JP2007257420A - Data map creation server, and method and program for creating data map - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To clearly visualize a thesaurus by formulating relations of characteristic words extracted from a document set. <P>SOLUTION: Word data with high frequency is extracted on the basis of the frequency of word data included in a plurality of pieces of document data extracted on the basis of a search keyword, directly higher or lower relations of the word data are determined on the basis of the relations of document sets including the word data, and when the directly higher or lower relations are established among the word data, the word data are displayed. When the relations among the word data display 51 are directly higher or lower, a connection line 52 connects the word data display 51 to display the word data display 51. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention relates to a data map creation server, a data map creation method, and a data map creation program for creating a data map that expresses a relationship between feature words appearing in a plurality of document data.

  In recent years, with the spread of the Internet, a large amount of information can be easily obtained, and a technique for narrowing down search results more efficiently, re-searching, or compiling search results in an easy-to-understand manner is required.

  As an example of an information search method, a search query expansion technique that presents an optimal keyword expressing a search request is known (Non-Patent Documents 1 and 2). In addition, there is a technology that increases knowledge on the search system side by acquiring a glossary of technical terms and a thesaurus for a specific field instead of corresponding to a search for each case (Non-Patent Document 3, 4, 5).

  A search system using such a technology is automatically used especially in fields where technical terms are not available in the first place, such as new fields, or in areas where a large amount of new documents are created every day, such as reputation information such as Blog. A technique for retrieving these new documents in association with each other is essential. Further, even in established fields, a common thesaurus or the like is required for knowledge sharing.

  In other words, clustering, summarization, and visualization are necessary to help understand search results. In addition to displaying the mutual relationship, it is important to assign an appropriate keyword representing the contents of the search result document or document set.

Hironori Tohi, Masaki Aono, Search query expansion method based on bi-clustering, IEICE 2nd Web Intelligence Study Group, IEICE SIG NOTES WI2-2005-18, 2005. Shinji Shinzato, Kentaro Torizawa, automatic acquisition of upper-lower relations between words from HTML documents, natural language processing, Vol.12, No.1, pp125-150, 2005 Jun Fujii, Tetsuya Ishikawa, Extraction and organization of dictionary knowledge information using the World Wide Web, IEICE Transactions, VOL. J85-D2, PP. 300-307,2002 Padiomini Srinivassan, Thesaurus Construction, in Information Retrieval: Data Structures and Algorithms, Edited by William B. Frakes and Ricarbo Baeza-yates, Prentice-Hall, 1992 Junji Umemura, Development of an automated construction system for unexplored thesaurus, Information Processing Promotion Society 2001 Report Collection, 2001

  However, in the conventional technique described above, the thesaurus construction has not been sufficiently formulated yet. Therefore, the relationship between the characteristic words extracted from the document set is not sufficiently visualized and does not function as a knowledge sharing thesaurus.

  An object of the present invention is to solve the above-described problems, formulate a relationship between characteristic words extracted from a document set, and clearly visualize the thesaurus.

  The data map creation server of the present invention is a data map creation server that creates a data map that maps a search keyword and related words extracted from a plurality of document data based on the search keyword, and stores the plurality of document data. Data storage means, related word extraction means for extracting one or more related words related to the search keyword from a plurality of document data stored in the data storage means, and the related word extraction means About the upper and lower relationship determining means for determining the upper and lower relationship of each map posting word according to a predetermined upper and lower relationship determining condition, with the extracted related word and the search keyword as a map posting word, and each map posting word When there are multiple map posting words An upper word determining means for determining an upper word that is a map posting word that is directly above, a word display area for displaying each of the map posting words, each map posting word, and a word immediately above each map posting word A data map creating means for creating a data map including a line segment display area for displaying line segments connected to each other, and a data map for transmitting the data map created by the data map creating means via a communication network And a transmission means.

  With the configuration described above, based on the number of frequent data included in a large number of documents extracted from the search request data, the upper and lower relations immediately above or directly below the document data related to the frequent data are displayed. Since the data map is created, it is possible to formulate the relationship between the characteristic words extracted from the document set and visualize the thesaurus more clearly.

  You may provide the search request receiving means which receives the search request which showed the search keyword via a communication network.

  You may provide the search request reception means which receives the search request which presented or designated the document data used as search object via a communication network.

  For example, for the first map posting word and the second map posting word to be determined, the upper / lower relationship determination condition is that the number of document data including the first map posting word is greater than the number of document data including the second map posting word. In many cases, when the first map posting word is included in the document data including the second map posting word in a predetermined ratio or more, the first map posting word is in a superior relationship with the second map posting word. It is stipulated to be decided. The predetermined ratio is, for example, 50%.

  The upper word determining means is, for example, only when the first map posting word is higher than the second map posting word and there is no upper word of the second map posting word with the first map posting word being higher. The map posting word is determined to be a word directly above the second map posting word.

  The word display area is an area for displaying, for example, a map posting word and the number of document data including the map posting word. With such a configuration, the data map can be displayed visually for the user.

  The data map creation method of the present invention is a data map creation method for creating a data map in which a search keyword and related words extracted from a plurality of document data are mapped based on the search keyword, and storing the plurality of document data. Extracting one or more related words related to the search keyword from a plurality of document data stored in the data storage means, and extracting the extracted related word and the search keyword as a map posting word, The step of determining the upper / lower relationship of each map posting word according to the predetermined upper / lower relationship determination condition, and when there are a plurality of map posting words that are upper relationships for each map posting word, Step to determine the top word that is the map posting word A data map including a word display area for displaying each map posting word and a line display area for displaying a line connecting each map posting word and the word immediately above each map posting word And a step of performing.

  A data map creation program according to the present invention is a data map creation program for creating a data map in which a search keyword and related words extracted from a plurality of document data are mapped based on the search keyword, wherein the computer includes the plurality of documents. A step of extracting one or more related words related to the search keyword from a plurality of document data stored in a data storage means for storing data; and the extracted related word and the search keyword are displayed in a map And determining the upper / lower relationship of each map posting word according to the predetermined upper / lower relationship determination conditions, and when there are multiple map posting words that have a higher relationship for each map posting word, With the map posting word Determining a word directly above, a word display area for displaying each map posting word, and a line segment display area for displaying a line connecting each map posting word and the word immediately above each map posting word And a step of creating a data map including:

  According to the present invention, the relationship between characteristic words extracted from a document set can be formulated, and the thesaurus can be clearly visualized.

Hereinafter, an embodiment of the present invention will be described with reference to the drawings.
FIG. 1 is a block diagram showing an example of the configuration of a data map creation system 100 according to an embodiment of the present invention.

  As shown in FIG. 1, the data map creation system 100 of this example includes a data map creation server 10 and a plurality of user terminals 30A to 30N. The data map creation server 10 and the plurality of user terminals 30A to 30N are each connected to a communication network 40 such as the Internet.

  The data map creation server 10 is configured by an information processing apparatus such as a WWW (World Wide Web) server or a workstation server, and is managed by an administrator of the system 100, for example.

  Each of the user terminals 30 </ b> A to 30 </ b> N is configured by an information processing apparatus such as a personal computer, and is managed by a user who uses the system 100. Each of the user terminals 30 </ b> A to 30 </ b> N is equipped with, for example, a browser and has various functions such as a function of transmitting / receiving various information via the communication network 40 and a function of displaying various information received via the communication network 40. Have.

  FIG. 2 is a block diagram illustrating an example of the configuration of the data map creation server 10. As shown in FIG. 2, the data map creation server 10 includes a data input unit 11, a data display unit 12, a data communication unit 13, a search request information receiving unit 14, a document data acquisition unit 15, and a frequent word extraction. It includes an arrangement unit 16, an upper / lower relationship determination unit 17, a data map creation unit 18, a document data DB 19, and a map creation / use DB 20.

  The data input unit 11 is configured by an information input device such as a keyboard and a mouse, for example, and is used when, for example, an administrator of the server 10 inputs various information. The data display unit 12 is configured by a display device such as an LCD (Liquid Crystal Display), and is used for displaying various information.

  The data communication unit 13 has a function of transmitting and receiving various types of information to and from the user terminals 30A to 30N via the communication network 40, for example.

  The search request information receiving unit 14 receives search keyword information (hereinafter simply referred to as “keyword”) search request information from the user terminals 30 </ b> A to 30 </ b> N and receives a search request (data map creation request). Has the function to execute.

The document data acquisition unit 15 extracts document data including a keyword from a plurality of document data to be searched according to the search request information received by the search request information receiving unit 14.
Here, the “keyword” is an element constituting the document data, and specifically includes, for example, a word, a idiom, a number, a mathematical expression, a symbol, and the like.

  The frequent word extraction / arrangement unit 16 extracts frequent words that appear in the extracted document data with high frequency as related words related to the keyword, and arranges the related words in the order of high frequency. “Frequently” means that the appearance frequency in the entire document data is high, or that the appearance frequency is high in the same sentence or paragraph as the keyword in the document data.

  The upper / lower relationship determination unit 17 uses the related words and keywords extracted by the frequent word extraction / arrangement unit 16 as map posting words, and determines the upper / lower relationship of each map posting word according to predetermined upper / lower relationship determination conditions. To do. In addition, when there are a plurality of map posting words that are in a higher-order relationship for each map posting word, the upper-lower relationship determination unit 17 determines the upper word that is the map posting word that is the upper-level relationship.

  The upper / lower relationship determining unit 17 is the first only when the first map posting word is higher than the second map posting word and there is no upper word of the second map posting word with the first map posting word being higher. The map posting word is determined to be a word directly above the second map posting word. For example, the upper relationship of the map posting word A is the map posting word B, C, D, the upper relationship of the map posting word B is the map posting word D, and the upper relationship of the map posting word C is the map posting word D. Then, the map posting words B and C become the words directly above the map posting word A.

  The data map creation unit 18 includes a word display area for displaying each map posting word, and a line segment display area for displaying a line segment connecting each map posting word and the word immediately above each map posting word. Create data map information. In other words, a process to create data map information for displaying a data map in which multiple map posting words (concepts that include both keywords and extracted related words) are arranged based on upper and lower relationships or directly above relationships is executed. Has the function of The data map information means electronic data of a data map displayed on the display screen of the display device, and may be simply referred to as “data map”.

  The document data DB 19 is constituted by a database device, for example, and stores various information such as a plurality of document data to be searched.

  The map creation / use data DB 20 is configured by a database device, for example, and stores various information such as data (map creation / use data) used when creating a data map.

  FIG. 3 is an explanatory diagram for explaining the storage state of the document data in the document data DB 19. As shown in FIG. 3, the document data DB 19 stores a plurality of document data 1 to N (N: any positive integer). The document data is created by a description language such as XML, for example. The unit of document data used for processing may be defined in any way, such as a complete document, each page of the document, each sentence of the document, each line of the document, or a combination thereof. .

Next, the operation of the data map creation system 100 of this example will be described.
FIG. 4 is a flowchart showing an example of search processing by the data map creation server 10 in the data map creation system 100 of this example.

  In this example, it is assumed that each document data of documents 1 to N is registered in the document data DB 19 (see FIG. 3). Here, it is assumed that a process for creating data map information is executed in response to a request from the user terminal 30A managed by the user A.

  In the data map creation process, first, the search request information receiving unit 14 of the data map creation server 10 receives the search request information from the user terminal 30A via the communication network 40, and receives the data map creation request (step S101).

  The search request information includes a search keyword. The search keyword may be in any form as long as it is a keyword for narrowing search, such as one or more words, idioms, symbols, expressions, and sentences.

  Here, it is assumed that the search request information receiving unit 14 receives “enzyme” and “activity” as search keywords in step S101.

  When the search request information receiving unit 14 receives the search request information, the document data acquiring unit 15 uses a search keyword “enzyme” and “activity” from a plurality of document data registered in the document data DB 19. Document data satisfying a predetermined search condition is extracted, and processing for narrowing down document data to be searched for related words is performed (step S102). In this example, it is assumed that document data composed mainly of terms and sentences related to biochemistry is stored in the document data DB 19.

  The search condition in step S102 includes, for example, that part or all of the search keyword is included in the document data, and the document constituent element that is the same as the search keyword or a similar document constituent element (for example, a synonym or related word) This is a condition used for a normal search engine, such as being included in

  Next, frequently used words that frequently appear in a plurality of document data including the keyword “enzyme” or “activity” extracted in step S102 (for example, more than half of the document data, more than 20% of the document data) are used as keywords. Extracted as related words (step S103).

  When the related word is extracted, the data map creating unit 18 creates data map information using the extracted related word and keyword, that is, using the map posting word (step S104).

  When the data map information is created by the data map creation unit 18, the communication unit 13 of the data map creation server 10 transmits the data map information to the user terminal 30A via the communication network 40 (step S105).

  When receiving the data map information from the data map creation server 10, the user terminal 30A displays a data map (see, for example, FIG. 10) on the display screen of the display device provided in the user terminal 30A.

  Next, a specific example of step S104 (data map creation processing) will be described with reference to FIG. The data map creation process is executed, for example, according to the flowchart shown in FIG. Here, it is assumed that “enzyme” and “activity” are designated as keywords, and each word shown in the table of FIG. 6 such as “activation” and “amino” is retrieved as related words.

  In the data map creation process, first, the frequent word extraction / arrangement unit 16 arranges map posting words such as keywords and extracted related words in the vertical and horizontal directions in the descending order of frequency included in a plurality of document data. A column table is created (step S201).

  In step S201, for example, as shown in FIG. 6, the related words “amino (95)” and “catalyst” extracted when the keywords “enzyme (309)” and “activity (465)” are searched. (70) "," in vivo (55) "," bioactive substance (72) "," activation (136) "," amino acid (93) "," nucleic acid (76) "," fatty acid (50) " , “Saccharide (66)” is created in the order of the number of document data shown in parentheses in the order of the number of document data, and arranged in rows and columns as item columns. In FIG. 6, the number in parentheses attached to each map posting word is the number of document data containing the corresponding map posting word among the document data narrowed down in step S102 described above. Yes, the number counted in step S102 (for keywords) or step S103 (for related words).

  Next, the upper / lower relationship determination unit 17 determines the upper / lower relationship of the map posting word (step S202). Here, the number of document data (document frequency) in “D” in which “w” appears when “D” is a set of document data and “w” is a word is represented by df (w, D). And Also, the number of document data in “D” in which both two words “u” and “v” appear is represented by df (u * v, D). In such a case, when the word “u” and the word “v” satisfy the relationship of the following expressions “Equation 1” and “Equation 2”, “the word“ u ”is the word“ v ”from the viewpoint of document frequency”. ”

  The fact that the word “u” and the word “v” satisfy the above-described expressions “Equation 1” and “Equation 2” indicates that the number of document data in which the word “u” appears is the word “v” as shown in FIG. ”Is larger than the number of document data in which“ ”appears (formula“ Formula 2 ”), and the word“ u ”appears in the majority of document data in which the word“ v ”appears. The right side of the expression “Equation 1” is not limited to 0.5, and may be another value such as 0.6 or 0.7, and is a constant between 0.0 and 1.0. I just need it. In particular, if the right side of the expression “Equation 1” is 1.0, the relationship “the word“ v ”always appears in a document in which the word“ u ”appears” is established. Further, the right side of the expression “Equation 1” may be a constant determined by the word “u” or the word “v”.

  In step S202, the upper / lower relationship determination unit 17 determines the upper / lower relationship for all sets of each map posting word in accordance with the above-described formulas "Equation 1" and "Equation 2". In the arrangement table shown in FIG. 6, when the map posting word arranged vertically in the row item is higher than the map posting word arranged right and left in the column item, the cell specified by the row and column If “1” is registered in the cell, and if it is not (when the upper-lower relationship is reversed or not in the upper-lower relationship), “0” is registered in the cell specified by the row and column, The sequence table is updated, for example, as shown in FIG. That is, the upper / lower relationship determination unit 17 performs a process of determining the upper / lower relationship for each map posting word and reflecting the determination result in the arrangement table.

  In the example shown in FIG. 8, since “1” is registered in the cell in which the row item is “enzyme” and the column item is “amino”, “enzyme” has a higher relationship than “amino”. Yes, “amino” means subordinate to “enzyme”. In addition, since “0” is registered in the cell in which the row item is “enzyme” and the column item is “bioactive substance”, “enzyme” does not have a higher relationship than “bioactive substance” Means.

  Next, when there are a plurality of map posting words that are upper-level relationships for each map posting word, the higher-level / lower-level relationship determining unit 17 determines the immediately above word that is the map posting word that is the upper-level relationship (step S203). The immediately above word can be obtained as a minimum element of a set of map posting words that are higher than a predetermined map posting word.

  Specifically, when the upper / lower relationship determination unit 17 has an upper / lower relationship as illustrated in FIG. 8, for example, when focusing on “amino acid” that is a map posting word in which a plurality of map posting words of the upper relationship exist, The top word is determined as follows. The upper / lower relationship determination unit 17 confirms that “amino” is higher than “amino acid” and that there is no upper word of “amino acid” with “amino” being higher, and “amino” is changed to “amino acid”. Determine the word directly above. Further, the upper / lower relationship determination unit 17 confirms that “enzyme” is higher than “amino acid” and that there is a higher word (“amino”) of “amino acid” with “enzyme” higher. Do not determine “enzyme” as the word immediately above “amino acid”. If there are a plurality of words directly above, all of them are determined as words directly above. Then, “1” registered in the cell corresponding to the map posting word that has not been directly above is updated to “0”, and the array table is updated as shown in FIG. 9, for example. In FIG. 9, an underbar is added to the cell updated from “1” to “0” in step S203.

  Next, in step S203, the data map creation unit 18 includes a map posting word that is a row item that is output as "1" in a cell in the matrix-like array table, and a map posting word that is a column item. A data map indicating the relationship between the map posting words is generated based on the upper and lower relationship (step S204). In this case, the data map is created by, for example, drawing the line segments sequentially from the map publication word directly below to the map publication word directly above, thereby connecting the map publication words with the line segments, and finally the map publication word. Is generated by adjusting the display position.

  For example, as shown in FIG. 10, the data map includes a word data display 51 that displays the name of the map posting word and the number of document data including the map posting word, and each word data display 51 that is directly above and below. And a connecting line 52 connecting the two. Here, for example, in the case of “in vivo”, there are two word data “enzyme” and “activity” which are directly related to each other.

  11 to 13 are explanatory diagrams showing examples of other data maps. The example shown in FIG. 11 is an example of a data map obtained when a predetermined document data group is prepared and “enzyme” is designated as a search keyword. The example shown in FIG. 12 is an example of a data map obtained when a predetermined document data group is prepared and “protein” is designated as a search keyword. The example shown in FIG. 13 is an example of a data map obtained when a predetermined document data group is prepared and “life”, “cell”, and “blood” are designated as search keywords.

  As shown in FIG. 13, in the data map, for example, word data (map posting words) branched from upper to lower may be displayed connected with lower word data. For example, “kouzagai” and “plants” cannot be found at first glance, but there is a “course surgery eye” that is commonly connected below them, so there is a certain relationship between them. Can be grasped visually.

  As described above, the data map information is created by the data map creation server 10, and the data map based on the data map information is displayed on the display screen of the display device of the user terminal 30A.

  As described above, in the above-described embodiment, the immediate relationship is determined based on the appearance frequency of the word data included in the plurality of document data extracted based on the search keyword. Since the data map in which the word data displays 51 are connected by the connecting line 52 is created, the relationship between the characteristic words extracted from the document set can be formulated, and the thesaurus can be clearly visualized. That is, it is possible to clearly grasp the relevance of each word data mapped on the map by looking at the data map in which the word data display 51 in which the immediate relationship is established is connected by the connecting line 52. It becomes.

  Specifically, for example, since the relationship immediately above “lipid 124” in the center of FIG. 11 is “nucleic acid”, “lipid” can be said to be a word that appears in the topic about “nucleic acid”. Further, the upper relationship of “lipid” is “protein”, and the upper relationship of “protein” is “enzyme”. Therefore, it can be seen that “lipid” is part of the topic of “protein”, and “protein” is part of the topic of “enzyme”.

  In addition, since the data map is created by connecting the word data display 51 in which the directly above relationship is established by the connecting line 52, a line segment is drawn between all the word data in the upper and lower relations. Therefore, the relationship between the word data is easy to see without complicating.

  Although not particularly mentioned in the above-described embodiment, each process described above executed by the server 10 is performed according to a program (data map creation program) provided in the server 10.

  In the embodiment described above, in the embodiment of the invention, the search process is performed when there is a search request that presents a search keyword. However, when there is a search request that presents or specifies document data, it is specified in advance. The search keyword may be used, and the search may be performed using the presented or designated document data as a search database. In this case, for example, the document data received via the communication network 40 at the time of a search request from the user terminal 30A, or the document data designated from the user terminal 30A among the document data stored in the document data DB 19, for example, is searched. The search process is executed as follows.

  According to the present invention, a data map capable of expressing the relationship of feature words appearing in document data such as patent document data, researcher data, accommodation facility data, electronic medical record, case data, company data, and product data is created. This is effective for adapting to a data map creation server or a data map creation system.

It is a block diagram which shows the structural example of the data map creation system in one embodiment of this invention. It is a block diagram which shows the structural example of a data map creation server. It is explanatory drawing for demonstrating the storage state of the document data in document data DB. It is a flowchart which shows the example of a search process. It is a flowchart which shows the detail of a data map creation process. It is a figure which shows the matrix used for the determination of a high-order subordinate relationship. It is a conceptual diagram of a high-order subordinate relationship. It is a figure which shows the example of the matrix used for determination of a high-order subordinate relationship. It is a figure which shows the example of the matrix used for determination of a high-order subordinate relationship. It is explanatory drawing which shows the example of the data map based on the word data used for the matrix figure shown in FIG. It is explanatory drawing which shows the example of the data map based on other word data. It is explanatory drawing which shows the example of the data map based on other word data. It is explanatory drawing which shows the example of the data map based on other word data.

Explanation of symbols

DESCRIPTION OF SYMBOLS 10 Data map creation server 11 Data input part 12 Data display part 13 Data communication part 14 Search request information reception part 15 Document data acquisition part 16 Frequent word extraction arrangement | positioning part 17 High-order subordinate relation determination part 18 Data map creation part 19 Document data DB
20 Map creation data DB
30A, 30B, ..., 30N User terminal 40 Communication network 100 Data map creation system

Claims (17)

A data map creation server that creates a data map that maps a search keyword and related words extracted from a plurality of document data based on the search keyword,
Data storage means for storing the plurality of document data;
Related word extracting means for extracting one or more related words related to the search keyword from a plurality of document data stored in the data storage means;
Upper and lower relationship determining means for determining the upper and lower relationship of each map posting word according to a predetermined upper and lower relationship determining condition, with the related word extracted by the related word extracting means and the search keyword as a map posting word; ,
When there are a plurality of map posting words that are upper-level relationships for each map posting word, a directly-upward word deciding means that determines the upper word that is the map posting word that is the upper-level relationship directly above,
Data for creating a data map including a word display area for displaying each map posting word and a line segment display area for displaying a line segment connecting each map posting word and the word immediately above each map posting word Map creation means;
A data map creation server comprising: a data map transmission unit configured to transmit the data map created by the data map creation unit via a communication network. The data map creation server according to claim 1, further comprising search request receiving means for receiving a search request presenting a search keyword via a communication network. The data map creation server according to claim 1, further comprising a search request reception unit that receives a search request that presents or specifies document data to be searched via a communication network. The upper / lower relationship determination condition is that, for the first map posting word and the second map posting word to be determined, the number of document data including the first map posting word is greater than the number of document data including the second map posting word, When the first map posting word is included in the document data including the second map posting word in a predetermined ratio or more, it is determined that the first map posting word is in a higher level relationship with the second map posting word. The data map creation server according to any one of claims 1 to 3, wherein the data map creation server is defined as follows. The data map creation server according to claim 4, wherein the predetermined ratio is 50%. The upper word determining means is the first map posting word only when the first map posting word is higher than the second map posting word and there is no higher word of the second map posting word that is higher than the first map posting word. 6. The data map creation server according to claim 4, wherein one map posting word is determined to be a word directly above the second map posting word. The data map creation server according to any one of claims 1 to 6, wherein each of the word display areas is an area for displaying a map posting word and a number of document data including the map posting word. A data map creation method for creating a data map that maps a search keyword and related words extracted from a plurality of document data based on the search keyword,
Extracting one or more related words related to the search keyword from a plurality of document data stored in a data storage means for storing the plurality of document data;
The extracted related word and the search keyword as a map posting word, and according to a predetermined upper and lower relationship determination condition, determining the upper and lower relationship of each map posting word,
Determining a top word that is a map posting word that is directly above when there is a plurality of top ranking map words for each map posting word;
Creating a data map including a word display area for displaying each map posting word, and a line segment display area for displaying a line segment connecting each map posting word and the word immediately above each map posting word And a data map creation method characterized by comprising: The data map creation method according to claim 8, further comprising a step of receiving a search request presenting a search keyword. The data map creation method according to claim 8, further comprising a step of receiving, via a communication network, a search request that presents or designates document data to be searched. The upper / lower relationship determination condition is that, for the first map posting word and the second map posting word to be determined, the number of document data including the first map posting word is greater than the number of document data including the second map posting word, When the first map posting word is included in the document data including the second map posting word in a predetermined ratio or more, it is determined that the first map posting word is in a higher level relationship with the second map posting word. The data map creation method according to any one of claims 8 to 10, wherein the data map creation method is defined as follows. In the step of determining the word immediately above, only when the first map posting word is higher than the second map posting word and there is no upper word of the second map posting word with the first map posting word being higher. The data map creation method according to claim 11, wherein the first map posting word is determined to be a word immediately above which is a superior relationship directly above the second map posting word. A data map creation program for creating a data map that maps a search keyword and related words extracted from a plurality of document data based on the search keyword,
On the computer,
Extracting one or more related words related to the search keyword from a plurality of document data stored in a data storage means for storing the plurality of document data;
The extracted related word and the search keyword as a map posting word, and according to a predetermined upper and lower relationship determination condition, determining the upper and lower relationship of each map posting word,
Determining a top word that is a map posting word that is directly above when there is a plurality of top ranking map words for each map posting word;
Creating a data map including a word display area for displaying each map posting word, and a line segment display area for displaying a line segment connecting each map posting word and the word immediately above each map posting word A data map creation program to execute. On the computer,
The data map creation program according to claim 13, further comprising the step of receiving a search request presenting a search keyword. On the computer,
14. The data map creation program according to claim 13, further comprising: a step of receiving a search request presenting or specifying document data to be searched via a communication network. The upper / lower relationship determination condition is that, for the first map posting word and the second map posting word to be determined, the number of document data including the first map posting word is greater than the number of document data including the second map posting word, When the first map posting word is included in the document data including the second map posting word in a predetermined ratio or more, it is determined that the first map posting word is in a higher level relationship with the second map posting word. The data map creation program according to any one of claims 13 to 15, wherein the data map creation program is defined as follows. In the computer, in the step of determining the word immediately above, the first map posting word is higher than the second map posting word, and the upper word of the second map posting word is higher than the first map posting word. The data map creation program according to claim 16, wherein a process for determining the first map posting word as a word directly above the second map posting word is executed only when not.