise appear on the concept map 80 of the secondary information, and there can be the possibility that a user who intends to grasp the features and tendency of a database according to the secondary information may be confused. Therefore, it is desired that pieces of information for which associative computation is performed share common subject with each other. To attain this, the function of classifying primary information by a subject is required.

The second embodiment shown in FIG. 16 has the above mentioned function. A component shown both in FIG. 6 and FIG. 4 showing the f irst embodiment is assigned a common number.

An automatic classification unit 210 classified the primary information 40, and generates a database 220 for each category (DB for each category) such as a database 220-1 containing articles (documents) only about, for example, the 'automobile industry', a database 220-1 containing articles (documents) only about 'new product information', etc.

According to the second embodiment, the associative dictionary generation unit 50 performs 46 associative computation using the database 220 for each category storing documents classified by subject, and generates an associative dictionary 60 with high purity (low noise).

An aspect can be specified in several methods. For example, they can be a method used according to part-of-speech information assigned based on a morphological analysis, a method used according to a character string pattern such as 1... business', etc.

Although it is difficult to classify a company name, a technical term, a product name, etc. by a part of speech or a. character string pattern, they can be used as an aspect if a dictionary for each category can be prepared, thereby extending the types of aspect on the concept map 80.

FIG. 17 is a flowchart of the entire operation according to the second embodiment with the above mentioned configuration. A step shown both in FIGS. 5 and 17 is assigned a common step number, and the explanation of the step is omitted here.

The third embodiment shown in FIG. 18 is performed using the dictionary for each category. Also, a component shown both in FIG. 18 and FIG. 4 showing the first embodiment is assigned a comm6n unit number.

47 In FIG. 18, an information extraction/thesaurus generation unit 310 automatically generates a dictionary 320 for each category such as a technical term dictionary 320-1, a product name dictionary 320- 2, etc. by extracting information from the primary information 40 and then generating a thesaurus. The map generation unit. 70 can arrange various aspects on the concept map 80 by referring to the dictionary 320 for each category when the concept map 80 is generated. A user can also input various aspects.

FIG. 19 shows a display example of the concept map 80 according to the third embodiment with the above mentioned configuration.

FIG. 19 actually shows a display example of the concept map 80 obtained when a company name dictionary is generated as the dictionary 320 for each category. As shown in FIG. 19A, when the user 110 inputs 'XYZ Corporation' as the key concept 150, 'company' as the focus 160, and 'company' as the aspect 140, the concept map 80 relating to the IXYZ Corporation' input as the key concept 150 is displayed as shown in FIG. 19B. In this example, the businesses (the Xl, Y1, and Z1 businesses) handled by XYZ Corporation are arranged at the vertexes of the triangle, and the triangle contains the names of other companies (AAA, BBB, and 48 is CCC) related to the businesses and the XYZ Corporation, each of them being connected with each other through relation lines, and appropriately positioned.

FIG. 20 is a flowchart of the entire operation according to the third embodiment with the above mentioned configuration. A step shown both in FIGS. 5 and 20 is assigned a common step number, and the explanation of the step is omitted here.

In the flowchart shown in FIG. 20, the information extraction/thesaurus generation unit 310 extracts information from the primary information obtained in step S21, and generates the dictionary 320 for each category by generating a thesaurus (step S111). Then, the associative dictionary generation unit 50 generates the associative dictionary 60 from the primary information 40 (step S112). Then, in the retrieval phase, the above mentioned processes in steps S23 through S28 are performed, and the primary information 40 is presented to the user 110.

Described below is the database access apparatus according to the fourth embodiment of the present invention.

FIG. 21 is a block diagram ofthe -system configuration according to the fourth embodiment of 49 the present invention. A component shown both in FIG.

21 and FIG. 4 showing the first embodiment is assigned a common unit number.

In FIG. 21, an automatic translation unit 410 cooperates with an associative dictionary generation unit 50a to automatically translate a keyword when the associative dictionary generation unit 50a generates the associative dictionary 60. The associative dictionary generation unit 50a cooperates with the automatic translation unit 410 to generate an associative dictionary 60a of the keywords translated into a mother tongue from the primary information 40 in a foreign language.

A map generation unit 70a presents a concept map 80a of keywords in the mother tongue based on the associative dictionary 60a. The map generation unit 70 translates the secondary information in a mother tongue selected on the concept map 80a into a foreign language, and notifies the retrieval engine 90 of the result.

FIG. 22 is a flowchart of the entire operation according to the fourthembodiment with the above mentioned configuration. A step shown both in FIG. 5 and 22, is assigned a common step number, and the explanation about the step is omitted here.

In the flowchart shown in FIG. 22, the automatic translation unit 410 cooperate with the associative dictionary generation unit 50a to generate the associative dictionary 60a by translating the primary information 40 in a foreign language into a mother tongue (steps S121 and S122). The processes in these steps are described below further in detail. That is, the associative dictionary generation unit 50a extracts a keyword in a foreign language entered in the associative dictionary 60a from the primary information 40, passes the keyword to the automatic translation unit 410 which translates it into the key concept in the mother tongue. Then, the associative dictionary generation unit 50a stores the keyword in the mother tongue received from the automatic translation unit 410 in the associative dictionary 60a. Afterwards, in the retrieval phase, the processes in steps S23 through S28 are performed, and the primary information 40 is presented to the user 110.

In the fourth embodiment, the associative dictionary 60a storing a keyword translated from the primary information 40 in a foreign language to the mother tongue is generated, and the concept nfap 80a on which a keyword is represented by a mother tongue 51 is presented to the user 110 as secondary information. Therefore, the user 110 can select target secondary information on the concept map 80a without considering the language in which the primary information 40 is represented. According to the present embodiment, the primary information 40 containing keywords selected by the user 110 on the concept map 80a is presented to the user 110 in a foreign language.

It is difficult to correctly grasp the outline of a document in a foreign language while browsing the document as compared with a document in a mother tongue. However, since the secondary information is presented in a mother tongue on the concept map 80a, it becomes much easier for the user 110 to browse the document.

According to the present embodiment, a keyword is automatically translated. Therefore, it is not necessary to strictly translate a sentence. As a result, technology without high translation precision will be available.

FIG. 23 is a block diagram of the system configuration of the database access apparatus according to the fifth embodiment of the present invention.

All functions of the fourth embodiment are 52 incorporated into the f if th embodiment of the present invention. For example, any of the above mentioned f irst through fourth embodiments can be selected by selecting a mode, etc.

An associative dictionary generation unit 50c has the function of the associative dictionary generation unit 50 and the associative dictionary generation unit 50a. An associative dictionary 60c contains the associative dictionary 60 and the associative dictionary 60a. A map generation unit 70c has the functions of the map generation unit 70 and the map generation unit 70a, and a concept map 80c contains the concept map 80 and the concept map 80a.

FIG. 24 is a flowchart of the entire operation according to the fifth embodiment with the above mentioned configuration. A step shown in the flowcharts in FIGS. 5, 17, 22, and 24 is assigned a common step number. The process flow shown in the flowchart in FIG. 24 contains the entire processes shown in the f lowcharts in FIGS. 5, 17, and 22. Step S134 contains steps S22, S112 and S122 shown in the flowcharts respectively in FIGS. 5, 17, 20, and 22.

When the primary information 40 is traced from the concept map 80, and when it is selected -f rom a candidate list, the primary information is not always 53 the primary information 40 requested by the user 110.

In this case, the user 110 wants to browse the presented primary information 40. However, it is not easy to browse it when the entire information is presented. Therefore, it is very convenient for the user 110 to be presented with a summary document before the entire primary information 40 is presented.

FIG. 25 is a block diagram of the system configuration of the database access apparatus according to the sixth embodiment having the function of summarizing the primary information 40. A component shown both in FIG. 25 and FIG. 4 showing the first embodiment is assigned a common unit number, and the function of the unit is omitted here.

An automatic summary unit 510 receives the entire text of the primary information 40 selected by the user 110 from a retrieval engine 90a, and generates a summary document 520. Then, the automatic summary unit 510 displays the summary document 520 on the display 23, thereby presenting it to the user 110.

FIG. 26 is a flowchart of the entire operation according to the sixth embodiment with the above mentioned configuration. A step shown both in FIGS.

and 26 is assigned a common step number, ahd the explanation of the process in the step is omitted 54 here.

When the primary information 40 is selected (step S27), the retrieval engine 90a transmits the original text 100 to the automatic summary unit 510. The is automatic summary unit 510 generates the summary document 520 of the original text 100 of the primary information 40 (step S141), displays the contents of the summary document 520 on the display 23, thereby presenting it to the user 110 (step S142).

When the primary information 40 is written in a foreign language, it is convenient for the user 110 if it can be automatically translated into a mother tongue and presented to the user 110.

FIG. 27 is a block diagram of the system configuration of the database access apparatus according to the seventh embodiment of the present invention with the function of translating the primary information 40 in a foreign language. A component shown both in FIG. 17 and FIG. 4 showing the f irst embodiment is assigned a common unit number, and the explanation of the function is omitted here.

An automatic translation unit 610 receives the original text 100 of the primary information 40 selected by the user 110 from a retrieval engine 90b through the concept map 80, automatically translates it, and generates a translated document 620 of the primary information 40. Then, the automatic translation unit 610 displays the contents of the translated document 620 on the display 23, and presents them to the user 110.

FIG. 28 is a flowchart of the entire operation according to the seventh embodiment with the above mentioned configuration. A step shown both in FIGS. 5 and 28 is assigned a common step number, and the explanation of the process in the step is omitted here.

When the primary information 40 is selected (step S27), the retrieval engine 90b transmits the original text 100 of the primary information 40 to the automatic translation unit 610. The automatic translation unit 610 generates the translated document 620 of the original text 100 of the primary information 40 (step S151), displays the contents (translated text) of the translated document 620 on the display 23, and presents them to the user 110 (step S152).

It is more convenient for the user 110 if the function of displaying the concept map 80a in a mother tongue according to the fourth embodiment is added to the seventh embodiment, and the listing of the primary 56 information 40 is displayed in a mother tongue.

FIG. 29 is a block diagram of the system configuration of the database access apparatus according to the eighth embodiment of the present invention having the above mentioned function. A component shown in FIG. 29, FIG. 4 showing the first embodiment, and FIG. 21 showing the fourth embodiment is assigned a common unit number, and the explanation of the function is omitted here.

In addition to the translating function of the automatic translation unit 610, an automatic translation unit 610a translates the list of the titles of the primary information 40 in a mother tongue received from a retrieval engine 90c, and displays the translation result on the display 23.

FIG. 30 is a flowchart of the entire operation according to the seventh embodiment with the above mentioned configuration. A step shown both in FIGS. 28 and 30 is assigned a common step number, and the explanation of the step is omitted here.

When secondary information is input on the concept map 80a written in a mother tongue (step S25), the retrieval engine 90b retrieves the primary information 40 containing the secondary information, generates the retrieval result (a list of the titles 57 of the primary information 40 in a foreign language), and transmits it to the automatic translation unit 610a. The automatic translation unit 610a translates the received list of the titles of the primary information 40 (step S161), displays it on the display 23, thereby presenting the list of the titles of the primary information 40 in a mother tongue to the user 110 (step S162). Then, the processes in steps S27, S151, and'S152 shown in the flowchart in FIG. 28 are sequentially performed.

FIG. 31 is a block diagram of the system configuration of the database access apparatus according to the ninth embodiment of the present invention.

The ninth embodiment is obtained by adding the function of translating a summary document of the primary information 40 to the sixth embodiment shown in FIG. 25. A component shown both in FIG. 31 and FIG. 25 according to the sixth embodiment is assigned a common unit number, and the explanation of the function is omitted here.

An automatic summary unit 510a transmits the summary document 520, generated by the automatic summary unit 510a itself, in a foreign language to an automatic translation unit 710. Theautomatic 58 translation unit 710 translates the summary document 520 into the mother tongue, and generates a translated document 720 of the summary document 520. Then, the automatic translation unit 710 displays the translated document (summary document) 720 in the mother tongue on the display 23.

FIG. 21 is a flowchart of the entire operation according to the ninth embodiment with the above mentioned configuration.. A step shown both in FIGS.

26 and 32 is assigned a common step number, and the explanation about the process of a part of the step is omitted here.

When the primary information 40 is selected (step S27), the retrieval engine 90a transmits the original text 100 of the primary information 40 to the automatic summary unit 510a. The automatic summary unit 510a generates the summary document 520 of the original text 100 of the primary information 40, and transmits the summary document 520 to the automatic translation unit 710 (step S141). The automatic translation unit 710 translates the received summary document 520 into a mother tongue, and generates the translated document 720 (step S171). Then, the automatic translation unit 710 displays the translated document 720 on the display 23, thereby presenting the 59 summary document (summary text) 520 in the mother tongue to the user 110 (step S172).

FIG. 33 is a block diagram of the system configuration of the database access apparatus according to the tenth embodiment of the present invention.

The tenth embodiment is obtained by adding the function of displaying the titles of the primary information 40 in a mother tongue to the above mentioned ninth embodiment. A component shown both in FIG. 33 and FIG. 31 showing the ninth embodiment is assigned a common unit number, and the explanation of the function is omitted here.

The retrieval engine 90b transmits a list of the titles of the primary information 40 in the mother tongue to an automatic translation unit 710a. The automatic translation unit 710a translates the list into a mother tongue, and displays the translation result on the display 23.

FIG. 34 is a flowchart of the entire operation according to the tenth embodiment with the above mentioned configuration. A step shown both in FIGS. 32 and 34 is assigned a common step number, and the explanation of the step is omitted here.

Upon receipt of the list of the titles of the primary information 40 in a foreign language from the retrieval engine 90b, the automatic translation unit 710a translates it into a mother tongue (step S181). Then, the automatic translation unit 710a displays the list of the titles of the primary information 40 in the mother tongue, thereby presenting the list to the user 110 (step S182). The user 110 selects a desired tile of the primary information 40 from the list in the mother tongue (step S182). Then, the processes in the above mentioned steps S141, S171, and S172 are sequentially performed, and the summary text of the primary information 40 selected by the user 110 is displayed on the display 23.

FIG. 35 is a block diagram of the system configuration of the database access apparatus according to the eleventh embodiment of the present invention.

The eleventh embodiment is different from the ninth and tenth embodiments. That is, it first translates the entire text of the primary information 40, and then automatically summarizes the entire text.

A component shown both in FIG. 35 and FIG. 31 showing the ninth embodiment is assigned a common unit number, and the explanation of the function is omitted here.

The retrieval engine 90c transmits the entire 61 text of the primary information 40 in a foreign language selected by the user 110 to an automatic translation unit 810. The automatic translation unit 810 translates the entire text into the mother tongue, generates a translated document 820 of the entire text of the primary information 40, and transmits it to an automatic summary unit 910.

The automatic summary unit 910 summarizes the entire text of the received primary information 40 in the mother tongue, generates a summary document 920, and displays it on the display 23.

FIG. 36 is a flowchart of the entire operation according to the eleventh embodiment with the above mentioned configuration. A step shown both in FIGS.

32 and 36 is assigned a common step, and the explanation of the process in the step is omitted here.

When the primary information 40 is selected (step S27), the retrieval engine 90c transmits the original text (entire text) 100 of the primary information 40 to the automatic translation unit 810. The automatic translation unit 810 translates the original text 100 of the primary information 40 into a mother tongue, generates the translated document 820, and transmits the translated document 820 to the automatic summary

62 unit 910 (step S191). The automatic summary unit 910 summarizes the received translated document 820, and generates the summary document 920 (step S192). Then, the automatic summary unit 910 presents the summary document (summary text) 920 in- the mother tongue by displaying the summary document 920 on the display 23 (step S193).

FIG. 37 is a block diagram of the system configuration of the database access apparatus according to the twelfth embodiment of the present invention.

The twelfth embodiment is obtained by adding the function of displaying the title of the primary information 40 in a mother. tongue to. the above mentioned eleventh embodiment. A component shown both in FIG. 37 and FIG. 35 showing the first embodiment is assigned a common unit number, and the explanation of the function is omitted here.

A retrieval engine.90d transmits a list of the titles of the primary information 40 in a mother tongue to an automatic translation unit 810a. The automatic translation unit 810a translates the list into the mother tongue, and displays the translation result on the display 23.

FIG. 38 is a flowchart of the entire operation 63 according to the twelf th embodiment with the above mentioned configuration. A step shown both in FIGS.

36 and 38 is assigned a common step number, and the explanation of the process in the step is omitted here.

Upon receipt of a list of the titles of the primary information 40 in a mother tongue, the automatic translation unit 810a translates it into the mother tongue (step S201). Then, the list of the titles of the primary information 40 in the mother tongue is displayed on the display 23, thereby presenting the list to the user 110 (step S202). The user 110 selects a desired title of the primary information 40 from the list in the mother tongue (step S203). Then, the processes in the above mentioned steps S191 through S193 are sequentially performed, and the summary text of the primary information 40 selected by the user 110 is displayed on the display 23.

FIG. 39 is a block diagram of the system configuration of the database access apparatus according to the thirteenth embodiment of the present invention.

The thirteenth embodiment has all functlons of the above mentioned first through the tenth 64 embodiments. A component in the first through tenth embodiments and also shown in FIG. 39 is assigned a common unit number.

An associative dictionary generation unit 50d has the functions of the above mentioned associative dictionary generation units 50, 50a, and 50c. An associative dictionary 60d comprises the above mentioned associative dictionaries 60, 60a, and 60c.

A map generation unit 70d has the functions of the above mentioned map generation units 70 and 70a. A concept map 80d comprises the above mentioned concept maps 80, 80a, and 80c.

In addition, the retrieval engine 90d has the functions of the above mentioned retrieval engines 90, 90a, and 90b. An automatic summary unit 510d has the functions of the above mentioned automatic summary units 510 and 510a. An automatic translation unit 610d has the functions of the above mentioned automatic translation units 610, 610a, 610b, 710, and 710a. A translated document 620d comprises the above mentioned translated documents 620 and 720.

The thirteenth embodiment with the above mentioned configuration can present all functions according to the above mentioned first through tenth embodiments to the user 110. The user 110can selectively use the function appropriate for each use as necessary.

FIG. 40 is a flowchart of an example of the entire operation according. to the thirteenth embodiment with. the above mentioned configuration.

A step shown both in FIGS. 36 and 40 is assigned a common step number.

When secondary information is input on the concept map 80d (step S25), the primary information 40 containing the secondary information is retrieved by the retrieval engine 90d, and the list of the titles of the primary information 40 retrieved by the retrieval engine 90d is translated (step S213). When the user.110 selects (a title of) certain primary information 40 from the list (step S214), for example, the process described below is selectively performed.

First, in the first mode, the retrieval engine 90d displays (presents) the original text 100 of the selected primary information 40 on the display 23 (step S214).

In the second mode, the automatic translation unit 610d translates the original text (body) 100 of the selected primary information 40 (step S215), and the automatic translation unit 610d displays (presents) the contents of the translated text 66 (translated document 600d) of the primary information 40 on the display 23.

Furthermore, in the third mode, an automatic summary unit 910d automatically summarizes the original text 100 of the selected primary information 40 (step S217), and the automatic translation unit 610d displays (presents) the contents of the summary text (summary document 520) ' obtained by the summarizing process (step S218).

In addition, in the fourth mode, the automatic translation unit 610d automatically translates the original text 100 of the selected primary information 40 (step S217), and the automatic summary unit 510d automatically summarizes the translated document 620d obtained by the automatic translating process, and displays the summary document (the above mentioned summary text of the primary information 40) obtained by the automatic summarizing process on the display 23 (step S220).

The flowchart in FIG. 40 shows the process based on the retrieval of the primary information 40 in a foreign language. When the primary information 40 is written in a mother tongue, for example, the process in the flowchart in FIG. 4 or 7 is performed.- FIG. 41 is a block diagram of the system 67 configuration of the database access apparatus according to the fourteenth embodiment of the present invention.

The fourteenth embodiment has all functions of the above mentioned first through eighth, eleventh, and twelfth embodiments. A component in the embodiments and also shown in FIG. 41 is assigned a common number.

An associative dictionary generation unit 50e. has the functions of the above mentioned associative dictionary generation units 50, 50a, and 50c. An associative dictionary 60e comprises the above mentioned associative dictionaries 60, 60a, and 60c.

A map generation unit 70e has the functions of the above mentioned map generation units 70 and 70a. A concept map 80e comprises the above mentioned concept maps 80, 80a, and 80c.

In addition, the retrieval engine 90e has the functions of the above mentioned retrieval engines 90, 90c, and 90d. An automatic summary unit 510e has the functions of the above mentioned automatic summary units 510 and 910. A summary document 520e comprises the summary documents 520 and 920. An automatic translation unit 610e has the functions of the above mentioned automatic translation units 610, 610a, 810, 68 810a, and 710a. A translated document 620d comprises the above mentioned translated documents 620 and 820.

The fourteenth embodiment with the above mentioned configuration can present all functions according to the above mentioned first through eighth, eleventh, and twelfth embodiments to the user 110. The user 110 can selectively use the function appropriate for each use as necessary.

In this specification, the flowchart of the entire operation according to the fourteenth embodiment is not shown. However, in the fourteenth embodiment as in the thirteenth embodiment, each function of the first through eighth, eleventh, and twelfth embodiments can be appropriately selected by the user 110 as necessary by selecting a mode, etc. Furthermore, it is apparent that one of ordinary skill in the art can easily generate the flowchart of the entire operation according to the fourteenth embodiment based on the flowcharts of the entire operations according to the above mentioned first through eighth, eleventh, and twelfth embodiments.

FIG. 42 shows an example of the system of retrieving the target primary information 40 from secondary information through the concept map 80. In this example, the primary information 40 is retrieved 69 using the keyword corresponding to a specif ed node 1001 as a retrieval key by specifying (f or example, clicking a mouse button) the node 1001 (a rectangle and an ellipse enclosing a keyword corresponding to the secondary information such as the 'Xl business', laaa strategy', etc.) arranged on the concept map 80 by the input device 22.

FIG. 43 is a flowchart of the outline of the process procedure of realizing the system of retrieving the primary information 40 by specifying the node as shown in FIG. 42.

First, the secondary information is presented by the node 1001 on the concept map 80 (step S301). When the user 110 clicks the button on the node 1001 (step S302), the keyword (secondary information) in the node is transmitted from the map generation unit 70 to the retrieval engine 90, and the retrieval engine 90 retrieves the primary information 40 containing the keyword (step S303). When the retrieval engine 90 completes the retrieval, it displays the list of the titles of the primary information 40 containing the keyword on the display 23 in, for example, the ranking form, etc. (step S304).

FIG. 44 shows another example of the sy.,jtem of retrieving the target primary information 40 from the secondary information through the concept map 80.

As shown in FIG. 44, a relation line 1003 connecting the nodes 1001 is displayed on the concept map 80 in addition to the node 1001. The relation line 1003 indicates that there. is any relationship between the two nodes 10,01 connected to either end of the relation line 1005. In the example shown in FIG. 44, by specifying (for example, clicking a mouse button) the relation line 1003 shown in FIG. 44, the primary information 40 associated by any relationship with keywords can be retrieved by performing AND retrieval on. the keywords using as retrieval keys the keywords (in this example, laaa strategy' and IZ1 business') in the two nodes 1001 at either end of the specified relation line 1003 shown in FIG. 44.

FIG. 45 is a flowchart of the outline of the process procedure of realizing the system of retrieving the primary information 40 by specifying the relation line shown in FIG. 44.

First, the secondary information is presented from the node 1001 through the concept map 80 (step S401). When the user 110 clicks the mouse button on the relation line 1003 on the concept map 80 (step S402), the two keywords in the nodes 1001 at -either end of the relation line 1003 are transmitted from the 71 map generation unit 70 to the retrieval engine 90. When the retrieval engine 90 receives the two keywords, the retrieval engine 90 performs the AND retrieval using the two keywords, and retrieves the corresponding primary information 40 (step S403).

When the retrieval engine 90 completes the retrieval, the list of the titles of the primary information 40 with which the two keywords are associated is displayed on the display 23 in, for example, a ranking form, etc. (step S404).

FIG. 46 shows the system of retrieving the primary information 40 using the relation line 1003 shown in FIG. 44.

In this third retrieval system, as in the second retrieval system shown in FIG. 44, the AND retrieval is performed using as retrieval keys the two keywords (hereinafter referred to as keywords at either end) in the nodes 1001 at either end of the relation line 1003 by specifying the relation line 1003. At this time, the keywords at either end is weighted depending on the position specified on the relation line 1003.

Practically, for example, when the center (line ratio of 1:1) of the relation line 1003 is specified (clicked) as shown in FIG. 46A, the AND retrieval is performed with the weights of the keywords at either 72 end set equal to each other.

In addition, when the position (in this example, the position of the line ratio of 1:2) closer to the vaaa strategy' on the relation line 1003 is specified as shown in FIG. 46B, the AND retrieval is performed with the weight of the 'aaa strategy' set to 2, and the weight of the 'Zl business' set to 1.

* FIG. 47 shows retrieval results of the cases (a) and (b) shown in FIG. 46. A retrieval result 1010 of the case (a) shown in FIG. 46 is shown on the right of (a) shown in FIG. 47. A retrieval result 1020 of the case (b) shown in FIG. 46 is shown on the right of (b) shown in FIG. 47.

As it is apparent when the retrieval result 1010 is compared with the retrieval result 1020, the ranking may be changed when the same document (primary information 40) is retrieved by changing the weights of the keywords at either.end of the relation line 1003 depending on the position specified on the relation line 1003. When the weight of the laaa strategy' is set double the weight of the 'Zl business' in FIG. 46B, the articles about the 'aaa strategy' rank first and second in the ranking as shown in FIG. 47A. On the other hand, in the dase of the retrieval shown in FIG. 4A, the articles about the 73 Zl business' and the 'aaa strategy' are ranked nearly equal.

FIG. 48 is a flowchart of the outline of the process procedure of realizing the third retrieval system shown in FIG. 46.

First, the secondary information is presented from the node 1001 through the concept map 80 (step S501). When the user 110 clicks the mouse buttonon the relation line 1003 on the concept map 80 (step S502), the click position on the relation line 1003 is detected. Based on the detected click position, the map generation unit 70 determines the weights of the keywords at either end of the relation line 1003, and transmits the keywords and their weights to the retrieval engine 90 (step S503).

When the retrieval engine 90 receives the keywords and their weights, the AND retrieval with the weight to the primary information 40 is performed (step S504). When the retrieval engine 90 completes the retrieval, the list of the titles of the primary information 40 with which the two keywords are associated is displayed on the display 23 in, for example, a ranking form, etc. (step S505). As a result, the list of the titles of the 15rimary information 40 is presented to the user 110 with the 74 weight depending on the click position reflecting on the list.

FIG. 49 is a f lowchart of an embodiment of a retrieving method in which the retrieval can be performed by either the first retrieval system shown in FIG. 42 or the second retrieval system shown in FIG. 44.

First, the secondary information is presented from the node 1001 through the concept map 80 (step S601). When the user 110 clicks the mouse button on the relation line 1003 on the concept map 80 (step S602), the keywords (secondary information) in the node is transmitted from the map generation unit 70 to the retrieval engine 90, and the retrieval engine 90 retrieves the primary information 40 containing the keywords (step S604).

On the other hand, when the user 110 clicks the mouse button on the relation line 1003 on the concept map 80. (step S603), the two keywords in the node 1001 on either end of the relation line 1003 are transmitted from the map generation unit 70 to the retrieval engine 90, the retrieval engine 90 retrieves the primary information 40 containing the keywords (step S604).

When the retrieval is completed in step S604, the retrieval engine 90 displays (presents) the list of the corresponding primary information 40 on the display 23 based on the retrieval result (step S605).

FIG. 50 is a flowchart of an embodiment of a retrieving method in which the retrieval can be performed by either the first retrieval system shown in FIG. 42 or the third retrieval system shown in FIG. 46.

First, the secondary information is presented from the node 1001 through the concept map 80 (step S701). When the user 110 clicks the mouse button on the node 1001 on the concept map 80 (step S702), the keywords (secondary information) in the node is transmitted from the map generation unit 70 to the retrieval engine 90, and the retrieval engine 90 retrieves the primary information 40 containing the keywords (step S705).

On the other hand, when the user 110 clicks the mouse button on the relation line 1003 on the concept map 80 (step S703), the click position on the relation line 1003 is detected. The map generation unit 70 determines the weight of the keywords at either end of the relation line 1003 from the detected click position, and transmits the keywords and their xeights to the retrieval engine 90 (step S704).

76 Upon receipt of the keywords on either end and their weights, the retrieval engine 90 performs the AND retrieval with a weight to the primary information 40 (step S705).

When the retrieval is completed in step S705, the retrieval engine 90 displays (presents) the list of the corresponding primary information 40 on the display 23 based on the retrieval result (step S706).

FIG. 51 is a flowchart of another embodiment of a retrieving method in which the retrieval can be performed by either the first retrieval system shown in FIG. 42 or the second retrieval system shown in FIG. 44. According to the embodiment, it is possible to specify (click on) a plurality of nodes 1001 and a plurality of relation lines 1003 on the concept map 80, and the AND retrieval can be performed using all specified keywords.

First, the secondary information is presented from the node 1001 on the concept map 80 (step S801).

When the user 110 click the mouse button on the node 1001 on the concept map 80, the map generation unit 70 stores the keywords (secondary information) in the node (step S802), and control is returned to step S801.

On the other hand, if the user 110 clicks the 77 mouse button on the relation line 1003 on the concept map 80, two keywords in the nodes 1001 at either end of the relation line 1003 are stored (step S 803), and control is returned to step S801).

The processes in steps S802 and S803 are repeated until a 'retrieval button' displayed at a predetermined position on the screen of the display 23 is clicked. If the retrieval button is clicked, the map generation unit 70 transmits all stored keywords to the retrieval engine 90 (step S804).

Upon receipt of a keyword from the map generation unit 70, the retrieval engine 90 retrieves the primary information 40 (step S805) by performing the AND retrieval using all received key concepts. When the retrieval is completed, the title of the corresponding primary information 40 is displayed (presented) on the display 23 (step S806).

FIG. 52 is a flowchart of another embodiment of a retrieving method in which the retrieval can be performed by either the first retrieval system shown in FIG. 42 or the third retrieval system shown in FIG.

46. In this embodiment, it is possible to specify (click on) a plurality of nodes 1001 and a plurality of relation lines 1003, and the AND retrieval-with a weight using all specified keywords is performed.

78 First, the secondary information is presented f rom the node 1001 through the concept map 80 (step S901). When the user 110 clicks the mouse button on the relation line 1001 on the concept map 80, the map generation unit 70 stores the keywords (secondary information) in the node (step S902), and control is returned to step S901.

on the other hand, when the user 110 clicks on the relation line 1003 on the concept map 80 (step S903), the click position on the relation line 1003 is detected. The map generation unit 70 determines the weights of the keywords at either end of the relation line 1003 from the detected click position, and stores the keywords and their weights (step S904), thereby returning control to step S901.

The processes in steps S902 through S904 are repeated until the 'retrieval button' displayed at a predetermined position of the screen on the display 23 is clicked. When the retrieval button is clicked (step S905), the map generation unit 70 transmits all stored keywords and their weights to the retrieval engine 90 (step S905). In this c ase, the weight of the keyword specified by the click on the node 1001 is appropriately set based on the specification etc.

Upon receipt of the keywords and their weights 79 from the map generation unit 70, the retrieval engine 90 retrieves the primary information 40 by performing the AND retrieval with a weight based on all received keywords and their weights (step S906). When the retrieval engine 90 completes the retrieval, it displays (presents) the title of the corresponding primary information 40 on the display 23 (step S907).

FIG. 53 shows an example of a user interface screen for realizing a retrieving method in which the user 110 can specify a plurality of nodes 1001 and a plurality of relation lines 1003 as described above by referring to the flowcharts shown in FIGS. 51 and 52.

In the example shown in FIG. 53, there is a is retrieval button 2001 on the right of the map generation button 173 of the menu screen shown in FIG. 12. The retrieval button 2001 corresponds to the retrieval button in the flowchart shown in FIGS. 51 and 52. By clicking the retrieval button 2001, for example, the concept map 80 as shown in FIG. 53 is displayed in the concept map display area 174.

FIG. 54 is a block diagram of the hardware configuration of a computer realizing each embodiment of the present invention.

A computer 3000 comprises a CPU 3001, ROM 3002, RAM 3003, an external storage device 3004, a storage medium drive device 3005, an input/output device 3006, and a communications interface 3007, and these components are interconnected through a bus 3008.

A program for realizing the functions of each embodiment according to the present invention is stored in a portable storage medium 3009 removable from the external storage device 3004, and the storage medium drive device 3005. A part or all of the above mentioned program can be stored in the ROM 3002.

The above mentioned program can be downloaded by the communications interface 3007 from an information provider 5000 through a network 4000. In this case, the program is downloaded into the portable storage medium 3009 installed in the external storage device 3004, the storage medium drive device 3005, or the RAM 3003.

The external storage device 3004 stores the primary information 40, the associative dictionary 60, the original text 100, the file 175, the database 220 for each category, the dictionary 320 for each category, the summary document 520, the translated document 620, etc. The summary document 520 and the translated document 620 can be stored in the RAM 3003.

The primary information 40 is stored in the 81 external storage device 3004 through the input/output device 3006 or the communications interface 3007. The primary information 40 can be stores in the portable storage medium 3009 installed in the storage medium drive device 3005. When the primary information 40 exists in an external database server, a Web site, etc. connected through the network 4000, the primary information 40 is accessed online in real time through the communications interface 3007, or is downloaded by the communications interface 3007 onto the external storage device 3004, etc. In addition, the primary information 40 can be generated through the -input/output device 3006.

-The input/output device 3006 comprises an input device having a pointing device such as a keyboard, a mouse, etc., a display such as a CRT display, an LCD (liquid crystal display) a PDP (plasma display), etc., a printing device such as a printer, etc.

The above mentioned program is invoked by an instruction from the input/output device 3006, and can be executed by the CPU 3001 loading from the external storage device 3004, the portable storage medium 3009, etc. to the RAM 3003. At this time, a user interface screen such as a GUI; etc. -is displayed on the display of the input/output device 3006, and the user invokes 82 the program by, for example, clicking a mouse button on an icon on the user interface screen. By invoking the program, a menu screen as shown in FIG. 12, etc. is displayed on the display of the input/output device 3006 under the control of the CPU 3001, and the user can retrieve desired primary information 40 in the above mentioned method on the menu screen. At this time, the concept map 80 is displayed on the display of the input/output device 3006, and the user specifies desired secondary information by clicking the mouse button on the node 1001 and the relation line 1003. According to the specification, the retrieval of the primary information 40 is started by the above mentioned AND retrieval, etc. When the retrieval is completed, a list of the retrieved primary information 40 is displayed on the display of the input/output device 3006 using a title, etc.

The portable storage medium 3009 can be a storage medium distributable as storing the above mentioned program such as a CD of various standards, a DVD of various standards, other optical disks, floppy disks, memory cards, etc. which can be distributed as storing the program.

The external storage device 3004 can be-a hard disk device, an optical disk device, a disk array, 83 modem a etc. The communications interface 3007 can be a ? terminal adapter, a LAN adapter, a router, a radio transmission/reception device, etc.

The network 4000 can be a public line, Internet, Intranet, Extranet, a radio circuit, a satellite circuit, etc. Therefore, the above mentioned program can also be downloaded into the computer 3000 from an application server, etc. through Internet. In addition, the program can be remotely executed by an information processing device of the information provider 5000. In this case, information about the concept map 80, a list of retrieval results of the -primary information 40, etc. is transmitted from the information processing device on the information provider side 5000 through the communications interface 3007 to be displayed on the display device of the input/output device 3006.

As described above, keywords highly relevant to a user-input keyword are extracted from an associative dictionary generated from the contents (primary information) of text information (a document database, a Web document, etc.), and a concept map on which the extracted keywords are arranged is presented to the user: When the user specifies the keyword on the concept map, 84 which is the secondary information, the original text of the primary information is retrieved, and a list and the text of the primary information highly relevant to the specified keyword are presented to the user.

Therefore, a group of keywords indicating the features and tendency of all or a part of the contents of text information relevant to a user- input keyword may be presented to the user as secondary information using a concept map. When the user selects a keyword from the concept map, a list of primary information containing the selected keyword and the original text of the primary information can be presented to the user.

Therefore, the user can detect and access target primary information by inputting any keyword even if the user cannot find a keyword appropriate for the retrieval, or cannot generate a correct keyword or keyword-based queries. Furthermore, the user can access the primary information of which the user has not been aware (that is, the user can be notified of useful primary information).

Claims (1)

1. CLAIMS:

   A text information access apparatus, comprising: an associative dictionary generation means (50) for generating an associative dictionary (60) by performing an associative analysis according to primary information (40) about text; a map generation means (70) for performing an associative retrieval process according to a f irst keyword input by a user, and visualizing associative information relevant to the first keyword as a concept map (80); and a retrieval means (90) for retrieving the primary information according to a second keyword specified is by the user on the concept map, and presenting the primary information (40) relevant to the second keyword.

   2. The apparatus according to claim 1, wherein: said first keyword is used by the user as a focus of a feature or a tendency of information; and said map generation means presents the concept map on which the keyword input as a focus, and keywords highly relevant to the keyword are arranged.

   86 3. The apparatus according to claim 2, wherein said first keyword can contain a keyword functioning as an aspect limiting a type of keyword highly relevant to the keyword as a focus.

   4. The apparatus according to claim 1, wherein said first keyword can contain a keyword which is a key concept indicating an item that the user demands to know, a keyword which is a focus indicating an attribute of the item, and a keyword which is an aspect limiting a type of keyword highly relevant to the keyword which is a focus.

   5. The apparatus according to claim 3 or 4, further comprising:

   a storage means (12) for preliminarily storing a set of the keyword which is a focus and the keyword which is an aspect; a menu display means (23) for displaying as a menu the set of the keyword which is a focus and the keyword which is an aspect stored in said storage means; and a selection means for selecting a set of a keyword which is a user-desired focus and a keyword which is an aspect from the menu.

   87 6. The apparatus according to any preceding claim, further comprising an automatic classification means (210) for automatically classifying the primary information, and generating a database (220) for each category, wherein said associative dictionary generation means (50) generates the associative dictionary based on the database for each category.

   7. The apparatus according to any preceding claim, further comprising an information extraction/ thesaurus generation means (310) for extracting information from the primary information, generating a thesaurus, and generating a dictionary (320) for each category, wherein said map generation means (70) generates the concept map by referring to the dictionary for each category.

   8. The apparatus according to any preceding claim, further comprising an automatic translation means (410) for automatically translating a keyword stored in the associative dictionary (60) in cooperation with the 88 associative dictionary generation means (50), wherein said map generation means (70) visualizes and displays a concept map on which a keyword translated by said automatic translation means (410) is arranged.

   9. The apparatus according to claim 1, further comprising:

   an automatic classification means (210) for automatically classifying the primary information, and generating a database (220) for each category; an information extraction/ thesaurus generation means (310) for extracting information from the primary information, generating a thesaurus, and generating a dictionary (320) for each category; and an automatic translation means (410). for automatically translating a keyword stored in the associative dictionary (60c) in cooperation with said associative dictionary generation means (50c), wherein said map generation means (70c) generates the concept map (80c) by referring to the associative dictionary (60c) for each category.

   10. The aRDa-ratus according to any preceding claim, rein: said retrieval means (90) presents a list-of the primary information before presenting entire text of 89 the primary information relevant to keyword; and a user selects desired primary information from the list.

   the second 11. The apparatus according to claim 10, further comprising:

   an automatic summary means (510) for automatically summarizing primary information s elected by the user, wherein summary text (520) generated by said automatic summary means (510) is presented before the entire text about the primary information is presented.

   12. The apparatus according to claim 11, further comprising:

   an automatic translation means (710) for automatically translating the summary text generated by said automatic summary means (510a), wherein said summary text is presented as translated text (720).

   13. The apparatus according to claim 12, wherein: said automatic translation means -(710a) automatically translates the list of the primary information selected by the user; and said list of the primary information is translated and presented.

   14. The apparatus according to claim 10, further comprising: an automatic translation means (810) for automatically translating the primary information selected by the user; and an automatic summary means (910) for generating summary text (910) of the primary information generated by said automatic translation means (810), wherein said summary text (920) generated by said automatic summary means (910) is presented before presenting the entire text about the primary information.

   15. The apparatus according to claim 14, wherein:

   said automatic translation means (810a) automatically translates the list of the primary information selected by the user; and said translated list of the primary information is and presented.

   91 16 The apparatus according to claim 1,f urther comprising:

   a first automatic translation means (610) for automatically translating primary information presented by said retrieval means, wherein translated text (620) of the primary information generated by said f irst automatic translation means (610) is presented.

   17. The apparatus according to claim 16, further comprising:

   a second automatic translation means (410) for automatically translating a keyword stored in the associative dictionary (60a) in cooperation with said associative dictionary generation means (50a), wherein said map generation means (70a) visualizes and displays a concept map (80a) on which keywords translated by said second automatic translation means are arranged.

   Ihe apparatus according to claim 16 or 17, ein:

   said automatic translation means (610a) automatically translates the list of the primary information selected by the user; and said list of the primary information is 92 translated and presented.

   19. The apparatus according to claim 9, further comprising:

   an automatic summary means (510d) for automatically summarizing the primary information selected by the user from the concept map; and an automatic translation means (610d) for translating the summary text (520) of the primary information generated by said automatic summary means (510d), wherein said translated summary text generated by said automatic translation means (610d) is presented before the entire text of the primary information is presented.

   20. The apparatus according to claim 9,further comprising:

   an automatic translation means (610e) for automatically translating the primary information selected by the user f rom the concept map (80e); and an automatic summary means (510e) for generating the summary text (520e) of the primary information generated by said automatic translation means (610e), wherein 93 said summary text generated by said automatic summary means (510e) is presented before the entire text of the primary information is presented.

   21. The apparatus according to any preceding claim, eln a first designation means designates at least one second keyword on the concept map (80).

   22. The apparatus according to my preceding claim, ein said concept map (80) is formed by second keywords (1001) and relation lines (1003) connecting the second keywords (1001), and further comprising:

   a.- designation means for designating at least one relation line (1003) on the concept map (80); and a determination means for detecting the relation line (1003) designated by the above designation means, and determining the second keyword used when said retrieval means retrieves the primary information from the relation line (1003).

   23. The apparatus according to claim 22, wherein:

   said determination means detects a designation position of the relation line (1003), and determines a weight of the second keyword based on position 94 information; and said retrieval means (90) retrieves the.primary information based on the weight for the second keyword determined by said determination means.

   24. A text information accessing method, comprising the steps of: generating an associative dictionary by an associative analysis from primary information of text (S21-S22); performing an associative retrieving process depending on a user-input first keyword, and visualizing associative information relevant to a first keyword as a concept map (S23); and retrieving the primary information based on a second keyword designated by the user on the concept map, and presenting the primary information relevant to the second keyword (S24-S28).

   25. The method according to claim 24, further comprising the steps of:

   preliminarily storing a set of a keyword which is a focus and the keyword which is an aspect; displaying as a menu the stored set of a keyword which is a focus and a keyword which is an aspect; and selecting a set of a keyword which is a userdesired focus and a keyword which is an aspect from the menu.

   26. The method according to claim 24 or 25, further comprising the steps of:

   automatically classifying the primary information, and generating a database for each category (S101,S102), wherein said associative dictionary is generated based on the database for each category.

   27. The method ac cording to claim 24, 25 or 26, further comprising the steps of:

   extracting information from the primary information, generating a thesaurus, and generating a dictionary for each category (Slll,Sll2), wherein said concept map is generated by referring to the dictionary for each category.

   28. The method according to any of claims 24 to 27, further comprising the steps of: automatically translating a keyword stored in the associative dictionary in cooperation wl-th generating 96 an associative dictionary (S121,S122), wherein said concept map on which a translated keyword is arranged is visualized and displayed.

   29. The method according to claim 24, further comprising the steps of:

   automatically classifying the primary information, and generating a database for each category (S101); extracting information from the primary information, generating a thesaurus, and generating a dictionary for each category (S111); and automatically translating a keyword stored in the associative dictionary (S121), wherein said concept map is generated by referring to the dictionary for each category.

   30. The method according to any of claims 24 to 29, further comprising the steps of:

   presenting alist of the primary information before presenting entire text of the primary information relevant to the second keyword (S26), wherein a user selects desired primary information from the list (S27).

   97 31. The method according to claim 30, further comprising the steps of: automatically summarizing primary information selected by the user (S141), wherein summary text generated in said automatic summarizing step is presented before entire text about the primary information is presented.

   32. The method according to claim 31, further comprising the steps of:

   automatically translating the summary text generated in said automatic summarizing (S171), wherein said summary text is presented as translated text (S172).

   33. The method according to claim 32, wherein said list of the primary information selected by the user is automatically translated (S181), and the list of the primary information after translation is presented (S182).

   34. The method according to claim 30, further comprising the steps of: automatically translating the primary information 98 selected by the user (S181); generating summary text of the primary information generated in said automatic translating (S141); and presenting said summary text generated in said automatically summarizing before presenting the entire text about the primary information (S172).

   35. The method according to claim 34, wherein the list of the primary information selected by the user is automatically translated,and the list of the primary information be translated is presented (S181,S182,S183).

   36 The method according to claim 35, further comprising the steps of:

   automatically translating primary information presented (S215); and presenting translated text of the primary information generated in said automatic translating (S216).

   37. The method according to claim 34, further comprising the steps of: automatically translating a keyword stored in the 99 associative dictionary (S121-S122), wherein said concept map on which keywords translated in said automatic translating are arranged is visualized and displayed.

   38. The method according to 37, further comprising the steps of: automatically translating the list of the primary information selected by the user (S201); and presenting the list of the primary information be translated (S202).

   39. The method according to claim 29, further comprising the steps of: automatically summarizing the primary information selected by the user from the concept map (S141); translating the summary text of the primary information generated in said automatically summarizing (S171); and 20 presenting said summary text before the entire text of the primary information is presented (S172).

   40. The method according to claim 29, further comprising the steps of:

   automatically translating the primary information -100selected by the user from the concept map (S181); generating the summary text of the primary information generated in said automatically translating (S141); and presenting said summary text before the entire text of the primary information is presented (S172).

   41. The method according to any of claims 24 to 40, further comprising the step of: a first designating step of designating at least one second keyword on the concept map.

   The method according to any of claims 24 to 41, wherein said concept map is formed by second keywords and relation lines connecting the second keywords, further comprising: a designating step of designating at least one relation line on the concept map; and detecting the designated relation line, and determining the second keyword used when the primary information is retrieved.

   43. The method according to claim 42, further comprising the steps of: detecting a designation position of the relation line, and determining a weight of the second keyword -101determined in the associated designating step based on the position, wherein the primary information is retrieved based on the weight determined for the second keyword determined.

   44. The method according to claim 24, wherein said concept map is formed by second keywords and relation lines connecting the second keywords, further comprising: a first designating step of designating at least one second keyword on the concept map; a second designating step of designating at least one relation line on the concept map; and detecting the designated relation line, and determining the second keyword used when the primary information is retrieved, wherein said primary information is retrieved based on th second keyword designated in said first designating step and the second keyword determined.

   45. The method according to claim 44, further comprising the steps of: detecting a designation position of the relation line, and determining a weight of the second keyword determined in said second designating step based on the position, wherein the primary information is retrieved based on the -102- weight determined for the second keyword determined.

   is 46. A computer-readable storage medium storing a program used to direct a computer to perform a process comprising the steps of: generating an associative dictionary by an associative analysis from primary information of text; performing an associative retrieving process depending on a user-input first keyword, and visualizing associative information relevant to the first keyword as a concept map; and retrieving the primary information based on a second keyword designated by the user on the concept map, and presenting the primary information relevant to the second keyword.

   47. The computer-readable storage medium according to claim 46, wherein said process further comprises: a first designating step of designating at least one second keyword on the concept map.

   48. A computer-readable storage medium storing a program used to direct a computer to perform a process comprising the steps of: generating an associative dictionary by an associative analysis from primary information of text; performing an associative retrieving process -103- depending on a user-input first keyword, and visualizing associative information relevant to the first keyword as a concept map; designating at least one relation line on the concept map; retrieving the designated relation line, and determining based on the relation line a second keyword on the concept map used when the primary information is retrieved; and retrieving the primary information based on the determined second keyword, and presenting the primary information relevant to the second keyword.

   49. A computer-readable storage medium storing a program used to direct a computer to perform a process comprising the steps of: generating an associative dictionary by an associative analysis from primary information of text; performing an associative retrieving process depending on a user-input first keyword, and visualizing associative information relevant to the first keyword as a concept map; a first designating step of designating at least one second keyword on the concept map; a second designating step of designating at least one relation line on the concept map; retrieving the designated relation line, and -104determining based on the designated relation line the second keyword on the concept map used when the primary information is retrieved; and retrieving the primary information based on the second keyword designated in said first designating step and the determined second keyword, and presenting the primary information relevant to the second keyword.

   50. The computer-readable storage medium according to claim 49, wherein said process further comprises: detecting a position of the designated relation line, and determining a weight of the determined second keyword based on the position; and retrieving the primary information based in addition on the weight determined for the determined second keyword.

   51. A text information access apparatus according to claim 1 and substantially according to any one of the embodiments thereof hereinbefore described with reference to the accompanying drawings.

   52. A text information accessing method according to claim 24 and substantially according to any one of the examples thereof hereinbefore described with reference to the accompanying drawings.

   -10553. A computer-readable storage medium according to claim 46, 48 or 49, and storing a program used to direct a computer to perform a process according to any one of the examples thereof hereinbefore described with reference to the accompanying drawings.