KR20210086370A - System and method for providing rmi of shared images rights - Google Patents

Abstract

The present invention relates to a system and method for providing rights management information for shared works. In particular, the system for providing rights management information for shared works finds shared works posted on each of several public domain sites, and provides reliable site rights management information among these as common rights management information for shared works of all sites.

Description

SYSTEM AND METHOD FOR PROVIDING RMI OF SHARED IMAGES RIGHTS

The present invention relates to a system and method for providing rights management information for a shared work, and more particularly, to find a shared work posted on each of several public domain sites, and to share the rights management information of a reliable site among all sites It relates to a system and method for providing rights management information for shared works that are provided as common rights management information for works.

A public domain is a work whose copyright has been donated by the author, whose copyright protection period has expired, or whose use is permitted to general users under certain conditions.

For these shared works, various organizations around the world are operating a search site that displays RMI (Rights Management Information), which indicates rights management information for a work, such as identification information of the work, author information, and terms of use information.

The shared works search site in operation can be easily used by anyone without any special restrictions as long as the scope of use for shared works such as use, copying, modification, and redistribution is complied with.

However, users often do not understand the RMI for shared works due to lack of understanding of copyright, different RMI system for each shared work search site, outdated, inconsistent or incorrect RMI expression system, etc. , which is a problem as it can lead to disputes over copyright infringement.

In particular, when users search for image-sharing works, while duplicate image-sharing works are found on multiple shared work sites, if the right holder partially updates the RMI information for only some sites, confusion is further aggravated by the different RMIs for each site. can be

Republic of Korea Patent No. 10-2049061 Republic of Korea Patent Publication No. 10-2017-0082025

The present invention is to solve the above problem, finds shared works posted on each of several public domain sites, and uses the rights management information of reliable sites among them as common rights management information for the shared works of all sites. We would like to provide a system and method for providing rights management information for shared works.

To this end, a shared work rights management information providing system according to the present invention includes: a web crawler for accessing a plurality of public domain sites to collect image shared works; an object extractor for extracting an object included in the collected image sharing work; a feature point extractor for extracting at least one feature point characterizing the extracted object; a duplicate extractor for extracting duplicate public domain sites on which the same shared work is posted by comparing the feature points of the object; a site extractor for inferring a site providing the most reliable rights management information among the duplicate public domain sites; and a copyright manager providing rights management information for the shared work based on the inferred rights management information of the site.

In this case, it is preferable that the object extractor extracts an object by processing the image-sharing work as an image signal, and generates a keyword for the object by labeling the extracted object.

In addition, it is preferable that the key point extractor extracts the key point only for the object of the image sharing work having the same keyword as the keyword generated by the object extractor.

In addition, it is preferable that the site extractor infer the site providing the most reliable rights management information by applying a weight model including the reliability of the site and the reliability of the posts registered in the site.

In addition, the reliability of the site is determined by including any one or more of the number of visits to the site, the number of post registrations, and the update method of the shared work, and the reliability of the post is determined by the number of views, the number of downloads, the date of the updater, and the meta for the shared work. It is preferable that reliability is determined by including any one or more of the number of data information.

In addition, it is preferable that the copyright manager creates a rights management information expression system that integrates different domestic and foreign rights management information expression systems into one, and provides rights management information for the shared work through the created rights management information expression system. .

On the other hand, the method for providing shared works rights management information according to the present invention includes a work collection step of accessing a plurality of public domain sites in a web crawler to collect image shared works; an object extraction step of extracting an object included in the collected image sharing work in an object extractor; a feature point extraction step of extracting at least one feature point characterizing the extracted object in a feature point extractor; a duplicate extraction step of extracting duplicate public domain sites in which the same shared work is published by comparing the feature points of the object in a duplicate extractor; a reliability determination step of inferring a site providing the most reliable rights management information among the duplicate public domain sites in the site extractor; and a management information providing step of providing rights management information for the shared work based on the rights management information of the site inferred from the copyright manager.

In this case, in the object extraction step, it is preferable to extract an object by processing the image-sharing work as an image signal, and to generate a keyword for the object by labeling the extracted object.

In addition, it is preferable to extract the key points only from the object of the image sharing work having the same keyword as the keyword generated by the object extractor in the key point extraction step.

In addition, in the reliability determination step, it is preferable to infer the site providing the most reliable rights management information by applying a weighted model including the reliability of the site and the reliability of the posts registered in the site.

In addition, the reliability of the site is determined by including any one or more of the number of visits to the site, the number of post registrations, and the update method of the shared work, and the reliability of the post is determined by the number of views, the number of downloads, the date of the updater, and the meta for the shared work. It is preferable that reliability is determined by including any one or more of the number of data information.

In addition, in the management information providing step, creating a rights management information expression system that integrates different domestic and foreign rights management information expression systems into one, and providing rights management information for the shared work with the created rights management information expression system desirable.

As described above, the present invention finds shared works posted on various public domain sites, and provides rights management information of reliable sites among them as common rights management information for shared works of all sites. Accordingly, the user can use the shared work by referring to the latest and reliable rights management information.

1 is a system diagram of a shared work rights management information providing system according to the present invention.
2 is a block diagram illustrating a shared work rights management information providing system according to an embodiment of the present invention.
3 is a diagram illustrating an object classification table according to the present invention.
4 is a diagram illustrating an object extraction image of the present invention.
5 is a diagram illustrating evaluation items for determining the reliability of redundant data according to the present invention.
6 is a diagram illustrating a weighted model for determining the reliability of redundant data according to the present invention.
7 is a diagram showing the RMI integrated expression system of the present invention.
8 and 9 are diagrams showing the conversion state of the RMI integrated expression system for each institution.
10 is a block diagram illustrating a shared work rights management information providing system according to another embodiment of the present invention.
11 is a flowchart illustrating a method for providing shared work rights management information according to the present invention.
12 to 19 show an experimental evaluation table for the present invention.

Hereinafter, a system and method for providing rights management information for shared works according to a preferred embodiment of the present invention will be described in detail with reference to the accompanying drawings.

As shown in FIG. 1 , the shared work rights management information providing system according to the present invention may be configured as a system system including the RMI server 100 , the database 101 and the integrated search server 102 as a preferred embodiment.

The RMI server 100 and the integrated search server 102 may be configured as a web server (WEB) or a WAP server (WAP). In addition, if necessary, it may be composed of a server that integrates them, and some functions may be distributed to an external third server and linked with each other.

The database 101 may also be integrally mounted on the server or independently constructed externally, and may be composed of one or a plurality of databases. Furthermore, the database can also distribute some functions to other servers or databases.

At this time, the RMI server 100 corresponds to an implementation of the system of the present invention as a server, and accesses public domain sites 10 to search and collect image sharing works and store them in the database 101, and the same / Extract sites where similar shared works are published.

In addition, the RMI server 100 determines the most recent or reliable site among the public domain sites 10 on which the shared work is posted, and provides RMI information for the shared work posted on all sites based on this.

The provided RMI information is provided to the user by, for example, the integrated search server 102 after the RMI server 100 or the integrated search server 102 integrates different RMI expression systems to build a unified RMI expression system.

As described above, the present invention finds shared works posted on several public domain sites 10 and provides rights management information of reliable sites among them as common rights management information for shared works of all sites.

Therefore, by using the shared work by referring to the latest and reliable rights management information, the user can increase the understanding of copyright for the shared work and prevent disputes in advance according to the accurate RMI information.

To this end, as shown in FIG. 2, the shared work rights management information providing system according to the present invention includes a web crawler 110, an object extractor 120, a feature point extractor 130, a duplicate extractor 140, a site extractor 150, and a copyright. and a manager 160 .

In addition, when the present invention is configured as a system server such as the RMI server 100, it includes a CPU 171 for establishing the basic functions of the server, a communication interface 172, and other operating system memory.

More specifically, the web crawler 110 automatically searches and indexes various information on the web (WAP), and repeatedly searches for image sharing works according to a method entered in advance of a computer program.

In particular, the web crawler 110 collects image sharing works by accessing a plurality of public domain sites 10 . The public domain site 10 includes all sites on which shared works are published. The collected image sharing work is, for example, stored in the database 101, and preferably, metadata is also collected and stored together.

The image sharing work includes an image such as a photo or video in which the shared work is included. Shared works include works whose protection period has expired, copyrights (including the right to create secondary works) donated, works that can be used under certain conditions (CCL), public works of public institutions, and works with permission for free use (KOGL), etc. includes

The object extractor 120 extracts an object included in the collected image sharing work, and retrieves the image sharing work stored in the database 101 after being collected by the web crawler 110 to extract the object.

In addition, the object extractor 120 generates a keyword for the object by labeling the extracted object after extracting the object by image signal processing of the image sharing work. The generated keyword designates the object.

This object extractor 120 performs object recognition for shared image works using deep learning technology for quick comparison of duplicate image shared works. Since the amount of collected image sharing works is huge, the image sharing works from which the same keyword is extracted based on the keyword of the object are compared.

As an embodiment, deep learning recognizes, labels, and classifies objects using a You Only Look Once (YOLO) algorithm. In order to extract an object from the collected image sharing work, learning about the object is performed first.

For example, object learning is performed based on the PASCAL VOC 2007 data set. Object classification can be divided into people, animals, vehicles, indoors, etc. as shown in FIG. 3, and can be extended to more various objects if necessary. According to this, it can be confirmed that it is classified into four 'horses' as shown in FIG. 4 .

The feature point extractor 130 extracts a feature point from the object of the image sharing work extracted by the object extractor 120 . The extracted feature points characterize the object and may include at least one or more of a color, a shape, and a shape.

As described above, the present invention reduces data throughput and improves processing speed by extracting feature points only from an image sharing work including an object having the same keyword as the keyword of the object generated by the object extractor 120 .

Feature point extraction of an image uses an ORB algorithm (Oriented FAST and Rotated BRIEF) as a preferred embodiment. ORB was developed by the OpenCV lab, and has relatively better computational cost and matching speed.

Therefore, based on the keyword extracted by the YOLO algorithm, only for the image shared work from which the same keyword is extracted, the feature point is extracted through the ORB algorithm, and the duplicate image shared work can be found in the matching module.

The duplicate extractor 140 extracts the duplicate public domain site 10 in which the same shared work is posted by comparing the feature points of the object. The 'duplicate public domain site' refers to sites where the same shared work as the corresponding shared work is posted among the public domain sites 10 .

As such, when the duplicate public domain sites 10 are extracted by the duplicate extractor 140 , the site extractor 150 infers and determines a site that provides the most reliable rights management information among the duplicate public domain sites 10 . do.

The highly reliable public domain site 10 extracted by the site extractor 150 means that the rights management information for the image sharing work provided by the site is the most reliable, and then updates are made based on this.

The site extractor 150 infers the site providing the most reliable rights management information by applying a weighted model including the reliability of the site and the reliability of the posts registered in the site.

That is, when matching data is detected as duplicate data in multiple shared work sites, it is necessary to determine which site is the most accurate data, so the update is performed based on the data determined to be the most reliable through the Weighted Scoring Model.

In order to provide a weighted scoring model technique for determining data reliability in the present invention, an algorithm for evaluating the reliability of the site and the reliability of posts registered in the site as shown in FIG. 5 was developed.

In this case, the reliability of the site is determined by including any one or more of the number of visits to the site, the number of registered posts, and the method of updating shared works. The reliability of a post is determined by including any one or more of the number of views, the number of downloads, the update date, and the number of metadata information for shared articles.

According to the evaluation table of the site for site reliability as shown, for example, the weight is different according to the number of visits to the site and the number of post registrations within the last month, the total number of post registrations, and the registration method.

For evaluation items such as the total number of post registrations and the number of visits to the site within the last month, the maximum value is obtained from the shared work sites, that is, the public domain sites 10, and scores are differentially distributed from the maximum value.

In addition, individual registration, which distinguishes the update method of shared works, is classified into two methods, such as 'Flickr', where the individual author uploads directly to the site, and 'Shared Madang', which uploads with the permission of the administrator.

In the case of receiving approval from the administrator in the update method, if you apply for a change of rights management information in the middle, the reflection may be delayed due to the processing time of the administrator, whereas the method of uploading directly to the site by an individual can always be modified immediately .

Therefore, a higher weight is calculated for the method in which an individual uploads directly among the update methods. However, the method of uploading directly by an individual is limited to the case of posting his or her own work.

In the case of evaluation items of post reliability, an evaluation table is prepared with different weights for items including the number of views, downloads, update date, and the number of meta information about shared works.

In the case of the number of views and downloads, the scores are distributed by comparing the posts of the work sites detected as duplicates, and the score is distributed according to how much meta information about the work is provided.

In this case, when the update date is the latest, it is preferable to give the greatest weight to the reliability score.

As above, each item of the two evaluation item goals was given different weights according to their importance, and the weights assigned to each item are as shown in FIG. 6, and based on this, the score for site reliability and post reliability is calculated as the following <Equation> is expressed

<Equation>

Site Reliability = 30%×nrv + 20%×nrp + 20%×tpr + 30%×ut

Article Reliability = 10%×nv + 25%×dc + 15%×mic + 50%×date

On the other hand, the copyright manager 160 provides a user with rights management information (RMI) based on the rights management information of a site inferred as the most reliable site among a plurality of public domain sites 10 having duplicate data for a shared work. provides

The rights management information may be provided as an example through the integrated search server 102 as described in FIG. 1, but is not limited thereto, and the RMI server 100 or each public domain site 10 in which the present invention is implemented. may be provided through

In particular, the present invention solves a problem that is difficult to understand due to different representation systems of rights management information for each institution, for example, rights management information that integrates different domestic and foreign rights management information representation systems in the copyright manager 160 into one. It creates an expression system and provides rights management information for shared works with the created rights management information expression system.

Accordingly, the present invention provides an integrated RMI integration expression system, thereby eliminating user confusion due to different rights management information expression systems for each shared work site and providing ease of management of the integrated RMI system.

As shown in Figure 7, as an embodiment, in order to integrate the RMI expression body of each site into one, CCL (Creative Common License) type 6 (rmi_01 to rmi_06) is integrated as a basis, and it is applied to freely available works such as expired works and orphan works It is classified as 'rmi_07', a work for free use.

8 and 9 show the integrated RMI to which the present invention is applied for 'Gongong Madang (Korea) and 'Flicker/Europeana', respectively.

As shown, it is integrated into 'FREE USE', rmi_07, in RMI where there is no restriction on using shared works except for the existing CCL format.

In addition, works that are not in the form of images for each shared work site or that users cannot freely use because the copyright holder does not allow the use for other purposes are excluded. Also excluded.

Through this, the RMI expression system provided by domestic 'Sharing Madang' as well as overseas 'Flicker' and 'Europeana' is integrated into one and provided to users. do.

Hereinafter, a method for providing shared copyrighted work rights management information according to the present invention as shown in FIG. 11 will be described with reference to FIG. 10 .

10 is a different representation of a shared work rights management information providing system according to the present invention, and FIG. 11 is a flowchart illustrating a shared work rights management information providing method according to the present invention.

As shown, the method for providing rights management information for shared works according to the present invention includes a work collection step (S110), an object extraction step (S120), a feature point extraction step (S130), a duplicate extraction step (S140), a reliability determination step (S150) ) and a management information providing step (S160).

As described with reference to FIG. 1 , the present invention may be implemented through the RMI server 100 , the database 101 and the integrated search server 102 as an embodiment.

Accordingly, it finds shared works posted on several public domain sites 10, and provides rights management information of reliable sites among them as common rights management information for shared works of all sites.

Therefore, by using the shared work by referring to the latest and reliable rights management information, the user can increase the understanding of copyright for the shared work and prevent disputes in advance according to the accurate RMI information.

To this end, in the works collection step (S110), a plurality of public domain sites 10 are accessed through the web crawler 110 to automatically search for image sharing works, and collect and store them.

The searched public domain site 10 includes all sites where shared works are posted. The collected image sharing work is, for example, stored in the HDFS-based Image Processing DB 101a, and preferably metadata is also collected and stored.

At this time, the image-sharing work includes images such as photos or videos that contain the shared work, and the shared work includes a work whose protection period has expired, a work whose copyright has been donated, a work that can be used under certain conditions (CCL), and a public institution. of public works and copyrighted works (KOGL).

Next, in the object extraction step (S120), an object included in the collected image sharing work is extracted. To this end, the object extractor 120 retrieves the image sharing work stored in the database 101 after being collected by the web crawler 110 to extract the object.

In addition, in the object extraction step (S120), the object extractor 120 generates a keyword for the object by labeling the extracted object after extracting the object by image signal processing of the image sharing work. The generated keyword designates the object.

For quick comparison of image-shared works, object recognition in images is performed using deep learning technology, and since the amount of collected image-shared works is huge, image-shared works are compared based on the keyword of the object.

As an embodiment, deep learning recognizes, labels, and classifies objects using a You Only Look Once (YOLO) algorithm. In order to extract an object from the collected image sharing work, learning about the object is performed first.

Next, in the keypoint extraction step ( S130 ), at least one keypoint characterizing the object extracted from the image sharing work is extracted through the keypoint extractor 130 . A feature point characterizes an object and may include one or more such as color, shape, and shape.

As such, the present invention reduces data throughput and improves processing speed by extracting key points only from an image sharing work that includes an object having the same keyword as the keyword of the object in the key point extraction step (S130).

Feature point extraction of an image uses an ORB algorithm (Oriented FAST and Rotated BRIEF) as a preferred embodiment. ORB was developed by the OpenCV lab, and has relatively better computational cost and matching speed.

Therefore, based on the keyword extracted by the YOLO algorithm, only for the image shared work from which the same keyword is extracted, the feature point is extracted through the ORB algorithm, and the duplicate image shared work can be found in the matching module.

Next, in the duplicate extraction step (S140), the duplicate public domain site 10 on which the same shared work is published is extracted. To this end, the duplicate extractor 140 compares the image sharing work of the public domain site 10 by comparing the feature points of the object.

'Duplicate public domain site' means sites where the same image sharing work as the corresponding shared work is posted among several public domain sites 10, and these site information may be stored in the search engine DB (Search Engine DB, 101b). have.

When the duplicate public domain sites 10 are extracted in this way, in the reliability determination step (S150), the site providing the most reliable rights management information among the duplicate public domain sites 10 is inferred through the site extractor 150 and judge

The highly reliable public domain site 10 means that the rights management information for the image-sharing work provided by the site is the most reliable, and then updates are made based on this.

In the reliability determination step (S150), the site providing the most reliable rights management information is inferred by applying a weighted model including the reliability of the site and the reliability of the posts registered in the site by the site extractor 150.

That is, when matching data is detected as duplicate data in multiple shared work sites, it is necessary to determine which site is the most accurate data, so the update proceeds based on the data determined to be the most reliable through the Weighted Scoring Model.

In this case, the reliability of the site is determined by including any one or more of the number of visits to the site, the number of registered posts, and the method of updating shared works. The reliability of a post is determined by including any one or more of the number of views, the number of downloads, the update date, and the number of metadata information for shared articles.

The site evaluation method for site reliability is, for example, different weights according to the number of site visits and the number of post registrations, the total number of post registrations, and the registration method within the last month.

For evaluation items such as the total number of post registrations and the number of visits to the site within the last month, the maximum value among the shared work sites, that is, the public domain sites 10, is obtained and scores are differentially distributed from the maximum value.

In addition, individual registration, which distinguishes the update method of shared works, is classified into two methods, such as 'Flickr', where the individual author uploads directly to the site, and 'Shared Madang', which uploads with the permission of the administrator.

In the case of receiving approval from the administrator in the update method, if you apply for a change of rights management information in the middle, the reflection may be delayed due to the processing time of the administrator, whereas the method of uploading directly to the site by an individual can always be modified immediately .

Therefore, a higher weight is calculated for the method in which an individual uploads directly among the update methods. However, the method of uploading directly by an individual is limited to the case of posting his or her own work.

In the case of evaluation items of post reliability, an evaluation table is prepared with different weights for items including the number of views, downloads, update date, and the number of meta information about shared works.

In the case of the number of views and downloads, the scores are distributed by comparing the posts of the work sites detected as duplicates, and the score is distributed according to how much meta information about the work is provided.

In this case, when the update date is the latest, it is preferable to give the greatest weight to the reliability score.

Next, in the management information providing step (S160), the rights management information of the site inferred as the most reliable site among the multiple public domain sites 10 having duplicate data for the shared work through the copyright manager 160 is based on to provide rights management information (RMI) to users.

The rights management information may be provided as an example through the integrated search server 102 as described in FIG. 1, but is not limited thereto, and the RMI server 100 or each public domain site 10 in which the present invention is implemented. may be provided through

In addition, the copyright manager 160 creates a rights management information expression system that integrates different domestic and foreign rights management information expression systems into one so as to solve the problem of difficulty in understanding due to the different expression systems of rights management information for each institution, The created rights management information expression system provides rights management information for shared works.

Accordingly, the present invention provides an integrated rights management information expression system, thereby eliminating user confusion due to different rights management information expression systems for each shared work site and providing ease of management of the integrated RMI system.

Accordingly, the present invention integrates the CCL (Creative Common License) type 6 (rmi_01 to rmi_06) as a basis to integrate the RMI expression body of each site into one, and freely available works such as expired works, orphan works, etc. The work is classified as 'rmi_07'.

Hereinafter, through the performance evaluation of the present invention as described above, the difference from the prior art is confirmed. Hereinafter, the present invention integrated the RMI expression system for each site and performed an experimental performance evaluation to know an efficient method through a deep learning-based comparative search technique.

The experimental results of 6000 images were compared between keywords for comparison between image sheets, and images having the same keyword were compared, and the result of finding the same image is shown in FIG. 12 .

For example, when the update operation of the image as shown in FIG. 13 is performed, the same image is detected at site A and site D. In this case, the reliability of each site is obtained by first collecting actual data on the evaluation items for each site and calculating a score for each item.

14 shows numerical values for evaluation items of each site, and FIG. 15 is a table showing reliability scores for each site for sites A to D.

The score for each item was calculated up to 50 points, and weight was applied to the score according to the calculation formula. As the last item, the upload method item, 15 points were calculated for sites B and D where the author directly registers/modifies works.

FIG. 16 shows numerical values for postings on sites A and D in which duplicate images are detected, and FIG. 17 is a table to obtain reliability of bulletin boards using numerical values. The score calculation formula for each evaluation item is the same as the formula for calculating the site reliability above, and additional points are given only to the site with the earliest update date.

In addition, the reliability of the two sites is compared using the previously calculated site reliability and bulletin board reliability. At this time, 40% weight was given to the site reliability and 60% weight was given to the bulletin board reliability.

In the case of site A, the total reliability score is 19.58, and in the case of site D, the total reliability score is 37.31. Therefore, it is determined that the data of site D is more reliable, and the update operation is executed based on the data of site D.

Therefore, it can be confirmed that the rights management information is updated (integrated) from rmi_04 to rmi_02 according to the posting reliability standard as shown in FIG. 19 when the image shared work has different rights management information and is uploaded to a different shared work site.

In the above, specific embodiments of the present invention have been described above. However, the spirit and scope of the present invention is not limited to these specific embodiments, but various modifications and variations are possible within the scope that does not change the gist of the present invention. You will understand when you grow up.

Accordingly, since the embodiments described above are provided to fully inform those of ordinary skill in the art to which the present invention belongs the scope of the invention, it should be understood that they are exemplary in all respects and not limiting, The invention is only defined by the scope of the claims.

100: RMI Server
101: database
102: Unified Search Server
110: web crawler
120: object extractor
130: feature point extractor
140: duplicate extractor
150: site extractor
160: Copyright manager

Claims (12)

a web crawler 110 that accesses a plurality of public domain sites 10 and collects image sharing works;
an object extractor 120 for extracting an object included in the collected image sharing work;
a feature point extractor 130 for extracting at least one feature point characterizing the extracted object;
a duplicate extractor 140 for extracting duplicate public domain sites 10 on which the same shared work is posted by comparing the feature points of the object;
a site extractor 150 for inferring a site providing the most reliable rights management information among the duplicate public domain sites 10; and
A shared work rights management information providing system comprising a; a copyright manager 160 for providing rights management information for the shared work based on the inferred rights management information of the site. According to claim 1,
The object extractor 120,
A shared work rights management information providing system, characterized in that the image-sharing work is subjected to image signal processing to extract an object, and the extracted object is subjected to labeling processing to generate a keyword for the object. 3. The method of claim 2,
The feature point extractor 130,
Shared work rights management information providing system, characterized in that extracting the feature point only for the object of the image sharing work having the same keyword as the keyword generated by the object extractor (120). According to claim 1,
The site extractor 150,
A shared work rights management information providing system, characterized in that the site providing the most reliable rights management information is inferred by applying a weight model including the reliability of the site and the reliability of the posts registered on the site. 5. The method of claim 4,
The reliability of the site is
Reliability is determined including any one or more of the number of visits to the site, the number of postings registered, and the method of updating shared works;
The credibility of this post is
A shared work rights management information providing system, characterized in that the reliability is determined by including any one or more of the number of views, the number of downloads, the update date, and the number of metadata information for the shared work. According to claim 1,
The copyright manager 160,
Shared work rights management information, characterized in that it creates a rights management information expression system that integrates different domestic and foreign rights management information expression systems into one, and provides rights management information for the shared work through the created rights management information expression system delivery system. a work collection step (S110) of accessing a plurality of public domain sites 10 in the web crawler 110 to collect image sharing works;
an object extraction step (S120) of extracting an object included in the collected image sharing work in the object extractor 120;
a feature point extraction step (S130) of extracting at least one feature point characterizing the extracted object in the feature point extractor 130;
a duplicate extraction step (S140) of extracting duplicate public domain sites 10 on which the same shared work is published by comparing the feature points of the object in the duplicate extractor 140;
a reliability determination step (S150) of inferring a site providing the most reliable rights management information among the duplicate public domain sites (10) in the site extractor (150); and
A method of providing rights management information for a shared work, comprising: a management information providing step (S160) of providing rights management information for the shared work based on the rights management information of the site inferred from the copyright manager 160; . 8. The method of claim 7,
In the object extraction step (S120),
A method for providing shared work rights management information, characterized in that the image-shared work is subjected to image signal processing to extract an object, and the extracted object is subjected to labeling processing to generate a keyword for the object. 9. The method of claim 8,
In the feature point extraction step (S130),
A method for providing shared work rights management information, characterized in that the feature point is extracted only for the object of the image sharing work having the same keyword as the keyword generated by the object extractor (120). 8. The method of claim 7,
In the reliability determination step (S150),
A method for providing shared works rights management information, characterized in that by applying a weight model including the reliability of the site and the reliability of the posts registered on the site, the site providing the most reliable rights management information is inferred. 11. The method of claim 10,
The reliability of the site is
Reliability is determined including any one or more of the number of visits to the site, the number of postings registered, and the method of updating shared works;
The credibility of this post is
A method for providing shared work rights management information, characterized in that reliability is determined by including any one or more of the number of views, the number of downloads, the update date, and the number of metadata information for the shared work. 8. The method of claim 7,
In the management information providing step (S160),
Shared work rights management information, characterized in that it creates a rights management information expression system that integrates different domestic and foreign rights management information expression systems into one, and provides rights management information for the shared work through the created rights management information expression system How to provide.

Images