KR102100346B1 - Apparatus and method for managing dataset - Google Patents

Abstract

According to one embodiment of the present invention, an apparatus for managing a dataset comprises: a meta data extraction unit extracting meta data from each of the plurality of datasets; a similarity calculation unit calculating similarity between the plurality of datasets based on degree of similarity between the meta data; an interlinking calculation unit calculating interlinking between the plurality of datasets based on whether each of the plurality of datasets refers each other; a semantic association calculation unit calculating semantic association between the plurality of datasets based on semantic information of each metadata; and a relation map generation unit generating a relation map displaying relation between the plurality of datasets based on at least one of the calculated similarity, the calculated interlinking, and the calculated semantic association. A researcher may view the plurality of datasets at a glance through the relation map.

Description

Data set management device and method {APPARATUS AND METHOD FOR MANAGING DATASET}

The present invention relates to a data set management apparatus and method.

With the development of the Internet, various types of data are being mass-produced, and technologies utilizing these data are also developing. For example, data mining that involves purifying or classifying data based on big data, or machine learning techniques that train modules to perform a predetermined function based on data, for example.

Here, in order to develop the technology for utilizing the above-described data, it is desirable that the researcher recognize what kind of data is secured at a glance. Furthermore, it is desirable not only to secure the type of data, but also to recognize at a glance how the various types of data relate to each other.

Accordingly, various types of data are collected and refined, and then technologies that enable listing or viewing at a glance are being developed or researched.

Korean Patent Publication No. 10-2005-0084866 (published on August 29, 2005)

The problem to be solved by the present invention is to propose a technique for managing a plurality of bundles (or datasets) of data having related topics.

For example, extracting what similarities, associations, and semantic associations between a plurality of these datasets are extracted from each other, and displaying the relationships between the plurality of datasets based on at least one of the extracted similarities, associations, or semantic associations Producing a relationship map and providing a technique provided to a researcher or the like can be included in the problem to be solved of the present invention.

However, the problem to be solved of the present invention is not limited to those mentioned above, and another problem not to be solved can be clearly understood by a person having ordinary knowledge to which the present invention belongs from the following description. will be.

The data set management apparatus according to an embodiment includes a meta data extraction unit for extracting metadata from each of a plurality of data sets, and a similarity calculation unit for calculating similarities between the plurality of data sets based on the similarity between the metadata. , Linkage calculation unit for calculating the degree of interlinking between the plurality of datasets based on whether each of the plurality of datasets refers to each other, and the plurality of datasets based on the semantic information of each of the metadata. A relationship between the plurality of datasets is displayed based on at least one of a semantic association calculation unit calculating a semantic association between datasets, and the calculated similarity, the calculated association degree, and the calculated semantic association degree. It includes a relationship map generating unit for generating a relationship map.

In addition, each of the plurality of data sets may include a plurality of data related to the subject.

In addition, the plurality of data includes content corresponding to each of at least two items, and the metadata extracting unit corresponds to at least a part of the at least two items and at least a part of the content corresponding to each of the at least two items, corresponding It can be extracted as metadata about the data set.

Further, the similarity between the metadata may be calculated by cosine similarity or jaccard similarity between vectors calculated for each of the metadata.

In addition, whether each of the plurality of data sets refers to each other may be determined as whether they are used together in the process of deriving a predetermined result.

In addition, the data set management apparatus further includes an interface unit that receives an input that is a basis for generating the relationship map among the calculated similarity, the calculated linkage, and the calculated semantic association, and the relationship map is calculated. The similarity, the calculated linkage, and the calculated semantic association may be generated based on the input through the interface unit.

In addition, the data set management apparatus further includes an interface unit that receives at least one of a first threshold for the similarity, a second threshold for the association, and a third threshold for the semantic association, and generates the relationship map. Based on at least one of the similarity of the first threshold or higher among the calculated similarities, the association of the second or higher thresholds among the calculated associations, and the semantic association of the third or higher thresholds among the calculated semantic associations, the You can create a relationship map.

In addition, the relationship map has a concentric circle structure in which a plurality of circles having different diameters form a concentric circle, one of the plurality of data sets is disposed at the center of the concentric circle, and the other is the center of the concentric circle among the plurality of circles. It can be placed on a circle rather than.

In addition, an interface unit for receiving the number of circles constituting the concentric circle may be further included.

In addition, the data set management apparatus further includes a keyword extracting unit for extracting keywords from each of the plurality of data sets, the similarity calculating unit calculates similarity between the keywords, and the linkage calculation unit calculates a linkage between the keywords The semantic association calculation unit calculates the semantic association between the keywords, and the relationship map generation unit displays the relationship between the keywords based on at least one of the similarity, linkage, and semantic association calculated between the keywords. You can create a relationship map.

The data set management method performed by the data set management apparatus according to an embodiment includes extracting metadata from each of a plurality of data sets and calculating similarities between the plurality of data sets based on the similarity between the metadata. Calculating a degree of interlinking between the plurality of data sets based on a step and whether each of the plurality of data sets refers to each other, and the plurality of data based on the semantic information of each of the meta data. Calculating a relationship map between the plurality of data sets based on at least one of the calculated similarity, the calculated association, and the calculated semantic association; Including the steps.

According to an embodiment, since a relationship map for a plurality of datasets can be provided to a researcher, etc., the researcher can view a plurality of datasets at a glance through such a relationship map.

1 is a system configuration diagram showing an overall configuration of a network in which a data set management apparatus according to an embodiment is operated.
2 is an exemplary conceptual diagram of a relationship map that can be provided by a data set management apparatus according to an embodiment.
3 is an exemplary configuration diagram of a configuration of a data set management apparatus according to an embodiment.
4 is an exemplary conceptual diagram of a screen displayed by a data set management apparatus according to an embodiment to a researcher and the like.
5 is a diagram illustrating an exemplary procedure for a data set management method according to an embodiment.

Advantages and features of the present invention, and methods for achieving them will be clarified with reference to embodiments described below in detail together with the accompanying drawings. However, the present invention is not limited to the embodiments disclosed below, but may be implemented in various different forms, and only the embodiments allow the disclosure of the present invention to be complete, and common knowledge in the art to which the present invention pertains It is provided to completely inform the person having the scope of the invention, and the present invention is only defined by the scope of the claims.

In describing embodiments of the present invention, when it is determined that a detailed description of known functions or configurations may unnecessarily obscure the subject matter of the present invention, the detailed description will be omitted. In addition, terms to be described later are terms defined in consideration of functions in an embodiment of the present invention, which may vary according to a user's or operator's intention or practice. Therefore, the definition should be made based on the contents throughout this specification.

1 is a system configuration diagram showing an entire configuration of a network in which the data set management apparatus 100 according to an embodiment is operated. Prior to the description of FIG. 1, the dataset management device 100 may be implemented in a device such as a personal computer, smart device, server, or cloud.

Referring to FIG. 1, the data set management apparatus 100 may be connected to various types of servers 200 and 210 through a network 300, and various types of data sets may be provided from these servers 200 and 210.

Here, the network 300 means a known network that can be implemented in a wired or wireless manner.

In addition, the various types of servers 200 and 210 described above may disclose their own datasets to the public, or only to subjects with specific authority. Hereinafter, a dataset disclosed to the public may be referred to as an open dataset, and a dataset disclosed only to a subject having specific authority, not to the public, will be referred to as an unopened dataset.

Meanwhile, a data set refers to a conceptual group or set containing a plurality of data having related topics. For example, one dataset may include data related to forest fires in 2018, and another dataset may include data regarding average height and weight of adolescents in the 2000s.

According to an embodiment of the present invention, the data set management management 100 obtains a plurality of data sets of various types from the servers 200 and 210, and similarities and linkages between the plurality of data sets will be described later. ) And semantic association (which will be described later on the semantic association) can be calculated, and based on at least one of the calculated values, a relationship map can be generated and then provided to various subjects. .

Here, the relationship map refers to the extent to which a plurality of datasets are similar to each other, and which of the other datasets each of the plurality of datasets refers to for which purpose, and also, the plurality of datasets are semantic (language It refers to a kind of graph that indicates the degree to which it is related).

Through this relationship map, researchers and others can not only view various types of datasets scattered at a glance, but also efficiently and effectively select datasets required for their research based on these relationship maps. You can. In general, data is meaningful in itself, but is more meaningful when it is combined or related with other data, and is useful for research. According to an embodiment, various types of datasets scattered in various places can be used. Likewise, it can be said that this is an effect according to an embodiment of the present invention in that the degree of binding or relevance between each other can be grasped at a glance.

FIG. 2 is a diagram conceptually showing a relationship map according to an exemplary embodiment and the structures underlying the relationship map.

Referring to Figure 2, there are three planes (20, 30, 40). The first plane 20 is a plane 20 in which a plurality of data sets 21 are present. The plurality of data sets 21 are connected with each other with similarity, linkage, and semantic association, and in FIG. 2, the similarity, linkage, and semantic association are displayed according to the shape of the line connecting the dataset 21. have. For example, similarity is indicated by a straight line, linkage by a dotted line, and semantic association by a spiral.

The second plane 30 is a plane 30 in which a plurality of meta data 31 is present. The plurality of meta data 31 is also connected to each other with the above-described similarity, linkage, and semantic association. In FIG. 2, the similarity, linkage, and semantic association depend on the shape of the line connecting the meta data 31. Is marked. For example, similarity is indicated by a straight line, linkage by a dotted line, and semantic association by a spiral.

Here, each meta data 31 may be extracted from each of the data sets 21 existing in the first plane 20 (definition of the meta data 31 and the meta data 31 data set ( The specific methodology extracted from 21) will be described later).

The third plane 40 is a plane 40 in which a relationship map according to an embodiment in which a plurality of data sets 41 are disposed on a concentric circle is present. Here, unlike the data set 21 existing in the first plane 20, the data set 41 existing in the third plane 41 contains a certain amount of data in each data set 41. It can be expressed in a predetermined manner, and it can also be expressed in a predetermined manner whether each data set 41 is an open data set or an unopened data set. Of course, even in the third plane 40, the plurality of data sets 41 are connected with similarity, linkage, and semantic association between each other, and in FIG. 2, the similarity, linkage, and semantic association are meta data 31 ) According to the shape of the line connecting them. For example, similarity is indicated by a straight line, linkage by a dotted line, and semantic association by a spiral.

Here, the process of generating the relationship map will be briefly described. First, a plurality of data sets 21 are provided as in the first plane 20. Thereafter, meta data 31 is extracted from each data set 21 and is disposed as in the second plane 30. Thereafter, similarity, linkage, and semantic association between the extracted meta data 31 are calculated. Thereafter, a plurality of metadata 31 are connected to each other as illustrated in FIG. 2 based on the similarity, linkage, and semantic association between the metadata thus calculated. Thereafter, based on the plurality of meta data 31 connected to each other, the plurality of data sets 21 existing in the first plane 20 are connected to each other as shown in FIG. 2.

Finally, when a predetermined request is received from a researcher or a user, a data set that satisfies this request is selected, and a relationship map centering the selected data set as a center of a concentric circle is shown in the third plane 40 of FIG. 2. Is generated. In this case, in the case of a plurality of datasets 41 disposed on concentric circles, similarity, linkage, and semantic association are based on the connection of the plurality of datasets 21 to each other in the first plane 20. It is connected to be displayed.

Here, what is displayed or provided to the researcher or the user may be a relationship map displayed on the third plane 40 among the planes illustrated in FIG. 2. A researcher or a user can view a plurality of data sets at a glance through the relationship map displayed as described above, and a predetermined meaning can be derived from the plurality of data sets through the relationship map.

Hereinafter, a detailed configuration of the data set management apparatus 100, which is a configuration for generating a relationship map, will be described in more detail.

3 is a schematic configuration diagram of a configuration of a data set management apparatus 100 according to an embodiment. Referring to FIG. 3, the data set management apparatus 100 includes a storage unit 110, a meta data extraction unit 120, a similarity calculation unit 130, a linkage calculation unit 140, and a semantic association calculation unit 150 ), A relationship map generating unit 160, an interface unit 170, a control unit 180, and a display unit 190, but does not include or are not mentioned at least one of the above-described configurations according to an embodiment. You may also include additional configurations

First, the data set management apparatus 100 and each configuration included therein can be implemented by a memory storing instructions programmed to perform a function to be described below and a microprocessor performing these instructions.

Among them, various types of data sets may be stored in the storage unit 110. The data set stored in this way may be obtained from the outside, for example, the data set management apparatus 100 may be obtained from the first server 200 to the second server 210 illustrated in FIG. 1. The data set stored in the storage unit 110 can be updated.

The meta data extraction unit 120 may extract meta data from each of a plurality of data sets stored in the storage unit 110.

Here, the meta data refers to information representing the properties of each data set. For example, the name of the provider that provided the data set, the name of the management department, the registered date, and the modified date, as well as the number of files included in the data set, extension, number of views or downloads, etc. may be examples of metadata.

The meta data extraction unit 120 may employ various methods to extract the meta data from the data set. For example, a plurality of data is included in a data set, and any one of them is an Excel file format list of the remaining data. The rest of the data is listed (listed) in this list, and items representing the properties of each data are provided, and content corresponding to each item may be included as content. Then, the metadata extracting unit 120 may extract at least some of these items and at least some of the content corresponding to each item as metadata. At this time, since the technology itself for extracting items or contents from an Excel file is a known technology, a description thereof will be omitted.

The similarity calculating unit 130 calculates the similarity between data sets based on the degree of similarity between meta data. Here, the meaning based on the degree of similarity between metadata may be to measure the similarity between the vectors when the metadata included in each dataset is expressed as a vector, but is not limited thereto. The method of determining the degree of similarity between metadata may be, for example, one of cosine similarity or jaccard similarity between vectors representing metadata, but is not limited thereto.

The linkage calculation unit 140 calculates the linkage based on whether each of the plurality of datasets refers to each other or has been previously referenced. For example, when the first dataset is a subway map of the Seoul area and the second dataset is a restaurant list in the Seoul area, the first dataset and the second dataset refer to each other in order to find the 'restaurant list in Gangnam, Seoul'. It has to be used. The linkage can be calculated based on whether the reference is made in this way. In order to calculate the linkage, in order to achieve a predetermined purpose, information on whether or not a plurality of datasets have been used together before may be secured and utilized.

The semantic association degree calculating unit 150 calculates a semantic association degree between datasets based on the semantic information of each meta data. Here, the semantic association degree is not simply based on the identity of meta data, but is a measure indicating how much the meta data has the same or similar semantics, that is, semantic information. Such semantic association may be calculated by a predetermined known module designed to extract semantic information from metadata.

The relationship map generation unit 160 generates a relationship map that displays a relationship between at least some of the plurality of data sets. Here, 'relationship' refers to at least one of similarity, linkage, and semantic association calculated as described above. The process or procedure of generating the relationship map has been described in FIG. 2, and thus a description thereof will be omitted.

The interface unit 150 is a component that accepts and processes various types of requests or inputs from a user or a researcher or an external subject of the data set management apparatus 100. For example, the interface unit 150 may be a touch screen, a keyboard, a microphone, or a mouse, as well as a device to configure an input signal from such an input device to understand its meaning.

Through the interface unit 150, various functions related to the relationship map may be performed in the data set management apparatus 100 according to an embodiment.

For example, in the process of generating the relationship map, the interface unit 150 may receive input of which of the above-described similarity, linkage, and semantic association will be used.

In addition, the interface unit 150 may receive the criterion so that data used for generating the relationship map can be selected as a predetermined criterion or higher. For example, the interface unit 150 may receive a first threshold, a second threshold, and a third threshold for each of the similarity, linkage, and semantic association described above. In this case, the relationship map generation unit 160 does not generate a relationship map by using each of a plurality of data sets, but only a data set connected to have similarity, linkage, and semantic association of at least the above-mentioned thresholds among the plurality of data sets. Can be used to create a relationship map.

In addition, when the relationship map forms a concentric circle structure, the interface unit 150 may receive the maximum number of circles constituting such concentric circles. If the maximum number of circles is compared with the case of 10 and 5, the size of the relationship map will be larger than the case of the former case and the number of datasets with a larger capacity or number may be displayed by being intertwined in the relationship map. There will be. Of course, in some cases, the researcher or the user may want a relationship map of a concentric circle structure having a smaller number of circles. In this case, the researcher or the user can select the desired number of circles for the concentric circle through the interface unit 150. You will be able to enter.

The controller 180 may be used for various types of control related to the relationship map. For example, the data set management apparatus 100 may be operated while the operating mode is toggled between the data mode and the keyword mode (to be described later with respect to keywords). The controller 180 may be a researcher or a user through the interface unit 150. When the operation mode selects the data mode, each configuration (110 to 190, etc.) is controlled so that a relationship map can be generated accordingly. Alternatively, when a researcher or user selects the interface unit 150 in the keyword mode, accordingly Each configuration (such as 110 to 190) can be controlled so that a relationship map can be generated.

In addition, the controller 180 may allow various types of services to be provided in relation to the relationship map according to the location of the user's mouse pointer. For example, a legend may be provided at a specific location, a window in which a keyword or dataset is searched may be displayed or provided, and an area in which a search result is displayed may be generated and provided, but is not limited thereto. An example of such a service provided by the controller 180 will be described in more detail in FIG. 4.

The display unit 190 is a device that displays or displays components related to various relationship maps provided by the relationship map or the controller 180, and can be implemented by a monitor, an LCD panel, and the like.

The keyword extraction unit 191 refers to information substantially included in data included in the data set. For example, if the dataset is related to forest fire information in 2019 in Gyeongsangbuk-do, the keywords may be 'wildfire', '2019', 'Gyeongsangbuk-do', but are not limited thereto. Here, since the technology itself for extracting keywords from data is known, detailed descriptions thereof will be omitted.

When the keyword extraction unit 191 extracts keywords, the similarity calculation unit 130, the linkage calculation unit 140, and the semantic association calculation unit 150 calculate similarity, linkage, and semantic association between each keyword It can be, the specific method may be the same as applied to the metadata or dataset.

In addition, the relationship map generator 160 may generate and provide a relationship map for these keywords, not a relationship map for the dataset. That is, the relationship map generating unit 160 may generate and provide a relationship map in which similarities, linkages, and semantic associations between keywords are displayed.

As described above, a relationship map representing a relationship between a plurality of datasets according to an embodiment may be generated and provided to a researcher or a user, and the researcher or user may control or process the relationship map in a desired form. From this, desired information can be obtained.

Hereinafter, an example of a relationship map and a screen displaying information related to the relationship map will be described with reference to FIG. 4. However, since FIG. 4 is merely exemplary, the spirit of the present invention is not limited to being illustrated in FIG. 4.

4, a relationship map is displayed. This relationship map may be the relationship map described so far. Here, the configuration representing each data set in the relationship map may be a circle, and the size and type of the circle indicate whether the corresponding data set is an open data set or an unopened data set, or the data set of a certain capacity. It may indicate whether is included.

In addition, when the mouse pointer or the like moves through the interface unit 150 and is positioned on a specific data set, the controller 180 can display the meta data included in the corresponding data set as a pop-up. It is exemplarily shown where data is popped up and displayed on the screen.

On the other hand, let's look at the configuration one by one from the top left of the screen.

The 'mode toggle button' is a means for determining an operation mode of the data set management apparatus 100, for example, between a data mode and a keyword mode (to be described later with respect to keywords), and can be implemented through the interface unit 170 described above. Do. If a researcher or a user selects an operation mode and a data mode, each configuration (110 to 190, etc.) is controlled so that a relationship map can be generated accordingly. Alternatively, the researcher or user searches the interface unit 150 in keyword mode. If is selected, it is possible to control each configuration (110 to 190, etc.) so that a relationship map can be generated accordingly.

The 'dataset' is a button provided to perform functions related to the dataset. For example, if you click 'Dataset', the information about the dataset can be listed and displayed as it is expanded, and the user can select a specific dataset (dataset) using this list.

The 'Classification System', 'Year', 'Extension', 'Region', and 'Limit of Use' are similarly designed to perform certain functions related to the dataset or relationship map.

Meanwhile, the 'search column' is a configuration provided to receive a desired data set or keyword from a researcher or user, and is implemented by the interface unit 170. When a researcher or a user inputs a data set or a keyword, the controller 180 selects a data set or keyword corresponding to the input data set or input keyword from the data set stored in the storage unit 110 or keywords included therein. , The relationship map generator 160 may generate a relationship map based on these data sets or keywords.

The 'data basket' is a conceptual storage that temporarily stores datasets, etc. selected by researchers or users for data analysis. Through this data basket, researchers or users can download data sets of their choice at once. Therefore, the time required for the procedure of fusion and analysis of these can be shortened.

The 'multi-dimensional map view' provides a function for selecting among similarity, linkage, and semantic association to be reflected in the relationship map, and can be implemented by the interface unit 170. Here, the degree of similarity may be displayed in the form of a similarity map, the degree of association may be in the form of a connection map, and the semantic association degree may be displayed in the form of a semantic map.

The 'data filter' is a structure that a researcher or a user manipulates to view a relationship map in a form desired by the user, and can be implemented by the interface unit 170 likewise. Such a data filter can be used to receive the number of circles forming the above-described threshold or concentric circle structure, and since it has already been described, a duplicate description will be omitted.

The 'search list' is a means or button for storing and showing the history searched by the researcher or user through the search field, and this can also be implemented by the interface unit 170.

5 is a flowchart illustrating a procedure of a dataset management method according to an embodiment, and such a dataset management method may be performed by the dataset management apparatus 100 described above. At this time, since the flowchart shown in FIG. 5 is merely exemplary, embodiments such as performed in a different order from that shown in FIG. 5 are not excluded by FIG. 5.

Referring to FIG. 5, extracting meta data from each of a plurality of data sets (S100), and calculating the similarity between the plurality of data sets based on the degree of similarity between the metadata, and each of the plurality of data sets Calculating the degree of interlinking between the plurality of datasets based on whether they are referred to each other, and calculating the semantic correlation between the plurality of datasets based on the semantic information of each of the metadata (S110) And, based on at least one of the calculated similarity, the calculated linkage and the calculated semantic association, generating a relationship map displaying a relationship between the plurality of data sets (S120) You can.

Hereinafter, since the data set management method is performed by the data set management apparatus 100, it has substantially the same technical idea as the data set management apparatus 100. The description portion of the data set management apparatus 100 will be used.

On the other hand, the data set management method according to an embodiment is implemented in the form of a computer-readable recording medium storing a computer program programmed to perform each step included in the method, or a computer program stored in the computer-readable recording medium. It is possible.

The above description is merely illustrative of the technical spirit of the present invention, and those of ordinary skill in the art to which the present invention pertains will be capable of various modifications and variations without departing from the essential quality of the present invention. Therefore, the embodiments disclosed in the present invention are not intended to limit the technical spirit of the present invention, but to explain, and the scope of the technical spirit of the present invention is not limited by these embodiments. The scope of protection of the present invention should be interpreted by the following claims, and all technical ideas within the scope equivalent thereto should be interpreted as being included in the scope of the present invention.

100: data set management device

Claims (13)

A metadata extracting unit that extracts metadata from each of the plurality of data sets
A similarity calculating unit calculating a similarity between the plurality of data sets based on the similarity between the metadata;
A linkage calculation unit for calculating an interlinking degree between the plurality of datasets based on whether each of the plurality of datasets refers to each other,
A semantic correlation calculating unit for calculating a semantic correlation between the plurality of data sets based on the semantic information of each of the metadata;
A relationship map generating unit generating a relationship map indicating a relationship between the plurality of data sets based on at least one of the calculated similarity, the calculated association, and the calculated semantic association;
And an interface unit that receives at least one of a first threshold for the similarity, a second threshold for the association, and a third threshold for the semantic association,
The relationship map generation unit,
The relationship based on at least one of the similarity of the first threshold or higher among the calculated similarities, the association of the second or higher thresholds among the calculated associations, and the semantic association of the third or higher thresholds among the calculated semantic associations. To generate a map
Dataset management device. According to claim 1,
Each of the plurality of data sets,
Containing a plurality of data related to the subject
Dataset management device. According to claim 2,
The plurality of data includes content corresponding to each of at least two items,
The metadata extraction unit,
Extracting at least a portion of the at least two items and at least a portion of content corresponding to each of the at least two items as metadata for a corresponding data set
Dataset management device. According to claim 1,
Similarity between the metadata,
It is calculated by the cosine similarity or jaccard similarity between vectors calculated for each of the metadata.
Dataset management device. According to claim 1,
Whether each of the plurality of data sets is referenced to each other,
Among the plurality of datasets, it is determined that the datasets used together in the process of deriving a predetermined result are referred to each other, and the datasets not used together in the process of deriving the result are referred to each other. It is judged not to,
In the process of deriving the result, information on whether some of the plurality of data sets are used together is information obtained in advance.
Dataset management device. According to claim 1,
The interface unit,
Among the calculated similarity, the calculated linkage, and the calculated semantic association, further receiving the basis for generating the relationship map,
The relationship map,
It is generated based on the more input through the interface unit
Dataset management device. delete According to claim 1,
The relationship map has a concentric circle structure in which a plurality of circles having different diameters form a concentric circle,
One of the plurality of data sets is disposed at the center of the concentric circle, and the other is disposed on a circle other than the center of the concentric circle among the plurality of circles.
Dataset management device. The method of claim 8,
The interface unit,
The number of circles constituting the concentric circle is further input.
Dataset management device. According to claim 2,
The data set management device,
Further comprising a keyword extraction unit for extracting keywords from each of the plurality of data,
The similarity calculating unit calculates the similarity between the keywords,
The linkage calculation unit calculates a linkage between the keywords,
The semantic correlation calculating unit calculates the semantic association between the keywords,
The relationship map generation unit,
Based on at least one of the similarity, linkage, and semantic association calculated between the keywords, generating a relationship map indicating the relationship between the keywords
Dataset management device. As a data set management method performed by the data set management device,
Extracting metadata from each of the plurality of data sets,
Calculating the similarity between the plurality of data sets based on the similarity between the metadata,
Calculating a degree of interlinking between the plurality of data sets based on whether each of the plurality of data sets refers to each other,
Calculating a semantic association between the plurality of data sets based on the semantic information of each of the meta data,
Generating a relationship map displaying a relationship between the plurality of data sets based on at least one of the calculated similarity, the calculated association, and the calculated semantic association;
And receiving at least one of a first threshold for the similarity, a second threshold for the association, and a third threshold for the semantic association,
The step of generating the relationship map,
The relationship based on at least one of the similarity of the first threshold or higher among the calculated similarities, the association of the second or higher thresholds among the calculated associations, and the semantic association of the third or higher thresholds among the calculated semantic associations. To generate a map
How to manage datasets. A computer-readable recording medium storing a computer program,
The computer program,
Comprising instructions for causing the processor to perform the method according to claim 11
Computer-readable recording media. A computer program stored on a computer-readable recording medium,
The computer program,
Comprising instructions for causing the processor to perform the method according to claim 11
Computer program.

Images