Apparatus and method for providing service for preventing illegal distribution of content

Abstract

The present invention relates to an apparatus and a method for providing a service for preventing illegal distribution of content and, more specifically, to an apparatus and a method for providing a service for preventing illegal distribution of content, which easily detect a website that illegally distributes content through image-based monitoring, and collect and provide evidence data for proving illegal distribution behavior of the corresponding website, thereby supporting eradication of illegal distribution of content. The present invention can support easy and accurate detection of a plurality of illegal sites operated in various languages, regardless of a used language, through image-based illegal site detection and evidence data collection, and support a right holder of content to be protected so that the right holder can easily take appropriate measures against the illegal sites by using the corresponding evidence data, thereby effectively protecting the copyright of the content to be protected by early blocking unauthorized distribution and spread of the content to be protected.

Description

Service provision device and method for preventing illegal distribution of content

[0001] The present invention relates to a service providing device and method for preventing illegal distribution of content, and more particularly, to a service providing device and method for preventing illegal distribution of content, which helps eradicate illegal distribution of content by easily detecting a website that illegally distributes content through image-based monitoring and collecting and providing evidence to prove the illegal distribution activities of the website.

[0002] Recently, various contents such as movies, dramas, and webtoons are being provided through the Internet, and accordingly, websites that provide such contents are supporting users to conveniently watch or subscribe to the contents by accessing them through their terminals.

[0003] Due to this user convenience, the content distributed through online spaces is continuously increasing, and the online content market is continuously revitalized. However, this development of the online content market also has a dark side, revitalizing the illegal distribution market where content is easily obtained online and distributed without permission.

[0004] Therefore, there is an urgent need to establish measures to track down and punish illegal websites that form these illegal distribution markets.

[0005] In order to track such illegal sites, text-based monitoring systems that track illegal sites through content monitoring using text-based keyword searches have emerged. However, these existing monitoring systems have the problem that the accuracy of text matching is very low in the case of content posted as variously modified strings because search results are determined based on the language of the search keyword.

[0006] For example, in order to find illegal overseas websites through monitoring using text-based keyword searches, search keywords must be used in the language of a specific country, which poses the problem of not being able to search for some copyrighted content posted in various languages ​​on various illegal overseas websites.

[0007] In other words, since overseas illegal websites change the titles of content related to copyrighted works posted to the language of the relevant country, it is difficult to accurately identify the copyrighted work based solely on the text on the page where the content is posted.

[0008] To improve this, a video identification system that uses video identification technology to identify illegal videos has recently emerged. This video identification system uses a crawler based on the site's menu and post list to download HTML text and illegal works, and then uses video identification technology to identify what kind of video it is.

[0009] However, these video identification systems also have technical limitations, such as the need to analyze the HTML structure of the site for crawling and set up a crawler specific to the site, and the need to download large amounts of video content through the crawler, which requires a lot of time and facilities.

[0010] In addition, existing video identification systems have the problem that they can only identify video materials registered in the video identification system, which limits the scope of application of identified works, and that excessive time, cost, and effort are required for collecting video works and registering characteristic points.

[0011] In addition, illegal sites that distribute content illegally block access from specific IPs or periodically change the domain address of the illegal site to hinder the collection of evidence. Therefore, even if an illegal site is detected, it is difficult to collect evidence against the illegal site in the future, making it difficult to punish the illegal site.

[0012] The present invention collects original images corresponding to content related to copyrighted works such as movies or webtoons, obtains the domain of a website corresponding to a web page based on a search result of a web page in which an image having the same characteristic points as the original image is posted, selects a domain requiring intensive management that repeatedly posts different content related to copyrighted works, generates a captured image of the web page, and analyzes the captured image based on deep learning to determine whether the website corresponding to the web page is an illegal site for illegally sharing content, and collects evidence data by securing the captured image for the illegal site, thereby enabling easy and accurate detection of illegal sites that share content related to copyrighted works without permission and easy collection of evidence data for such illegal sites.

[0013] A service providing device for preventing illegal distribution of content according to an embodiment of the present invention may include a content collecting unit that collects and stores original images corresponding to protected content in order to detect protected content requiring copyright protection, a feature extraction unit that extracts feature points of the original image to generate feature point information and then stores the feature point information by matching it with the original image, a search unit that obtains a search result list for one or more candidate images similar to the original image through a search engine based on the original image, a detection unit that extracts feature points for each of one or more candidate images included in the search result list in conjunction with the feature point extraction unit and then compares the feature points with the original image to detect candidate images that are similar to the original image by a preset reference value or more as surveillance target images, an evidence collection unit that accesses a web page corresponding to the surveillance target image based on the search result list to generate a captured image by capturing the web page, and generates evidence data information including the captured image and stores the captured image by matching it with the original image, and a judgment unit that analyzes the captured image according to the evidence data information according to a preset algorithm to determine whether a website corresponding to the web page is an illegal site.

[0014] As an example related to the present invention, the search result list may be characterized by including one or more search results each corresponding to one or more candidate images and including address information of a web page corresponding to the candidate images.

[0015] As an example related to the present invention, the judgment unit may include a learning model in which a correlation between an image and a type of a website is learned, and may be characterized in that the captured image is applied as an input to the learning model to derive a type of a website corresponding to the captured image through the learning model, and if the derived type of the website is a type of a preset illegal website, the website corresponding to the address information according to the evidence information is determined to be an illegal website.

[0016] As an example related to the present invention, the service providing device may further include a site management unit that obtains address information corresponding to the surveillance target image based on the search result list in conjunction with the detection unit, obtains domain information from the address information and accumulates the number of acquisitions of the domain information, selects a website corresponding to the domain information as a surveillance target site when the number of acquisitions of the domain information corresponding to the address information is greater than or equal to a preset reference value, transmits query information for querying the type of the surveillance target site, sets the type of the surveillance target site based on response information received in response to the query information, and generates learning data in which a captured image corresponding to the surveillance target image and a type according to the response information are mutually matched in conjunction with the evidence collection unit, and the determination unit may be characterized in that it trains the learning data generated through the site management unit to the learning model.

[0017] As an example related to the present invention, the site management unit may be characterized in that it stores evidence data information including address information including a domain address according to the domain information by matching the domain information, and for each of one or more pieces of domain information, it generates and provides statistical information including at least one of the number of evidences by content type, traffic information of a website corresponding to the domain information, and language type corresponding to the domain information based on one or more pieces of evidence data information matching the domain information.

[0018] As an example related to the present invention, the evidence collection unit may be characterized by accessing the web page through a preset VPN server.

[0019] As an example related to the present invention, the evidence collection unit may be characterized in that it generates the captured image based on one or more images generated by capturing while scrolling a web page corresponding to the surveillance target image according to a preset evidence collection algorithm.

[0020] As an example related to the present invention, the service providing device may further include an information providing unit that, when detecting an illegal site using evidence data information corresponding to the original image in conjunction with the judgment unit, generates notification information based on the evidence data information and the original image, and then transmits the notification information to a user terminal that has registered the original image.

[0021] As an example related to the present invention, the service providing device may further include a blockchain linkage unit that stores the evidence data information in a blockchain.

[0022] A service providing method for preventing illegal distribution of content of a service providing device according to an embodiment of the present invention may include the steps of collecting and storing original images corresponding to protected content in order to detect protected content requiring copyright protection, extracting feature points of the original image to generate feature point information and then matching and storing the information with the original image, obtaining a search result list for one or more candidate images similar to the original image through a search engine based on the original image, extracting feature points for each of one or more candidate images included in the search result list in conjunction with the feature point extraction unit, and then detecting candidate images similar to the original image by a preset reference value or more as surveillance target images through a comparison of feature points with the original image, accessing a web page corresponding to the surveillance target image based on the search result list to generate a captured image by capturing the web page, generating evidence data information including the captured image, and matching and storing the captured image with the original image, and analyzing the captured image according to the evidence data information according to a preset algorithm to determine whether a website corresponding to the web page is an illegal website.

[0023] The present invention detects a web page of a website that has an image similar to an original image that constitutes protected content requiring copyright protection posted through image-based analysis, and captures the web page as it is to easily collect evidence data, and furthermore, identifies the type of website corresponding to the captured image only through a deep learning-based model that has learned the correlation between the type of illegal website that distributes content without permission and the image, and easily determines whether the identified website type corresponds to an illegal website, and if it corresponds to an illegal website, notifies the rights holder of the protected content that illegal sharing of the protected content is taking place along with evidence data for sanctioning the illegal website. Thus, through such image-based illegal website detection and evidence data collection, it is possible to easily and accurately detect a large number of illegal websites operating in various languages ​​regardless of the language used, and it is possible to easily support the rights holder of the protected content to take appropriate measures against the illegal website using the evidence data. Thus, the unauthorized distribution and spread of the protected content can be blocked early, and the copyright of the protected content can be effectively protected.

[0024] In addition, the present invention can easily collect the history of illegal sharing of protected content through bypass access or capture-based evidence acquisition even when an illegal site continuously changes its access address to avoid collecting evidence or blocks access from a specific IP, thereby neutralizing such IP blocking and continuous changes in access addresses, thereby supporting easy legal action against illegal sites.

[0025] In addition, the present invention can support efficient detection of illegal sites by continuously training a deep learning model on the types of images and illegal sites, thereby providing robust identification performance for content posted on various illegal sites using different languages.

[0026] Figure 1 is a configuration diagram of a service provision system for preventing illegal distribution of content according to an embodiment of the present invention.

[0027] Figure 2 is a detailed configuration diagram of a service providing device for preventing illegal distribution of content according to an embodiment of the present invention.

[0028] Figures 3 to 9 are operation examples of a service providing method for preventing illegal distribution of content of a service providing device according to an embodiment of the present invention.

[0029] Hereinafter, detailed embodiments of the present invention will be described with reference to the drawings.

[0030] Figure 1 is a configuration diagram of a service provision system for preventing illegal distribution of content according to an embodiment of the present invention.

[0031] As illustrated, the service providing system according to the present invention may include one or more user terminals that register content subject to protection related to a work requiring copyright protection, and a service providing device (100) that communicates with the one or more user terminals through a communication network and monitors a website that illegally shares or distributes the content subject to protection.

[0032] At this time, the communication network described in the present invention may include a wired / wireless communication network, and examples of such wireless communication networks include Wireless LAN (WLAN), Digital Living Network Alliance (DLNA), Wireless Broadband (Wibro), World Interoperability for Microwave Access (Wimax), Global System for Mobile communication (GSM), Code Division Multi Access (CDMA), Code Division Multi Access 2000 (CDMA2000), Enhanced Voice-Data Optimized or Enhanced Voice-Data Only (EV-DO), Wideband CDMA (WCDMA), High Speed ​​Downlink Packet Access (HSDPA), High Speed ​​Uplink Packet Access (HSUPA), IEEE 802.16, Long Term Evolution (LTE), Long Term Evolution-Advanced (LTE-A), Wireless Mobile Broadband Service (WMBS), 5G mobile communication service, This may include Bluetooth, LoRa (Long Range), RFID (Radio Frequency Identification), Infrared Data Association (IrDA), Ultra Wideband (UWB), ZigBee, Near Field Communication (NFC), Ultra Sound Communication (USC), Visible Light Communication (VLC), Wi-Fi, and Wi-Fi Direct.Additionally, wired communication networks may include wired Local Area Networks (LANs), wired Wide Area Networks (WANs), Power Line Communications (PLCs), USB communications, Ethernet, serial communications, and optical / coaxial cables.

[0033] In addition, the user terminal may include various terminals such as a smart phone, a portable terminal, a mobile terminal, a personal digital assistant (PDA), a personal computer, a notebook computer, a slate PC, a tablet PC, an ultrabook, etc., having a communication function.

[0034] In addition, the service providing device (100) may be configured as a server, and the service providing device (100) may communicate with various devices and external servers in addition to the user terminal through a communication network.

[0035] Based on the above-described configuration, the service providing device (100) according to an embodiment of the present invention searches for a web page containing an image similar to an image of the protected content related to a work registered by a user for copyright protection through image-based analysis, captures the searched web page to collect evidence data, and analyzes the type of the website corresponding to the searched web page using the captured image, and if the website is an illegal website that illegally uses the protected content, provides information and evidence data on the website, thereby enabling accurate detection of illegal websites and easily providing data for punishing illegal websites. This will be described in detail with reference to the drawings below.

[0036] First, FIG. 2 is a detailed configuration diagram of a service providing device (100) for preventing illegal distribution of content according to an embodiment of the present invention.

[0037] As illustrated, a service providing device (100) according to an embodiment of the present invention may be configured to include a communication unit (110), a storage unit (120), a control unit (130), etc.

[0038] At this time, the communication unit (110) can communicate with one or more user terminals, various external devices, external servers, etc. through the communication network.

[0039] In addition, the storage unit (120) can store various types of information, and the storage unit (120) can be configured in various forms such as an HDD (Hard Disk Drive), an SSD (Solid State Drive), etc., and can be configured as a DB or can include one or more DBs.

[0040] For example, the storage unit (120) may be configured to include a content DB in which information related to content to be protected is stored, a collection DB in which information related to evidence data is stored, a member DB in which one or more member information corresponding to one or more user terminals is stored, etc.

[0041] At this time, the content DB, evidence DB, member DB, etc. may each be configured as individual database servers and configured to be linked (communicated) with the service providing device (100).

[0042] In addition, the control unit (130) performs the overall control function of the service providing device (100), and the control unit (130) may include RAM, ROM, CPU, GPU, and a bus, and the RAM, ROM, CPU, GPU, etc. may be connected to each other through a bus.

[0043] Additionally, the communication unit (110) and storage unit (120) may be configured to be included in the control unit (130).

[0044] In addition, the control unit (130) can perform the overall control function of the service providing device (100).

[0045] In addition, as illustrated, the control unit (130) may be configured to include a content collection unit (131), a feature extraction unit (132), a search unit (133), a detection unit (134), an evidence collection unit (135), a judgment unit (136), a site management unit (137), an information provision unit (138), etc.

[0046] At this time, it goes without saying that at least one of the components constituting the control unit (130) may be included in another component.

[0047] In addition, the plurality of components constituting the control unit (130) may be implemented by a processor capable of data processing, and each may be implemented separately by a different processor, or may be functionally separated within one processor.

[0048] In addition, the control unit (130) or at least one of the various components constituting the control unit (130) can communicate with various terminals or devices such as user terminals, various external devices, external servers, etc. through the communication unit (110), and the communication configuration through the communication unit (110) will be omitted below.

[0049] Based on the above-described configuration, an operation for performing a service providing method for preventing illegal distribution of content of a service providing device (100) according to the present invention is described with reference to FIGS. 3 and 4.

[0050] First, the content collection unit (131) can collect protected content information (or content registration information) including an original image corresponding to the protected content in order to detect protected content requiring copyright protection from each of one or more user terminals and store the collected information in the content DB.

[0051] At this time, the protected information may further include content identification information of the protected content, rights holder identification information (or user identification information) of the rights holder (or user) who owns the protected content, etc.

[0052] Alternatively, the content collection unit (131) may automatically collect original images through web crawling for general websites such as portal sites or information provision sites, and may create protected information based on the original images and store the information in the content DB.

[0053] At this time, the original image described in the present invention means an image corresponding to the content to be protected, and may also mean a representative image representing the content to be protected.

[0054] Examples of such original images include various images that can be used, such as poster images for movie content, still images for video content or webtoon content, cover images for book content, and cover images for webtoon content.

[0055] In addition, the content types of the protected content described in the present invention may include various content types such as movies, dramas, entertainment, webtoons, books, etc.

[0056] That is, the service providing device (100) according to the present invention can monitor regardless of the type of content.

[0057] In addition, the feature point extraction unit (132) can extract feature points of the original image included in the protected information stored in the content DB to generate feature point information, and can match the feature point information with the original image or the protected information and store it in the content DB.

[0058] At this time, the feature point information may include content identification information corresponding to the original image from which the feature point was extracted, feature point data for the feature point extracted from the original image, etc.

[0059] In addition, the feature point extraction unit (132) can extract feature points from an image through a preset image analysis algorithm.

[0060] In addition, the search unit (133) may apply the original image included in the protected information as a search query (or query image) to a search engine preset to a search engine or a search engine provided by an external server through communication with an external server, and obtain a search result list corresponding to the search query from the search engine through the search engine.

[0061] At this time, the search engine can search various websites using the search query, and generate a search result list including one or more search results that include candidate images similar to the original image according to the search query.

[0062] As an example of such a search engine, the search engines of 'NAVER' or 'GOOGLE' can be used, and when the search engine is included in the service providing device (100), execution data related to the search engine can be stored in the storage unit (120) of the service providing device (100).

[0063] At this time, the search engine may be preset in the search unit (133).

[0064] In addition, the search result list may be configured in the form of a web page, and may include one or more search results (or search result information) that include candidate images similar to the original image and address information (or URL (uniform resource locator) address) of web pages corresponding to the candidate images.

[0065] Additionally, the search unit (133) can provide the search result list to the detection unit (134).

[0066] In addition, the detection unit (134) can extract feature points from each of one or more candidate images corresponding to one or more search results included in the search result list in conjunction with the search unit (133) and the feature point extraction unit (132), and then compare the feature points with the feature points according to the feature point information corresponding to the original image to analyze the similarity with the original image, and detect (select) a candidate image whose similarity is higher than a preset standard value as a surveillance target image.

[0067] That is, the feature point extraction unit (132) can extract feature points from the candidate image according to the request (control) of the detection unit (134), and the detection unit (134) can detect (set) a candidate image that is similar to the original image by a preset standard value or more as a surveillance target image through a feature point comparison with the original image for each of one or more candidate images included in the search result list.

[0068] At this time, the detection unit (134) can detect one or more surveillance target images among one or more candidate images corresponding to a specific original image.

[0069] Meanwhile, as illustrated in FIG. 5, the evidence collection unit (135) can capture a web page corresponding to the surveillance target image based on the search result list in conjunction with the detection unit (134) to create a capture image.

[0070] For example, the evidence collection unit (135) can access a web page of a website corresponding to the surveillance target image based on a search result corresponding to the surveillance target image detected by the detection unit (134) in the search result list.

[0071] At this time, if the website accessed by the above-mentioned evidence collection unit (135) is an illegal website for illegally sharing content, a method of blocking access to IPs (internet protocols) of a specific country can be used to avoid surveillance or tracking, so the above-mentioned evidence collection unit (135) can access the web page of the website corresponding to the surveillance target image through a preset VPN (virtual private network) server.

[0072] That is, the above-mentioned evidence collection unit (135) can communicate with the external VPN server, receive an IP address, and access the web page of the website corresponding to the surveillance target image.

[0073] Additionally, the web page accessed by the above-mentioned evidence collection unit (135) may include the image of the surveillance target.

[0074] In addition, the above-mentioned evidence collection unit (135) can capture a web page containing the surveillance target image.

[0075] At this time, the evidence collection unit (135) can capture a web page corresponding to the surveillance target image while scrolling according to a preset evidence collection algorithm to create one or more images, and can create a capture image based on the one or more images created.

[0076] For example, the evidence collection unit (135) can generate the captured image by combining one or more of the generated images.

[0077] In addition, the above-described evidence collection unit (135) can change the language of the web page corresponding to the surveillance target image to a preset reference language according to the above-described evidence collection algorithm, and generate a capture image based on the web page changed to the reference language as described above.

[0078] In addition, data related to an execution program, such as a web browser for executing the web page, can be stored in the storage unit (120), and the web page can be executed through the execution program to generate the captured image.

[0079] In addition, the evidence collection unit (135) can generate evidence data information including the captured image.

[0080] At this time, the evidence data information may include content identification information of the protected content corresponding to the original image used as a search query to obtain the surveillance target image, address information (detection address) of the web page including the surveillance target image, search date and time (detection date and time) corresponding to the search result list, the surveillance target image (or detected image), etc.

[0081] In addition, the above-mentioned evidence collection unit (135) can store the above-mentioned evidence data information in the above-mentioned evidence DB, and can also store the information in the above-mentioned content DB by matching it with the protected information stored in the above-mentioned content DB corresponding to the original image used as the above-mentioned search query.

[0082] Meanwhile, the judgment unit (136) can analyze the captured image according to the evidence data information in conjunction with the evidence collection unit (135) according to a preset algorithm to determine whether the website corresponding to the web page is an illegal website.

[0083] For example, the judgment unit (136) may apply a captured image included in the evidence data information when the evidence collection unit (135) generates the evidence data information to a learning model (or deep learning model) preset (included) in the judgment unit (136) to derive the type of a website that includes a web page corresponding to the captured image through the learning model.

[0084] At this time, the learning model can learn the correlation between the type of image and website.

[0085] Additionally, the learning model (or deep learning model) may be composed of a deep learning algorithm, and the deep learning algorithm may be composed of one or more neural network models.

[0086] In addition, the neural network model (or neural network) described in the present invention may be composed of an input layer, one or more hidden layers, and an output layer, and various types of neural networks such as a deep neural network (DNN), a recurrent neural network (RNN), a convolutional neural network (CNN), a generative model, etc. may be applied to the neural network model.

[0087] Accordingly, the judgment unit (136) applies the captured image included in the evidence data information as input to the learning model, and outputs the type of the website corresponding to the captured image through the learning model, and if the type of the website thus output is a type of a preset illegal site (matches the type of the illegal site), the website corresponding to the address information according to the evidence data information can be judged as an illegal site.

[0088] To this end, the judgment unit (136) may be preset with a plurality of different website types (or website types) and judgment criteria information on whether or not a website is illegal by website type, and the judgment unit (136) may extract the address information from the evidence data information including the captured image applied as input to the learning model when the type of the website calculated through the learning model to correspond to the captured image is a website type (or type of illegal site) corresponding to illegality according to the judgment criteria information, and may determine that the website corresponding to the extracted address information is an illegal site that shares or distributes protected content corresponding to the original image without permission.

[0089] At this time, the types of websites learned by the learning model may be limited to two types, such as illegal site types and legal site types, or two types, such as whether it is illegal or not, and accordingly, the judgment unit (136) may apply the captured image as input to the learning model and output the illegal site type or legal site type as the type of website output through the learning model, or output whether it is illegal or not.

[0090] In addition, the judgment unit (136) may output information about at least one of the probability that the type of website corresponding to the captured image is an illegal site type and the probability that the type of website corresponding to the captured image is a legal site type through the learning model, or may output an infringement degree for the probability that the website corresponding to the captured image infringes on the protected content corresponding to the captured image through the learning model.

[0091] At this time, the above-mentioned degree of infringement may be the probability that it is the type of illegal site.

[0092] Accordingly, the judgment unit (136) can determine whether the website corresponding to the captured image is an illegal site (whether it is an illegal site or not) based on the output information output through the learning model without the judgment reference information.

[0093] Alternatively, the above judgment criterion information may set a criterion for determining an illegal site, and the judgment unit (136) may determine that the website corresponding to the captured image is an illegal site if the probability of the illegal site type or the degree of infringement calculated through the learning model is greater than the preset criterion according to the judgment criterion information.

[0094] In addition, the judgment unit (136) may generate judgment result information including the evidence information when it determines that the website corresponding to the captured image is an illegal site.

[0095] At this time, the judgment result information may further include site verification result information regarding whether the website corresponding to the captured image is an illegal site (illegal or not).

[0096] In addition, the judgment unit (136) may generate judgment result information including the evidence information even if the website corresponding to the captured image is not an illegal site.

[0097] In addition, the information providing unit (138) may, in conjunction with the judgment unit (136), extract an original image corresponding to the evidence information from the content DB when the website corresponding to the evidence information is an illegal site based on the judgment result information when generating the judgment result information, and may generate notification information based on the extracted original image and the evidence information included in the judgment result information.

[0098] In addition, the information provision unit (138) can transmit the notification information to a user terminal that has registered the original image extracted from the content DB or the protection target information corresponding to the extracted original image.

[0099] At this time, the information provision unit (138) can identify member information corresponding to the content information in the member DB based on the content information corresponding to the extracted original image, and can identify the user terminal that is the transmission target (or reception target) of the notification information based on the identified member information.

[0100] In addition, the evidence collection unit (135) may include an original image corresponding to the evidence collection data information in the evidence collection data information, and the information providing unit (138) may generate the notification information based on the evidence collection data information when the original image is included in the evidence collection data information.

[0101] Accordingly, as illustrated in FIG. 6, a user terminal that has received the notification information can display the notification information to enable a user (rights holder) who has the copyright for the content corresponding to the extracted original image to take action against a website determined to be an illegal site.

[0102] At this time, the information provision unit (138) may transmit the notification information to a preset administrator terminal.

[0103] Meanwhile, in the above-described configuration, the service providing device (100) may include a site management unit (137), and the site management unit (137) may track and manage websites where different original images are continuously posted, and perform operations to increase the accuracy and reliability of website type judgment of the learning model included in the judgment unit (136), which will be described in detail with reference to FIG. 7.

[0104] First, the site management unit (137) can obtain address information from a search result corresponding to the surveillance target image based on a search result list including the surveillance target image when the surveillance target image is detected by the detection unit (134) in conjunction with the detection unit (134).

[0105] In addition, the site management unit (137) can obtain domain information from the acquired address information according to a preset domain acquisition rule, and store the domain information in the evidence DB.

[0106] For example, if the address information corresponding to the surveillance target image is configured as https: / www.torrentxxx.com / m / view.php?b_id=1111, the site management unit (137) can obtain https: / www.torrentxxx.com as domain information from the address information according to a preset domain acquisition rule.

[0107] Alternatively, the site management unit (137) may compare the acquired address information with other address information included in each of one or more pieces of evidence data information stored in the evidence DB, and may identify duplicate texts among the texts included in the acquired address information that overlap with the acquired address information and one or more pieces of other address information and whose number of overlaps is greater than a preset standard number of times, and may acquire the identified duplicate texts as domain information.

[0108] At this time, the above domain information may mean a domain address.

[0109] In addition, as illustrated in FIG. 8, the site management unit (137) can, in conjunction with the detection unit (134), accumulate the number of acquisitions (or number of detections or number of detections) of existing domain information stored in the evidence DB that matches the acquired domain information each time domain information is acquired from a plurality of different search result lists.

[0110] In addition, the site management unit (137) can select a website corresponding to the specific domain information as a site to be monitored if the number of acquisitions of the specific domain information corresponding to the target image of the monitored image is greater than a preset standard.

[0111] In addition, the site management unit (137) can transmit query information for querying the type (website type) of the site to be monitored to a preset administrator terminal that manages the service providing device (100).

[0112] In addition, the site management unit (137) can receive response information in response to the query information from the administrator terminal, and set the type of the site to be monitored based on the response information.

[0113] For example, the site management unit (137) can match type information for a website type corresponding to a specific surveillance target site set by the administrator terminal according to the response information with domain information corresponding to the specific surveillance target site and store the information in the evidence DB.

[0114] At this time, the type of the site to be monitored may be one of multiple different types of websites or information on whether the site to be monitored is illegal.

[0115] In addition, the site management unit (137) can obtain a capture image corresponding to a specific surveillance target image in conjunction with the evidence collection unit (135), and when receiving response information corresponding to the specific surveillance target image, can generate learning data including the type (type information) of the surveillance target site according to the received response information and the capture image corresponding to the specific surveillance target image in a mutually matched state.

[0116] At this time, the site management unit (137) can generate learning data that includes the type information and the captured image corresponding to the specific surveillance target image in a mutually matched state, if the number of acquisitions of specific domain information corresponding to the surveillance target image is greater than a preset standard and the type information matching the specific domain information is already stored in the evidence DB.

[0117] In addition, the judgment unit (136) can train the learning model with learning data generated through the site management unit (137) in conjunction with the site management unit (137), thereby allowing the learning model to learn various images and types of various websites, and continuously train the learning model so that the type of website corresponding to the image can be accurately analyzed and identified.

[0118] For example, the judgment unit (136) may train the learning model with learning data including captured images of web pages included in illegal sites (illegal sharing websites) such as torrent sites or file sharing sites and the types of the corresponding illegal sites in a mutually matched state so that the structure of web pages related to illegal sites shown in the captured images is learned by the learning model, and may train the learning model with learning data including captured images of web pages included in legal sites such as portal sites or information providing sites (e.g., blogs, SNS (Social Network Service) related sites) and the types of the corresponding legal sites in a mutually matched state so that the structure of web pages related to different legal sites shown in the captured images is learned by the learning model, so that the learning model can robustly identify the types of illegal sites from the types of legal sites by using only images.

[0119] At this time, the service providing device (100) or the storage unit (120) may be configured to further include a learning DB in which the learning data is stored, and the site management unit (137) may store the learning data in the learning DB when generating the learning data, and the judgment unit (136) may extract the learning data stored in the learning DB and apply the learning data to the learning DB.

[0120] In addition, the judgment unit (136) may be preset with judgment criteria information in which one or more types of websites judged as illegal sites are set as described above, and the site management unit (137) may, in conjunction with the judgment unit (136), add the type of website included in the response information to the judgment criteria information if the type of website included in the response information does not exist in the judgment criteria information.

[0121] At this time, the above judgment criteria information may be preset to one or more types of websites corresponding to illegal sites.

[0122] Alternatively, the above judgment criteria information may be set together with one or more types (website types) that correspond to (belong to) an illegal site and one or more types (website types) that do not correspond to (belong to) an illegal site, and in this case, the response information may include type setting information (or category information) regarding whether the website type included in the response information corresponds to an illegal site.

[0123] For example, the judgment criterion information may be set to include one or more types belonging to a first category related to the illegal site and one or more types belonging to a second category related to a legal site that is not the illegal site, and the site management unit (137) may set the additional target type as a type belonging to the first category in the judgment criterion information when the type of the website according to the response information is an additional target type that is an additional target type in the judgment criterion information and when the additional target type is confirmed to belong to the first category according to the type setting information included in the response information.

[0124] Accordingly, the site management unit (137) can add a website included in the response information to the judgment criteria information based on the type setting information included in the response information, and set whether the type of the added website is an illegal site in the judgment criteria information.

[0125] In addition, the site management unit (137) may identify at least one piece of evidence data information including domain information of the surveillance target site and address information including the domain information in the evidence DB when the surveillance target site is an illegal site according to the type of the surveillance target site according to the response information and the type of illegal site (web site type) preset in the judgment criterion information, and store the identified at least one piece of evidence data information in the evidence DB so as to be matched with the domain information.

[0126] At this time, the site management unit (137) can, in conjunction with the evidence collection unit (135), identify domain information corresponding to address information included in the evidence collection data information in the evidence collection DB each time the evidence collection unit (135) generates evidence data information, and store the generated evidence data information in the evidence collection DB by matching it with the identified domain information.

[0127] For example, the site management unit (137) can store evidence data information including address information including a domain address according to the domain information in the evidence DB by matching it with the domain information.

[0128] In addition, as illustrated in FIG. 9, the site management unit (137) can generate statistical information on the number of evidences for one or more preset content types (content types) based on information stored in the evidence DB and information stored in the content DB, and one or more website information corresponding to the type of illegal site in the evidence DB and one or more domain information corresponding to each of the information.

[0129] Accordingly, the site management unit (137) can generate the statistical information and transmit it to the user terminal or administrator terminal when receiving request information for the statistical information from the user terminal or administrator terminal.

[0130] In addition, as illustrated in FIG. 8, the site management unit (137) can access a website corresponding to the domain information to identify the language type (language code) used in the website, or collect traffic information of the website, and after generating additional information including the identified language type or traffic information, can set the additional information to the domain information or match it with the domain information and store it in the evidence DB, and can provide the additional information by including it in the statistical information.

[0131] In addition, when the information providing unit (138) receives request information for the statistical information from a user terminal or an administrator terminal, it requests the statistical information from the site management unit (137), and the site management unit (137) can generate the statistical information and provide it to the information providing unit (138). The information providing unit (138) can generate the statistical information in the form of a web page and transmit it to the user terminal or administrator terminal that transmitted the request information.

[0132] Meanwhile, the control unit (130) of the service providing device (100) may further include a blockchain linkage unit (139) that links with a block chain, and the blockchain linkage unit (139) links with a block chain (or blockchain system) that can guarantee the integrity and security of data, and whenever data is stored in the evidence DB, a block containing the data stored in the evidence DB is created and stored in the blockchain (or blockchain system), thereby preventing falsification of the data stored in the evidence DB.

[0133] For example, the blockchain linkage unit (139) can generate a blockchain-based block including the generated information whenever information to be stored in the evidence DB, such as the domain information, evidence data information, and the additional information, is generated, and can transmit the block to a preset blockchain system and store it in the blockchain (or blockchain system).

[0134] At this time, the blockchain described in the present invention means a blockchain algorithm that organizes a plurality of transaction information into blocks and connects multiple blocks like a chain using a hash, and the blockchain system may mean a plurality of nodes that execute the blockchain algorithm to jointly manage information requiring security by distributing and storing blocks on digital devices of nodes that are participants (peers) of the blockchain in a peer-to-peer (P2P) network distributed environment.

[0135] To this end, the communication unit (110) configured in the service providing device (100) can communicate with one or more nodes constituting the blockchain system.

[0136] Additionally, the service providing device (100) may be configured as any one of a plurality of nodes constituting the blockchain system.

[0137] In addition, the above-mentioned evidence DB may be configured as a shared ledger used for the blockchain when the service providing device (100) is configured as a node constituting a blockchain system.

[0138] Through the above-described configuration, the service providing device (100) according to the present invention can increase the security and integrity of data stored in the evidence DB.

[0139] As described above, the present invention detects web pages of websites that post images similar to original images constituting protected content requiring copyright protection through image-based analysis, and captures the web pages as they are to easily collect evidence data, and furthermore, identifies the type of website corresponding to the captured image only through a deep learning-based model that has learned the correlation between the type of illegal website distributing content without permission and the image, and easily determines whether the identified website type corresponds to an illegal website, and if it corresponds to an illegal website, notifies the rights holder of the protected content that illegal sharing of the protected content is taking place along with evidence data for sanctioning the illegal website. Thus, through such image-based illegal website detection and evidence data collection, it is possible to easily and accurately detect a large number of illegal websites operating in various languages ​​regardless of the language used, and it is possible to easily support the rights holder of the protected content to take appropriate measures against the illegal websites using the evidence data. Thus, the unauthorized distribution and spread of the protected content can be blocked early, and the copyright of the protected content can be effectively protected.

[0140] In addition, the present invention can easily collect the history of illegal sharing of protected content through bypass access or capture-based evidence acquisition even when an illegal site continuously changes its access address to avoid collecting evidence or blocks access from a specific IP, thereby neutralizing such IP blocking and continuous changes in access addresses, thereby supporting easy legal action against illegal sites.

[0141] In addition, the present invention can support efficient detection of illegal sites by continuously training a deep learning model on the types of images and illegal sites, thereby providing robust identification performance for content posted on various illegal sites using different languages.

[0142]

[0143] The components described in the embodiments of the present invention may be implemented using one or more general-purpose computers or special-purpose computers, such as hardware such as a memory, a processor, a controller, an arithmetic logic unit (ALU), a digital signal processor, a microcomputer, a field programmable gate array (FPGA), a programmable logic unit (PLU), a microprocessor, software including an instruction set, or a combination thereof, or any other device capable of executing and responding to instructions.

[0144] The various devices and components described herein may be implemented using hardware circuitry (e.g., CMOS-based logic circuitry), firmware, software, or a combination thereof. For example, they may be implemented using transistors, logic gates, and electronic circuits in various electrical structures.

[0145] Those skilled in the art will appreciate that modifications and variations of the above-described content may be made without departing from the essential characteristics of the present invention. Therefore, the embodiments disclosed herein are intended to illustrate, rather than limit, the technical spirit of the present invention, 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 construed according to the following claims, and all technical ideas within the scope equivalent thereto should be construed as being included within the scope of the present invention.

[0146] Description of the symbol

[0147] 100: Service providing device 110: Communication unit

[0148] 120: Storage unit 130: Control unit

[0149] 131: Content Collection Unit 132: Feature Extraction Unit

[0150] 133: Search Unit 134: Detection Unit

[0151] 135: Evidence Collection Department 136: Judgment Department

[0152] 137: Site Administration Department 138: Information Provision Department

[0153] 139: Blockchain Integration

Claims

1. A content collection unit that collects and stores original images corresponding to protected content to detect protected content requiring copyright protection; A feature point extraction unit that extracts feature points of the original image, creates feature point information, and then matches it with the original image and stores it; A search unit that obtains a search result list for one or more candidate images similar to the original image through a search engine based on the original image; A detection unit that extracts feature points in conjunction with the feature point extraction unit for each of one or more candidate images included in the search result list, and then detects candidate images that are similar to the original image by a preset standard or more as surveillance target images through a comparison of the feature points with the original image; An evidence collection unit that accesses a web page corresponding to the surveillance target image based on the search result list, creates a captured image by capturing the web page, and creates evidence data information including the captured image and stores it in a match with the original image; and A judgment unit that analyzes the captured image based on the above evidence information according to a preset algorithm and determines whether the website corresponding to the above web page is an illegal website. A service providing device for preventing illegal distribution of content including .

2. In claim 1, A service providing device for preventing illegal distribution of content, characterized in that the search result list includes one or more search results each corresponding to one or more candidate images and including address information of a web page corresponding to the candidate images.

3. In claim 1, A service providing device for preventing illegal distribution of content, characterized in that the judgment unit includes a learning model in which a correlation between an image and a type of a website is learned, applies the captured image as an input to the learning model to calculate a type of website corresponding to the captured image through the learning model, and if the type of the calculated website is a type of a preset illegal website, determines the website corresponding to the address information according to the evidence information as an illegal website.

4. In claim 3, A site management unit that obtains address information corresponding to the surveillance target image based on the search result list in conjunction with the detection unit, obtains domain information from the address information and accumulates the number of acquisitions of the domain information, selects a website corresponding to the domain information as a surveillance target site when the number of acquisitions of the domain information corresponding to the address information is greater than or equal to a preset reference value, transmits query information for querying the type of the surveillance target site, sets the type of the surveillance target site based on response information received in response to the query information, and generates learning data in which a captured image corresponding to the surveillance target image and a type according to the response information are mutually matched in conjunction with the evidence collection unit, A service providing device for preventing illegal distribution of content, characterized in that the judgment unit trains the learning data generated through the site management unit on the learning model.

5. In claim 4, A service providing device for preventing illegal distribution of content, characterized in that the site management unit stores evidence data information including address information including a domain address according to the domain information by matching the domain information, and, for each of one or more pieces of domain information, generates and provides statistical information including at least one of the number of evidences collected by content type, traffic information of a website corresponding to the domain information, and language types corresponding to the domain information based on one or more pieces of evidence data information matching the domain information.

6. In claim 1, A service providing device for preventing illegal distribution of content, characterized in that the above-mentioned evidence collection unit accesses the web page through a preset VPN server.

7. In claim 1, A service providing device for preventing illegal distribution of content, characterized in that the above-mentioned evidence collection unit generates the captured image based on one or more images generated by capturing while scrolling a web page corresponding to the surveillance target image according to a preset evidence collection algorithm.

8. In claim 1, A service providing device for preventing illegal distribution of content, characterized in that it further includes an information providing unit that, when detecting an illegal site using evidence data information corresponding to the original image in conjunction with the judgment unit, generates notification information based on the evidence data information and the original image, and then transmits the notification information to a user terminal that has registered the original image.

9. In claim 1, A service providing device for preventing illegal distribution of content, characterized in that it further includes a blockchain linkage that stores the above-mentioned evidence data information in a blockchain.

10. In a service provision method for preventing illegal distribution of content of a service provision device, A step of collecting and storing original images corresponding to protected content to detect protected content requiring copyright protection; A step of extracting feature points of the original image, generating feature point information, and then matching it with the original image and storing it; A step of obtaining a search result list for one or more candidate images similar to the original image through a search engine based on the original image; A step of extracting feature points in conjunction with the feature point extraction unit for each of one or more candidate images included in the search result list, and then detecting a candidate image that is similar to the original image by a preset standard value or more as a surveillance target image through a feature point comparison with the original image; A step of accessing a web page corresponding to the surveillance target image based on the search result list, creating a captured image by capturing the web page, and creating evidence data information including the captured image and storing it in a match with the original image; and A step of analyzing the captured image according to the above evidence information according to a preset algorithm to determine whether the website corresponding to the above web page is an illegal website. A method of providing a service to prevent illegal distribution of content including .

Images