KR20180111646A - Device and method for chronological big data curation system - Google Patents

Abstract

The present invention relates to a curation device based on chronological information and a controlling method thereof. More specifically, the present invention relates to a curation device based on chronological information for providing event flow information and a method thereof. The present invention summarizes various types of information scattered on the internet through topics and associations, easily and accessibly shows the summarized data to users and enables the summarized data to reuse to generate valuable information through maintaining, managing and providing for the information. The curation device comprises: a data collecting part; a graph modeling part; and a chronological analyzing part.

Description

Technical Field [0001] The present invention relates to a chronological information-based curation apparatus for providing event flow information, and a control method thereof. [0002]

The present invention relates to a chronological information-based curation apparatus and method, and more particularly to a chronological information-based curation apparatus and method for providing event flow information.

Due to the advent and development of web technologies, vast quantities of different kinds of data are being produced rapidly and the amount of information is increasing considerably. In this big data era, consumers can get more data and information than ever before, but it is not easy to find valuable data out of a lot of information and knowledge. Finding valuable information from large amounts of data is becoming increasingly important, and many countries and companies are investing a lot of time and money in data acquisition and analysis.

Digital curation refers to the task of organizing various kinds of information scattered on the Internet through themes and associations, showing the organized data easily and easily to the users, and making the data reusable . Digital curation can generate valuable information through maintenance, management and provision of information, and can improve accessibility and reusability of information.

One of the important issues is to create a beneficial and reliable digital curation device that meets the user's needs, as it takes a lot of effort and cost to search, understand and identify important topics in this big data age.

Currently existing information systems provide search services based on queries.

Semantic search is a search that can improve the search accuracy by understanding the searcher's intentions and the contextual meaning of the words in order to get better performance and get more accurate results. However, since the current search system including such a semantic system considers only time-independent parameters, it has a great difficulty in expressing and ranking search data considering time damping effect.

Some events occur over a long period of time, and the importance and interests associated with the events change over time. When various events and / or accidents are intertwined with others, it becomes difficult to understand the underlying and fundamental implications associated with search data.

Accordingly, there is a need for research on a digital curation apparatus and system that consider the time of occurrence of an event.

The present invention is directed to solving the above-mentioned problems and other problems. Another object of the present invention is to provide a culity apparatus capable of classifying data more efficiently in consideration of occurrence time between events.

The present invention provides a chronological information-based curation apparatus and method for providing relevant key information in chronological order to help a comprehensive understanding of a specific event or accident that a user is searching for in a big data environment.

The present invention collects event or accident data from a variety of sources, analyzes the information chronologically, provides a modeling of a particular event or knowledge over time, is presented through visualization tools, This invention provides a chronological information-based curation apparatus and method that can reduce repetitive search operations and generate reusable information.

It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory and are not restrictive of the invention, unless further departing from the spirit and scope of the invention as defined by the appended claims. It will be possible.

According to an aspect of the present invention, there is provided a data collecting apparatus including: a data collecting unit collecting a plurality of topic data from a data source on a web; A graph modeling unit for extracting an association between the collected plurality of topic data; And a chronological analyzer for classifying an internal node set directly related to a predetermined topic among the plurality of topic data and an external node set indirectly related to the predetermined topic based on the extracted correlation, Lt; / RTI >

Here, the data source may include a post of a social network service (SNS), and a news article.

The predetermined topic may be a topic for a keyword of a keyword inputted from a user.

In addition, the association between the two topic data can be quantified in the number of documents in which the two topic data are included together in a single document.

The chronological analyzer classifies the topic data having a relation with the predetermined topic into an internal node set and classifies the topic data related to the generation time into an external node set although it is not related to the predetermined topic .

The chronological analyzer may sort the plurality of classified topics in order of generation time.

According to another aspect of the present invention for achieving the above or other objects, there is provided a method comprising: collecting a plurality of topic data from a data source on a web; Extracting an association between the collected plurality of topic data; And classifying into an external node set indirectly related to the predetermined topic, based on the extracted association, an internal node set directly related to a predetermined topic among the plurality of topic data Control method.

Advantages of the curation apparatus and the control method thereof according to the present invention are as follows.

According to at least one embodiment of the present invention, it is possible to organize various types of information scattered in the Internet through topics and associativity, to display the organized data easily for users, and to reuse the organized data And maintains, manages, and provides information, thereby generating valuable information.

Further scope of applicability of the present invention will become apparent from the following detailed description. It should be understood, however, that the detailed description and specific examples, such as the preferred embodiments of the invention, are given by way of illustration only, since various changes and modifications within the spirit and scope of the invention will become apparent to those skilled in the art.

1 is a block diagram of a cueing apparatus according to an embodiment of the present invention.
FIG. 2 graphically illustrates the nature of SNS, blog, and news articles that are distinguished in accordance with one embodiment of the present invention.
3 is a diagram showing a topic alignment structure for aligning a main topic and a sub topic according to an embodiment of the present invention.
4 is an exemplary diagram illustrating a time relation graph according to an embodiment of the present invention.
5 is an exemplary diagram illustrating a similarity relation graph according to an embodiment of the present invention.
FIG. 6 and FIG. 7 are diagrams showing an example in which the reuse unit 400 stores a tree structure according to an embodiment of the present invention.
FIG. 8 is a flowchart illustrating a method of controlling a curation apparatus according to an embodiment of the present invention.
FIG. 9 is a graph showing a result of a search by a famous politician Ahn Cheol-Su of a curation apparatus according to an embodiment of the present invention.
FIG. 10 is a graph of a search result for the "Nonghyup" hacking case, which is a Korean bank, by a curlation apparatus according to an embodiment of the present invention.
11 is a graph showing a result different from the above experimental results by the curation apparatus according to an embodiment of the present invention.

Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings, wherein like reference numerals are used to designate identical or similar elements, and redundant description thereof will be omitted. The suffix "module" and " part "for the components used in the following description are given or mixed in consideration of ease of specification, and do not have their own meaning or role. In the following description of the embodiments of the present invention, a detailed description of related arts will be omitted when it is determined that the gist of the embodiments disclosed herein may be blurred. It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory and are intended to provide further explanation of the invention as claimed. , ≪ / RTI > equivalents, and alternatives.

Terms including ordinals, such as first, second, etc., may be used to describe various elements, but the elements are not limited to these terms. The terms are used only for the purpose of distinguishing one component from another.

It is to be understood that when an element is referred to as being "connected" or "connected" to another element, it may be directly connected or connected to the other element, . On the other hand, when an element is referred to as being "directly connected" or "directly connected" to another element, it should be understood that there are no other elements in between.

The singular expressions include plural expressions unless the context clearly dictates otherwise.

In the present application, the terms "comprises", "having", and the like are used to specify that a feature, a number, a step, an operation, an element, a component, But do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, or combinations thereof.

The present invention relates to a curation apparatus and a control method thereof, and curation refers to a method for determining how to display various data. For example, in a news article, search systems such as the existing "Naver News" will only get the same articles (titles only slightly different) at the same time when you enter certain keywords. As a result, the results of such a system can be viewed as a rather difficult search result screen for verifying various results related to the search keyword. In order to overcome this, according to the present invention, a plurality of topics (multiple keywords of articles according to time) related to a specific keyword for searching are formed, and a plurality of topics are appropriately divided into various parameters to provide an effective search result to the user The purpose is to do.

Particularly, in the present invention, in constructing a plurality of topics, it is proposed to consider not only direct relevance but also indirect relevancy. The indirectness is estimated by taking into consideration the time at which the topic occurred closest to the topic occurrence time of the subject (keyword) inputted at the time of searching. Indirectness can mean a criterion that a person has not yet judged.

 "In the example of the experimental results of" forest fire in Pohang ", it is common that when searching with the keyword" forest fire " In addition, according to the embodiment of the present invention, since the indirectness by time is further taken into account, it is possible to see "" forest fire "occurring in other areas even when searching with the keyword" POHANG ". Because "POHANG" and "fire" do not have a direct relationship (except in the case other than the case), they are topics that occur at the same time, so that indirect relevance can be recognized according to the present invention. Due to this indirect relevance, when a search term "Pohang" is input, the topic "forest fire" is composed together (hereinafter, these topics will be described as an external topic) So that it can be shown to the user.

Consider an external node group (hereinafter, referred to as an external node group) in which a directly related topic group (internal node group, hereinafter referred to as an internal node group) and indirectly related temporal association exists do. In addition, it will be possible to determine which search results will be provided with a certain threshold value (using Maximum Cut), so that only the results that the user is interested in can be properly displayed. The difference in time is small (recent topics around the topic), and associativity will provide the most significant results through the appropriate thresholds, which are maximum (topic-like topics). In other words, the Internal Node Group searches for a topic by intuitiveness, and the External Node Group is a topic that is provided to search for the area implied.

1 is a block diagram of a cueing apparatus according to an embodiment of the present invention.

1, the chronological information-based curation apparatus according to an embodiment of the present invention includes a data collecting unit 100, a graph modeling unit 200, a chronological analysis unit 300, and a reuse unit 400 Lt; / RTI >

The data collection unit 100 collects a plurality of topic data from a data source on the web. Here, the data source may include a social network service (SNS) publication, a news article, and the like, and may include any source capable of providing data through a www (world wide web) May be included.

And, the topic data may mean keyword data on an event or a subject that can receive the attention or attention of the people. For example, it could be the words 'presidential election' or 'fire.' Specific methods of collecting the topic data will be described later in detail.

The data collecting unit 100 collects not only the topic data but also the time at which the topic data is generated (may include both the date and the time, hereinafter referred to as the generation time including both the generated date and time). ), As well as information on the other.

The generation time, topic data, and association data collected from each data source in the data collecting unit 100 are processed through a preprocessing process (a step of the graphic modeling unit) to be described later, .

Particularly, in one embodiment of the present invention, it is proposed to collect data by dividing the nature of the social network service (SNS) among the data sources into at least two types as follows.

For example, it can be divided into a first SNS in which users can easily share their daily lives in a mobile or a PC, and a second SNS provided in order to share detailed information. Examples of the first SNS include kakaotalk, twitter, instagram, and facebook. Examples of the second SNS include a blog or a brunch brunch).

The reason for distinguishing the characteristics of the SNS is that the types of data collected may vary depending on the personality.

FIG. 2 graphically illustrates the nature of SNS, blog, and news articles that are distinguished in accordance with one embodiment of the present invention. In the diagram, the SNS means the first SNS mentioned above, and the blog means the second SNS mentioned above.

In the case of the first SNS for daily conversation, the length of the document is very short and the propagation power is very high because users easily write their opinions, but it will be somewhat lacking in reliability.

Even in the case of the second SNS for sharing information, the length of the document is on the short side, but the propagation power is somewhat lacking, and in terms of reliability, it may be a middle level higher than the second SNS.

Contrary to the above two examples, news articles are very much more reliable in terms of reliability than the above-mentioned SNS, and will be very high in terms of propagation power because they are written by professional reporters. In addition, since the length of the document is also secured to a predetermined length or more, it can be higher than the case of the above-described two SNSs.

In an embodiment of the present invention, when the propagation power is high as in the first SNS, but is somewhat low in terms of document length and reliability, it is proposed to collect only the generation time of topic data and topic data.

Further, in the embodiment of the present invention, since the second SNS and the news article have a somewhat higher reliability, it is proposed to collect the relationship between the data.

The association thus collected is used to classify the topic data below.

The extraction (collection) of such an association is performed in the graph modeling unit 200.

The graph modeling unit 200 performs a preprocessing process to obtain meaningful data from the data collected by the data collecting unit 100. The preprocessing process is a process that encompasses all the various processes necessary for selecting only the important data necessary for analysis.

The graph modeling unit 200 extracts a relation between the plurality of topic data. And, the association can be extracted (collected) from a data source whose credibility is recognized to some extent, such as the second SNS (blog and the like) and the news article in the above example. The data for this association may be data using at least one of similarity and coexistence frequency between the plurality of topic data. That is, the graph modeling unit 200 uses at least one of similarity and coexistence frequency in determining an association between any one topic data and another topic data.

The frequency of simultaneous occurrence may refer to the number of documents including the two topic data together in a single document. That is, the number of cases in which words indicating two topic data are included in one document at the same time. For example, when it is desired to grasp the association between the first topic data and the second topic data, the number of the single documents including the first and second topic data at the same time is referred to as a relation between the first and second topic data It will be possible. In this case, the higher the correlation, the higher the correlation between the two topic data.

If there is a strong association between the two topic data (that is, if the association is high), then the main topic and sub-topic are classified according to the generation time order. The relationship between topic data can be expressed in a graphical structure that can be extended to map and reduce for parallel processing.

As will be described later in detail, in graphically displaying collected data, each topic data is represented by a node on the network, and the association is represented by a link structure linking nodes and nodes.

3 is a diagram showing a topic alignment structure for aligning a main topic and a sub topic according to an embodiment of the present invention. The illustrated topic alignment structure is a hierarchical structure of topic data associated with a predetermined topic (a specific event).

Each topic data may include sub-topic data according to a generation time (event creation time), and may recursively have a sub-topic (sub-topic data below the sub-topic data). For each major topic, the subtopics are hierarchically represented using a directional graph structure.

As shown in FIG. 3, the graph includes topic data as a node, and a directional link that connects two nodes with each other. Directionality means the order of the generation time relationship, and the value can mean the difference in the generation time. Although the values are not shown in Fig. 3, these values will be described later with reference to Figs. 4 and 5 below.

In FIG. 3, the node A 301-1 represents a main topic (for example, a topic for a keyword of a search word input from a user) and the other nodes 301-2 to 301-6 refer to a sub topic. This topic alignment structure is useful for understanding the hierarchical structure of topics. Analysis of the graph can analyze which events (events) occurred as a series of processes. In FIG. 3, node C 301-3 has nodes E, F (301-5, 301-6) for two lower and lower subjects E and F, and the main topic A has an implicit relationship with E, F . It can be seen that the subtopic C occurred after the main topic A, and that the topics E and F occurred after the topic C.

Two types of links for graphically representing the relationship between two topic data will be described with reference to FIGS. 4 and 5. FIG.

4 is an exemplary diagram illustrating a time relation graph according to an embodiment of the present invention. 5 is an exemplary diagram illustrating a similarity relation graph according to an embodiment of the present invention.

Referring to FIG. 4, a value (numerical value) of a link connecting two topic data means a difference in generation time of two topic data. That is, the difference between the generation time of the first topic data and the generation time of the second topic data is numerically expressed. The temporal relationship link is directional depending on the generation time prefix.

In Fig. 4, there is shown a temporal relationship link between the five topic data 401-1 to 401-5. Indicates that topic data B (401-2) has been generated after 25 hours of topic data A (401-1) and topic data D (401-4) has occurred one hour after topic data A (401-1).

5 shows a link showing only the degree of similarity not related to the directionality. Such a similarity link may have a value obtained by quantifying the above-mentioned simultaneous occurrence frequency using the TF-IDF (Term Frequency - Inverse Document Frequency).

After the time relation link is completed as shown in FIG. 4, the similarity between the topic data of FIG. 5 can be calculated. This is because similarity exists between two topic data that occur simultaneously in the document, as described above. In order to quantify the association, the association between each topic data (keyword) is quantified by using the frequency of simultaneous appearance of the blog and news.

1, and the chronological analyzer 300 will be described.

The chronological analyzer 300 analyzes an internal chronical node group, which is a set of a series of events occurring in a time order, with respect to a specific topic (main topic) (External Chonical Nodes Group), which is a set of potential events for which no association exists. Based on the association between the topic data extracted by the graph modeling unit 200, the topic data is divided into an internal node set and an external node .

The internal node set represents a set of topic data arranged in a sequence of chronological order and associated with the main topic. An external node set is a collection of topic data that is indirectly related to a major topic.

The relationship between the parent node (parent node) and the child node (child node) is determined according to the time order.

The relationship between the external node set and the main topic is low, but a similar event occurred at a similar time. That is, as shown in FIG. 3, the outer topic 304-1 at the top of the outer node set may have no association with the main topic 301-1, or may be low, There is relevance.

The number of nodes that will ultimately be shown on the topic alignment structure of FIG. 3 may be limited to an appropriate number depending on the generation time and the similarity value between links. In particular, according to the embodiment of the present invention, the number of frames can be limited to a suitable number based on a maximum cut method. That is, it is possible to control the number of nodes to be displayed through a maximum cut threshold according to a preset or user input.

On the other hand, if another topic appears after a long time since the first topic appeared, the relationship between the two topics may not be important. Or, a high relevance topic may reappear after a long period of time despite the high degree of similarity between the two topics. In order to satisfy the two cases, the nodes to be shown to the user use a maximum cut threshold value in which the related weight is high but the temporal weight is minimum.

The reuse unit 400 can visualize a graph according to the above-described method, and store the generated result. That is, the result of analyzing the chronological order according to the above-described method is visually displayed and the analyzed result is stored in a tree structure for later use.

FIG. 6 and FIG. 7 are diagrams showing an example in which the reuse unit 400 stores a tree structure according to an embodiment of the present invention.

Referring to FIGS. 6 and 7, when the graph according to the above-described method is configured, all information of the graph is stored in a standardized form. As a result, the availability and reusability of data will increase and may be useful for future maintenance.

The analysis result is stored in a tree structure method for use in the later chronological analysis step. For example, the tree structure is expressed using XML (eXtensible Markup Language) format as shown in FIG. That is, the graph data of FIG. 6 can be stored in the XML format as shown in FIG. The results of this analysis can be provided to the API or reused when the user types a query and enters keywords for key events.

FIG. 8 is a flowchart illustrating a method of controlling a curation apparatus according to an embodiment of the present invention.

According to the drawing, in step S810, a plurality of topic data may be collected from a data source on the web. As an example of such, it is possible to collect topic data and generation time from the first SNS data source as described above, and to collect the association between the data in the second SNS data source, news article, such as a blog.

In step S830, the chronological analyzer 300 may classify the topic data into an inner node set and an outer node set.

In step S840, the reuse unit 400 stores analysis results in a tree structure for visualization and reuse of the results of the chronological analysis.

In the following drawings and related explanations, data analyzed through actual experiments will be described in detail.

We used Twitter, Naver Blog, and Naver News for one month period for the experiment. The total number of data collected through Twitter is 1.5 million tweets and the total size is about 30 GB. We removed the meaningless words from Twitter and extracted a total of 600 topic data, and the graph was constructed using this topic data. It also collected 22,692 Naver blogs and 16,288 Naver news articles. Blog data is an average of 328 words per document, and news data is an average of 254 words per document. The links in the graph are composed of Naver blog and Naver News data, and 123,894 links are created in the topic alignment structure. In this experiment, we limited the range of datasets for SNS, blog, and news, but they can be easily extended by adding a preprocessing module to collect data from a wider variety of sources.

We used the noun extractor 'Komoran' to extract the topic data from the data source. Generation time and keywords were used to extract topic data. Related blogs and news data using keywords extracted from Twitter data sources. The frequency of simultaneous occurrence of words in the document was extracted from the blog and news data, the subject and TF-IDF pairs were calculated, and the similarity link corresponding to the association was normalized between 0 and 1.

The time relation link was calculated on the basis of the creation time of the article containing the topic extracted from the Twitter data. The time relation link value is calculated by the difference between the document creation time and the normalized value (or average value), and is also normalized from 0 to 1. Graphs are constructed using subject, similarity, and time relationships (associations). The set of internal and external nodes are organized according to chronological analysis. If the similarity link representing the association has a value higher than a certain threshold value, these nodes are included as parent nodes and child nodes. This set is called the Internal Chronological Group. Other nodes that do not exceed a certain threshold have been classified as External Chronological Groups.

FIG. 9 is a graph showing a result of a search by a famous politician Ahn Cheol-Su of a curation apparatus according to an embodiment of the present invention. "Ahn, Chul - soo" is the main topic, and "Jong - hoon Kim" is composed of the subtopic. "Ahn Chul - soo" was a candidate for the opposition party, and Kim Jong - hoon was regarded as a ruling party candidate. After that, "Ahn Cheol - su" resigned from the presidential candidate to support the opposition candidate. This series of events is visualized as a graph separated into a set of internal nodes and a set of external nodes associated with Ahn.

FIG. 10 is a graph of a search result for the "Nonghyup" hacking case, which is a Korean bank, by a curlation apparatus according to an embodiment of the present invention. One of the main suspicions related to the incident at the time was that the suspect was "North Korea" after the incident. And people thought that this hacking incident was related to "Cheonan". The external link with "North Korea" is "Kim Kwan Jin", the Minister of National Defense of Korea at the time of the accident. These series of events are visualized as a graph separated into an internal node set and an external node set related to "Nonghyup". There are keywords that seem to be unrelated, but in most cases they showed results related to "Nonghyup hacking".

11 is a graph showing a result different from the above experimental results by the curation apparatus according to an embodiment of the present invention. "White Day" is the anniversary on March 14th when a man presents candy to a woman. You can see that there is a sub topic "candy" in "White Day". There is semantic similarity between these two keywords, but it is difficult to confirm that there is a temporal relationship. The 'Chronological Big Data Curation' method, which is suggested when entering various search keywords, works hard in most cases and helps to understand chronological topics in a series of topics. However, we confirmed that there is a limit in the case of general topics or one-off events that do not last long.

The chronological information-based curlation method according to an embodiment of the present invention is compared with an existing method and evaluated. An element of the assessment is to answer the question "Is it possible to use a search term to comprehend a comprehensive understanding of the topic?" Inventors of the present invention compared Naver News portal, which is the most widely used news portal site in Korea.

Fig. 12 shows a Naver news result in which the search keywords "Ahn-kun" and "hacking" Referring to FIG. 12 (a), which is a result of "Ahn," only the articles of March 31, which is the latest date, are displayed in the search results of the first page even if they are sorted in order of accuracy as a search option. Two days ago, the article on March 29 was exposed for the 130th time. As you can see from this picture, it is difficult to understand the trend of the overall events, only report the news results. On the other hand, Fig. 12 (b) which is the result of the search keyword "hacking "

For a more accurate assessment, we conducted quantitative assessments of traditional assessment methods and 'Chronological Big Data Curation'. For quantitative evaluation, we measured the only words that appeared in the article.

FIG. 13 shows a result of comparing a chronological information-based curation method and a Naver News search result according to an embodiment of the present invention.

The average value of the following items was calculated based on the search term. 1) Percentage of unique words: Percentage of proper nouns that can identify topics for news, 2) Time distribution: Average time range of article creation time displayed in the final result, K @ N is the upper It is a value calculated by N number.

When comparing the top five to fifteen search results, it can be seen that the rate of inclusion of the unique word has a higher value than the existing Naver News search result and the curlation method according to an embodiment of the present invention has a higher value . Also, it can be seen that the curation method according to the embodiment of the present invention has a higher value in the search result time interval.

That is, it can be recognized that the result according to the present invention can provide a more comprehensive and accurate search result than the Naver News search result.

While the present invention has been particularly shown and described with reference to exemplary embodiments thereof, it should be understood that the same is by way of illustration and example only and is not to be taken by way of limitation, And the scope of the technical idea of the present invention is not limited by the description with reference to the drawings or the drawings. It will also be appreciated by those skilled in the art that the concepts and embodiments of the invention set forth herein may be used as a basis for modifying or designing other structures for carrying out the same purposes of the present invention It will be understood by those skilled in the art that various changes in form and details may be made therein without departing from the spirit and scope of the present invention as defined by the appended claims and their equivalents. And various changes, substitutions, and alterations can be made without departing from the scope of the invention.

Claims (7)

A data collection unit for collecting a plurality of topic data from a data source on the web;
A graph modeling unit for extracting an association between the collected plurality of topic data; And
And a chronological analyzer for classifying the set of internal nodes into an external node set indirectly associated with the main topic, based on the extracted association, the internal node set directly related to the main topic among the plurality of topic data,
Curation device.
The method according to claim 1,
The data source comprising:
Including social network services (SNS) posts, news articles,
Curation device.
The method according to claim 1,
Wherein the main topic is a topic for a keyword of a keyword inputted from a user,
Curation device.
The method according to claim 1,
The association between the two topic data,
The number of documents included together with the two topic data in a single document,
Curation device.
5. The method of claim 4,
Wherein the chronological analyzer comprises:
The topic data having an association with the main topic is classified into an internal node set,
Characterized in that the topic data related to the generation time is classified into an external node set although it is not related to the main topic.
Curation device.
The method according to claim 1,
Wherein the chronological analyzer comprises:
And sorting the classified plurality of topics in order of generation time.
Curation device.
Collecting a plurality of topic data from a data source on the web;
Extracting an association between the collected plurality of topic data; And
Grouping into an internal node set directly related to a predetermined topic among the plurality of topic data and an external node set indirectly related to the predetermined topic based on the extracted association.
A control method of a curation device.

Images