KR20230100220A - Copyright sales brokering method for literary and dramatic works using computing apparatus - Google Patents

Abstract

The present invention relates to a transaction mediation method of literary works IP using a computer device, and more specifically, based on natural language processing, transaction mediation information is automatically converted by converting a sales work and a purchase request into the same summary information metadata. It relates to a method for creating and brokering IP transactions and a system implementing the same. The IP transaction mediation method according to the present invention may be operated by a processor, wherein the processor receives at least one piece of work information for sale from a seller of a work IP, and the processor requests at least one purchase request from a buyer of the work IP. Receiving information, extracting, by the processor, first summary information from the sales work information; extracting, by the processor, second summary information from the purchase request information; 2 calculating a story evaluation index based on correlation between summary information; calculating, by the processor, a matching score between the sales work information and the purchase request information based on the story evaluation index; and, by the processor, the IP seller and transmitting information about the matching score to one or both of the IP purchasers.

Description

Transaction brokerage method of literary works IP using computer device {COPYRIGHT SALES BROKERING METHOD FOR LITERARY AND DRAMATIC WORKS USING COMPUTING APPARATUS}

The present invention relates to a transaction mediation method of literary works IP using a computer device, and more specifically, based on natural language processing, transaction mediation information is automatically converted by converting a sales work and a purchase request into the same summary information metadata. It relates to a method for creating and brokering IP transactions and a system implementing the same.

The distribution of cultural works based on various creative stories is on the rise. In particular, in recent years, primary works announced by individual authors have been produced as secondary works of various types, resulting in commercial success. For example, for the purpose of commercializing a literary work by visualizing it as a secondary work, the intellectual property (IP) market for primary works is expanding.

Conventionally, producers promoting the commercialization of secondary works had various difficulties in the process of searching for primary works or their copyright holders. Due to the vast amount of published primary works, in IP transactions, a platform in charge of distribution of the primary works or a content supplier (CP) intervenes as an intermediary between the producer of secondary works and the writer of primary works. was frequent. However, such an intermediary distribution structure of works restricts the diversity of search for works from the perspective of producers who want to produce secondary works, and writers of primary works are forced to enter into unfavorable or unfair contracts in relation to IP supply conditions, etc. used to cause problems.

In particular, from the standpoint of the primary author, I want to directly expose my copyrighted IP to various secondary copyright producers without going through an intermediary, and more safely protect my IP-related revenue from infringement due to the existence of an intermediary. We expect service experiences that are difficult to expect in the current distribution structure, such as conducting market research directly on producers of secondary works and reflecting them in commercial creative activities.

Nevertheless, as mentioned above, since the amount of primary works is vast and the number of producers of secondary works that want to acquire the copyright of the primary works is not small, there is a limit that there is no choice but to have an entity that mediates IP transactions between the two. there is.

As a conventional solution in this area, there was a technology that recommends primary works in a similar way to products. These technologies have been used in places such as bookstores, movie theaters, and video content platforms for consumers or subscribers who directly consume primary works. In the prior art, analyzes such as analyzing the consumption patterns of a large number of consumers and assigning weight to the number of consumers with a high number of subscriptions or analyzing the continuous selection of consumers and recommending related works have been utilized. In addition, a method of enabling search for the properties of a work by assigning a tag to the primary work was utilized.

However, in order to recommend primary works by analyzing consumers' consumption patterns, results must be derived from the record of a number of consumers already consuming copyrighted works, so the accuracy of recommendations targeting consumers who have newly entered the market or platform has been low. . In addition, it was difficult to respond to those who focused on selecting carefully selected primary works without the premise of mass consumption, such as producers of secondary works. After all, even using the prior art, there was a difficulty in excluding the intervention of intermediaries in the transaction of copyrighted works.

The present invention is intended to automate the transaction mediation process of literary works in order to overcome the limitations of the prior art described above. The essential limitation of the prior art comes from the vast amount of primary works. Direct transactions between writers and producers are virtually impossible due to the existence of a vast amount of literary works, and it is difficult to form a reasonable market without the intervention of full-time mediators. The above-mentioned unfair treatment by CP and platform is also largely due to the inevitability of these mediators. Accordingly, an object of the present invention is to minimize intermediary intervention in such mediation procedures and to automate most of the procedures.

In order to solve the above problem, a method for brokering a transaction of literary works IP (Intellectual Property) using a computer device including a processor according to an aspect of the present invention, wherein the processor obtains at least one sales work information from a seller of the work IP Receiving an input, the processor receiving at least one piece of purchase request information from a purchaser of a work IP, the processor extracting first summary information from the sales work information, the processor extracting second summary information from the purchase request information Extracting summary information; Calculating, by the processor, a story evaluation index based on correlation between the first summary information and second summary information; calculating a matching score between requested information, and transmitting, by the processor, information about the matching score to one or both of the IP seller and the IP buyer; can include

The first summary information and the second summary information may be copyrighted metadata including at least one item among content type, genre, subject, person, incident, and background, and a story information value corresponding to each item. .

In the step of calculating the story evaluation index, the processor calculates at least one matching rate by comparing information values assigned to each item for each of at least one item constituting the first summary information and the second summary information. and configuring, by the processor, a story evaluation index corresponding to the match rate for each of the items.

The method in which the processor compares the information values specified in the respective items with each other, reads the first information value specified in the first item of the first summary information, and reads the first information value specified in the first item in the second summary information. 2 information values are read, the first information value and the second information value are input to the natural language processing device to calculate the agreement rate in the word vector space, and the agreement rate in the word vector space is the agreement rate for the first item. , and the process may be performed by the processor for each item.

The selling work information includes desired price information set by the seller for the selling work, the purchase request information includes expected price information that the buyer is willing to pay for any selling work, and the matching score is , the matching score obtained by merging the above story evaluation indices using one or more calculation methods among summation, average, weighted sum, and weighted average, the number of user views of the sales works, user evaluation scores, user review content, and the volume or playback time of the sales works The processor may weight-add a popularity score based on at least one and a price score determined based on a comparison between the desired price information and the price of the expected price information.

The weighted summing method of the matching scores by the processor may include multiplying the matching score by a first percentage weight, the popularity score by a second percentage weight, and the price score by a third percentage weight, respectively, and summing the results. it could be

The transmitting of information on the matching score by the processor may include generating, by the processor, recommended price information of a work for sale based on the desired price information and expected price information; A step of transmitting together with the matching score may be further included.

The step of transmitting, by the processor, information about the matching score may include: transmitting, by the processor, a plurality of matching points calculated between one piece of sales work information and a plurality of pieces of purchase request information; The method may further include at least one of transmitting a plurality of matching scores each calculated between the sales work information and the piece of purchase request information.

At least one of the first summary information and the second summary information may be extracted by a natural language processing device.

The transaction mediation method of literary works IP using a computer device including the processor may further include, by the processor, reading at least one of the sales work information and the first summary information from a database. there is.

A computer device including a processor having a transaction mediation function of literary works IP according to an aspect of the present invention for solving the above-mentioned problems, a first input unit for receiving at least one piece of sales information from a seller of a work IP, a work A second input unit for receiving at least one piece of purchase request information from a purchaser of an IP, a first extractor for extracting first summary information from the sales work information, and a second extractor for extracting second summary information from the purchase request information. , an index calculation unit that calculates a story evaluation index based on the correlation between the first summary information and the second summary information, and a matching calculator that calculates a matching score between the sales work and the purchase request information based on the story evaluation index , and a transmission unit for transmitting information about the matching score to one or both of the IP seller and the IP buyer.

The first summary information and the second summary information,

It may be copyrighted metadata including at least one item among content type, genre, subject, person, event, and background, and a story information value corresponding to each item.

The index calculation unit calculates the information value for each item based on at least one matching rate calculated by comparing information values assigned to each item for each of the at least one item constituting the first summary information and the second summary information. It may be configured to output a story evaluation index corresponding to the matching rate.

The index calculation unit compares the information values designated for each item with each other by reading the first information value designated for the first item of the first summary information, and reading the value specified for the first item in the second summary information. The second information value is read, the first information value and the second information value are input to the natural language processing device to calculate the matching rate in the word vector space, and the matching rate in the word vector space is calculated for the first item. The matching rate may be determined, and the above procedure may be performed by the processor for each item.

The selling work information includes desired price information set by the seller for the selling work, the purchase request information includes expected price information that the buyer is willing to pay for any selling work, and the matching score is , the matching score obtained by merging the above story evaluation indices using one or more calculation methods among summation, average, weighted sum, and weighted average, the number of user views of the sales works, user evaluation scores, user review content, and the volume or playback time of the sales works The matching calculation unit may weight-add a popularity score based on at least one and a price score determined based on a comparison of amounts of the desired price information and the expected price information.

The matching calculation unit may be configured to multiply the matching score by a first percentage weight, the popularity score by a second percentage weight, and the price score by a third percentage weight, and then sum the results to output the matching score. can

The matching calculation unit may be configured to generate recommended price information of a work for sale based on the desired price information and expected price information, and the transmission unit to transmit the recommended price information together with the matching score.

The transmission unit may include at least one of a plurality of matching scores calculated between a piece of work for sale information and a plurality of pieces of purchase request information, and a plurality of matching scores respectively calculated between a plurality of pieces of work for sale information and one piece of purchase request information. It can be configured to transmit one more.

At least one of the first extraction unit and the second extraction unit may be operated by a natural language processing device.

The computer device including a processor having a transaction mediation function of the literary work IP may further include a database storing at least one of the sales work information and the first summary information.

A computer readable recording medium according to an aspect of the present invention for solving the above problems includes instructions that can be executed in a processor, and the instructions, when executed in a processor, cause the processor to obtain at least from a seller of a work IP Receiving one sales work information, receiving at least one purchase request information from a buyer of the work IP, extracting first summary information from the sales work information, extracting second summary information from the purchase request information, A story evaluation index is calculated based on the correlation between the first summary information and the second summary information, a matching score between the sales work information and the purchase request information is calculated based on the story evaluation index, and the IP seller and It may be configured to transmit information about the matching score to one or both of the IP purchasers.

By automating the mediation process of IP transactions through metadata of literary works, it is possible to maximize the efficiency of mediation while minimizing the area in which intermediaries can intervene. Through this, a more direct IP transaction can be concluded between the primary author and the secondary author, resulting in beneficial results for both parties.

1 is a conceptual diagram summarizing the basic concept of the present invention;
2 is a conceptual diagram showing the summary informationization process of literary works according to the present invention and the IP transaction mediation method based on the process;
3 is a conceptual diagram of a process of deriving a story evaluation index through comparison of information values for each item of first summary information and second summary information;
4 is an example of transaction brokerage information corresponding to a plurality of purchase request information for a plurality of sales work information;
5 is a conceptual diagram of a computer device including a processor for implementing a system according to an embodiment of the present invention;
6 is a conceptual diagram showing an embodiment of a common format for sales work information and purchase request information;
7 is a conceptual diagram illustrating an operation of a natural language processing device for calculating a matching degree.

Since the present invention can make various changes and have various embodiments, specific embodiments are illustrated in the drawings and described in detail. However, this is not intended to limit the present invention to specific embodiments, and should be understood to include all modifications, equivalents, and substitutes included in the spirit and scope of the present invention.

Terms such as first and second may be used to describe various components, but the components should not be limited by the terms. These terms are only used for the purpose of distinguishing one component from another. For example, a first element may be termed a second element, and similarly, a second element may be termed a first element, without departing from the scope of the present invention. The term “and/or” includes any combination of a plurality of related listed items or any of a plurality of related listed items.

It is understood that when an element is referred to as being "connected" or "connected" to another element, it may be directly connected or connected to the other element, but other elements may exist in the middle. It should be. On the other hand, when an element is referred to as “directly connected” or “directly connected” to another element, it should be understood that no other element exists in the middle.

Terms used in this application are only used to describe specific embodiments, and are not intended to limit the present invention. Singular expressions include plural expressions unless the context clearly dictates otherwise. In this application, the terms "include" or "have" are intended to designate that there is a feature, number, step, operation, component, part, or combination thereof described in the specification, but one or more other features It should be understood that the presence or addition of numbers, steps, operations, components, parts, or combinations thereof is not precluded.

Unless defined otherwise, all terms used herein, including technical or scientific terms, have the same meaning as commonly understood by one of ordinary skill in the art to which the present invention belongs. Terms such as those defined in commonly used dictionaries should be interpreted as having a meaning consistent with the meaning in the context of the related art, and unless explicitly defined in the present application, they should not be interpreted in an ideal or excessively formal meaning. don't

When the term "device" or "unit" is used in describing the invention in this application, it is intended to describe an embodiment showing any function of the present invention that can be implemented as a device or functional unit, and necessarily the corresponding function It does not imply that it must be implemented as a single, independent device or function. For example, one device may be implemented with a plurality of devices performing the same function, or conversely, one device may be installed to simultaneously perform the functions of a plurality of devices. Functions of certain devices or functional units may be implemented by other devices or functional units through software means or by general computers and information processing devices. In addition, a plurality of devices or functional units may be connected to each other only through an information communication network and separated from each other in a physical space. Since this is an area of various embodiments that can be taken by a person skilled in the art in the information processing technology field to implement the same technical idea, any detailed implementation method should be construed as being included in the technical idea area of the present invention.

Hereinafter, with reference to the accompanying drawings, preferred embodiments of the present application will be described in more detail. In order to facilitate overall understanding in the description of the present invention, the same reference numerals are used for the same components in the drawings, and redundant descriptions of the same components are omitted.

The present invention has an object to automate the process of brokering transactions of literary works. To this end, in the present invention, summary information in a standardized metadata form is extracted from literary works posted for sale, while summary information in the same format is also extracted from purchase request information of producers who want to acquire the copyright of such literary works. By extracting, it is possible to automatically compare a large number of works and purchase requests.

1 is a conceptual diagram summarizing the basic concept of the present invention. The sales work information 220 is information provided by a copyright holder of a literary work. The sales work information 220 may include the content body of the sales work or summary information thereof. Meanwhile, producers who want to produce secondary works by obtaining copyrights provide purchase request information 225 . The purchase request information 225 may include a sentence describing the original work being sought, a document such as a proposal, and a search query.

Both the sales work information 220 and the purchase request information 225 may be documents written in natural language. Here, a natural language is a concept contrasted with an artificial language, and means that it is not emotionalized by any artificial rule or language system, but is written by a written or colloquial language used daily. FIG. 6 is an exemplary diagram illustrating an embodiment of a common format of sales work information 220 and purchase request information 225. Referring to FIG. The "story set" shown in FIG. 6 is an information sheet that can be prepared and submitted directly by a copyright owner or seller who wants to sell a work for sale or by a creator of a secondary work who wants to acquire copyright. In order to achieve the object of the present invention, it is necessary to compare similarities between these natural language-based summary information using an automated method.

It is very difficult to compare documents written in natural language due to their linguistic characteristics. For example, it can be judged by human intuition that a certain literary work and a certain manufacturer's plan describe the same content, but it is very difficult to determine similarity by mechanically comparing the originals of the two document information (100). Since the same meaning can be expressed using different types of sentences or words, a separate technical measure beyond simple one-to-one comparison of information is required.

In the present invention, sales work information 220 is converted into first summary information 250 and purchase request information 225 is converted into second summary information 255, respectively. Preferably, each conversion process may be operated by a natural language processing device implemented in the same manner. Also preferably, the first summary information 250 and the second summary information 255 may be metadata formats having the same format. Since the natural language processing device may have a function of automatically summarizing a large amount of language content and extracting its gist, when the metadata of the same form is extracted by inputting two works to the same natural language processing device, the two pieces of information are easily compared. (110) can be.

In the process of comparison 110, each metadata may be compared in units of semantic sets. For example, if the 'genre' field of the metadata included in the first summary information 250 defines the genre of a work for sale as 'fantasy', this degree corresponds to the 'genre' field included in the second summary information 255. can be compared directly with the metadata of In this case, each metadata field may be compared by a numerical judgment according to the degree of match, rather than a yes/no decision on whether or not the fields match. For example, when the value of the 'genre' field metadata described in the second summary information 255 is 'fantasy' the same as that of the first summary information 250, it can be judged to be completely matched, while the corresponding When the value is 'SF', it can be determined that there is a point that is semantically similar to 'fantasy' and therefore partially coincides.

Determination of partial match as described above may also be performed through a natural language processing device. Natural language processing techniques can be used to determine the degree of semantic similarity of certain words to other words. For example, a conventional methodology such as Word2Vec converts a word into vector space information using an artificial intelligence learning system, and compares the difference between vectors pointed by two words in the vector space to determine the degree of association between words. Provides a tool that can be measured by automatic means. In the present invention, the degree of association between words can be identified using the aforementioned Word2Vec or other algorithms having similar functions, and comparison can be performed based on the degree.

As described above, if the method of mutually collating information changed in the form of standardized summary information is used, a large amount of sales work information and a large amount of purchase request information can be compared at high speed, and natural language processing as described above can be performed. In the case of passing through a semantic similarity judgment based on the semantic similarity, it can be derived by quantifying the degree of match, not whether or not it matches. After going through this process, it is obvious that if a high matching rate is confirmed between a specific sales work and a specific purchase request, it is possible to mediate copyright transactions between each other based on this. Through the present invention, it is possible to mediate transactions by efficiently connecting demand and supply of literary works IP using such an automated method.

2 is a conceptual diagram showing the summary information processing of literary works according to the present invention and the IP transaction mediation method based thereon.

In the literary works IP transaction mediation system 200, a seller 210 and a buyer 215 are assumed as participants in the transaction. The seller 210 is an author and/or a copyright holder of a literary work, and means a person who possesses an IP of a primary work and intends to sell it to others. The seller 210 may be the author of the literary work himself or may be a third party entrusted with IP transaction affairs from the author himself. The purchaser 215 is a person who wants to purchase the IP of a literary work, and means a person who intends to purchase the IP of a primary work under a predetermined contract condition in order to produce a secondary work. For example, the buyer 215 may be an adaptation creator, a video product producer, or a content supply company (CP).

The seller 210 uploads the sales work information 220 to the IP transaction mediation system 200 . The sales work information 220 may be a set of information including an original copy of a literary work targeted for IP sales, that is, a sales work, and related supplementary information. Preferably, the sales work information 220 includes original data of the sales work, summary information such as the title, author, genre, background, event, person, tag of the sales work, and desired price information set by the seller 210 for the sales work. It may be an information set including at least one of In addition, the IP transaction intermediary system 200 may receive and utilize a plurality of sales work information 220 from a plurality of sellers 210 .

The sales writing information 220 is input to the system 200 through the first input unit 230 . The sales work information 220 input to the first input unit 230 is transferred to the first extraction unit 240 . The first extractor 240 converts the sales work information 220 into first summary information 250 .

The first extractor 240 may use the natural language processing device 238 to extract the first summary information 250 from the sales work information 220 . The natural language processing device 238 is a device operated by a computer processor having a natural language interpretation capability, and may be driven by artificial intelligence for interpretation.

According to an embodiment of the present invention, the interpretation artificial intelligence of the natural language processing device 248 may be implemented to operate based on a pre-learning language model. Preferably, a BERT (Bidirectional Encoder Representations from Transformers) series model, more specifically, a model having a synonym discrimination function such as "XML-RoBERTA-paraphrase" may be used. The model is specialized for a sentence analysis task, and may be a model in which fine-tuning has been performed in advance with data specialized for a related task. As an example, the conventional model presented in a paper published by Conneau, A., etc. in 2019 through arXiv may be used, but is not limited thereto, and any language model capable of achieving the purpose of natural language processing of the present invention may be used. Even if used, it will be included in the present invention.

According to the learning model as described above, the natural language processing device 238 can identify semantic or intrinsic information of an input sentence and, if necessary, perform an operation of analyzing similarities.

For natural language processing artificial intelligence learning of the natural language processing device 238, initial learning may be performed. For initial learning, an unstructured learning database that matches the original text of a previously known literary work with information summarizing it can be used.

The unstructured learning database may include, for example, summary information summarizing the body of a literary work published by a publisher and the title, author, genre, background, event, person, tag, and plot of the literary work. The summary information as described above may be used as learning data for the natural language processing device 238 of the present invention to automatically convert the text of a literary work into summary information.

Alternatively, the unstructured learning database may include a data set and a corpus based on Korean daily conversations collected through SNS or other means, and data capable of understanding the context of words. The above data can be used as learning data for the natural language processing device 238 of the present invention to compare synonyms or synonyms and to derive the degree of semantic agreement between them.

In one embodiment of the present invention, the unstructured learning database is for learning the Korean pre-learning language model of the natural language processing device 238, and may be a machine learning parallel corpus dataset derived from SNS usage data. As an example, the above dataset is a machine learning translation parallel corpus dataset corresponding to about 100 languages provided by Facebook AI Research presented in a paper published through arXiv by Conneau, A., etc. in 2019. It is okay to use .

The unstructured learning database may be preprocessed and input for refinement. The preprocessing process may be a process of applying word meaning correction, typo correction, spelling correction, etc. to data that the native language processing device cannot normally process. In the pre-processing process, paragraph and sentence separation, numbers or English expressions, special characters or emoticons, punctuation marks, etc. are converted into the same form or removed from the work to be studied using a regular expression-based refinement program. It can strengthen the data uniformity of the original of the literary work that becomes the

In the preprocessing process, a tokenizing procedure for pre-segmenting sentences may be applied in order to further improve the uniformity of data to be learned. The natural language processing device 238 may build a vocabulary knowledge body (Vocabulary) by segmenting through methods such as spacing between words of a given natural language sentence, morpheme analysis of parts of speech, and byte pair encoding.

The natural language processing device 238 trained as described above may be configured to use either or both of a deep neural network and a shallow neural network. For example, a process of extracting summary information by analyzing the original text of a literary work by a deep neural network, and a process of determining the semantic agreement rate of words by a word matching method including Word2Vec by a shallow neural network are configured to be performed. It can be.

The first summary information 250 extracted by the first extractor 240 using the natural language processing device 248 is in the form of copyrighted work metadata in a standardized format used in the system 200 of the present invention, , at least one of content type, genre, subject, person, event, and background, and a story information value corresponding to each item. Among the above items, the type of content may enumerate type information of a content medium in which a secondary work is desired to be produced through a work for sale. For example, as shown in the first summary information example 250a of FIG. 3, the first summary summarized in the form of "content type: webtoon, TV drama, movie; genre: romance; event: love; background: beach" The information 250 may express that the corresponding sales content information 220 is a romance genre work that is desired to be adapted into a webtoon, TV drama, or movie, and is a love story that takes place on the beach.

Preferably, the first summary information 250 may be stored and utilized in the sales writing database 260 . As described above, a plurality of sellers 210 may participate in the IP transaction brokerage system 200 at the same time. The seller 210 provides the sales work information 220 to the IP transaction intermediary system 200 at any desired time, and the system 200 converts the information into the form of first summary information 250 and the sale After accumulating in the sales works database 260 together with the work information 220, at any time, preferably at the time when the purchase request information 225 of the buyer 215 is entered, the sales works database 260 Intermediation of IP transactions may be attempted by using the stored information on a plurality of works for sale 220 and the first summary information 250 .

Meanwhile, the purchaser 215 uploads purchase request information 225 to the IP transaction mediation system 200 . The purchase request information 225 is information containing content about the work IP that the buyer 210 wants to acquire for the purpose of producing a secondary work, a search query in the form of words or sentences to search for the work, and a purchase request. At least among the expected price information on the amount that the buyer 215 is willing to pay for the purchase of summary information such as genre, background, event, person, tag, etc. of the literary work desired by the buyer, and IP purchase of any work for sale It may be an information set including one. In addition, the IP transaction brokerage system 200 may receive and utilize a plurality of purchase request information 225 from a plurality of buyers 215 .

The purchase request information 225 is input to the system 200 through the second input unit 235 . The purchase request information 225 input through the second input unit 235 is transmitted to the second extraction unit 245 . The second extractor 245 converts the purchase request information 225 into second summary information 255 .

The second extractor 245 may use the natural language processing device 238 to extract the second summary information 255 from the purchase request information 225 . The natural language processing device 238 is a device operated by a computer processor having a natural language interpretation capability, and may be driven by artificial intelligence for interpretation.

Also, preferably, it may be a device that is learned by the same learning data as the natural language processing device 238 called by the first extractor 240 to extract the first summary information 250 and operates in the same way. . This is because the metadata format of the first summary information 250 and the second summary information 255 should be the same as possible so that the IP transaction mediation process to be performed by the matching unit 270 is easy, but it must be limited to this implementation. It is not, and if it is for the purpose of achieving the object of the present invention by smooth matching, the natural language processing device 238 called by the first extraction unit 240 and the second extraction unit 245 each have different learning data And / or may operate in a way.

The second summary information 255 extracted by the second extractor 245 using the natural language processing device 248 is in the form of copyrighted work metadata in a standardized format used in the system 200 of the present invention, , at least one of content type, genre, subject, person, event, and background, and a story information value corresponding to each item. Among the above items, content type may enumerate type information of content media for which the purchaser 215 promotes production of a derivative work. For example, as in the second summary information example 255a of FIG. 3, second summary information summarized in the form of "content type: TV drama, movie; genre: romance; event: love; background: mountain" ( 255) may express that the corresponding purchase request information 225 is searching for a work containing a love story taking place in the mountains in the romance genre for adaptation into webtoons, TV dramas, movies, and the like.

The second summary information 255 may be generated in response to the purchase request information 225 of the buyer 215 being received, but in another embodiment of the present invention, after being generated in advance, the purchase request database 265 ) and can be used. As described above, a plurality of buyers 215 may participate in the IP transaction brokerage system 200 at the same time. The buyer 215 provides the purchase request information 220 to the IP transaction brokerage system 200 in advance, and the system 200 converts the information into the form of second summary information 255 and the purchase request information 225 After accumulating in the purchase request database 265 together with ), a plurality of purchases stored in the purchase request database 265 at an arbitrary point in time, preferably when the seller 215 requests matching with purchase request information. IP transaction mediation may be attempted using the request information 225 and the second summary information 255 .

As described above, when at least one of the first summary information 250 and the second summary information 255 is prepared, the matching unit 270 may perform an IP transaction mediation procedure based on this. The operation of the matching unit 270 can be largely divided into an index calculation unit 272 and a matching calculation unit 274.

The index calculation unit 272 may calculate a story evaluation index based on the correlation between the first summary information 250 and the second summary information 255 . Preferably, the first summary information 250 and the second summary information 255 may be in the same metadata format. Accordingly, information values assigned to each item of metadata constituting each piece of summary information can be compared with each other.

A process of comparing the information value according to an embodiment of the present invention will be described in more detail. First, the first information value specified in the first item of the first summary information 250 is read, and the second information value specified in the same first item is read in the second summary information 255. That is, information is read from each summary information for the same item. Then, a degree of agreement between the first information value and the second information value may be determined. At this time, the degree of matching may be determined by simply determining whether the first information value and the second information value are the same, but preferably, the first information value and the second information value are input to the natural language processing device 248. And, it may be determined by inputting the words constituting each information value into the natural language processing device 248 and calculating the degree to which the meanings of each other coincide.

In particular, according to an embodiment of the present invention, the calculation of the degree of matching may be implemented by a method of projecting the first information value and the second information value into a word vector space by the natural language processing device 248 and then comparing them. can A word vector space is a concept used in artificial intelligence-based natural language processing methods such as Word2Vec. A word is defined as a vector representing a predetermined meaning through correlation with other words, and this vector is defined as a vector representing another word within the vector space. By comparing with vectors, it means a method of determining the degree of congruence of word meaning or appropriateness in context through mathematical calculation.

In one embodiment of the present invention, the operation of the natural language processing apparatus for calculating the degree of matching may be performed by a procedure shown in FIG. 7 . The first information value 710 and the second information value 715 may be tokenized by being input to a tokenizer that pre-divides sentences to further improve data uniformity (720). Each of the tokenized information values 730 and 735 may be input to the pre-learning language model 740 and converted into sentence vectors 750 and 755 and then subjected to mutual comparison processing. As described above, the pre-learning language model may be a BERT-based model. An indexing table 760 may be used in the vector analysis process for the mutual comparison process. In the mutual comparison process, a semantic/intrinsic similarity score by comparing the use of a language model between input sentences and an expressive/explicit similarity score by a degree in common between tokens extracted through lexical segmentation are determined, respectively, and the two similarities are determined. The overall similarity score can be obtained by deriving the harmonic mean of the scores.

The embodiment of the comparison process corresponds to one of several embodiments for the implementation of the present invention, and any other natural language processing algorithm that can compare the degree of semantic, contextual, and expressive matching of two different words. Alternatively, an implementation method may be used.

The above comparison procedure may be repeatedly performed for each item by the index calculation unit 272. Through this, information values of all items belonging to each of the first summary information and the second summary information may be compared. This comparison result can be formed as new metadata indicating the degree of agreement for each item, which is referred to as a story evaluation index in the present invention.

The process of deriving the story evaluation index through the comparison of information values for each item of the first summary information 250 and the second summary information 255 described above by the index calculation unit 272 will be briefly described with reference to FIG. . Since the description below is only an artificial case made by simplifying the comparison process of the information value conceptually, FIG. 3 and the description below do not limit the present invention, and the same items, information values, and comparisons as illustrated It does not imply that results, and resultant figures, must be derived.

In FIG. 3 , the first summary information example 250a is “content type: webtoon, TV drama, movie; Genre: Romance; event: love; Background: beach", and the second summary information example 255a assumes a case of content type: webtoon, TV drama, movie; genre: romance; event: romance; background: mountain". By comparing (300) each item of the two exemplary summary information (250a, 255a), the index calculator (272) tries to derive the story evaluation index (310).

In the first summary information example 250a and the second summary information example 255a, when the information values of the "content type" items are read and compared with each other (300), the information values listed in the two items are "TV drama, It can be seen that the "movie" matches. In this case, the content type desired by the seller 210 through the sales work information 220 and the content type desired by the buyer 215 through the purchase request information 225 match at least partially, and the transaction is highly likely to be concluded. can be considered Accordingly, this item may be determined to be “matched” between the two pieces of summary information, and the details thereof may be included in the story evaluation index 310 and output.

If the information values of the "genre" items are read and compared with each other (300) in the first summary information example 250a and the second summary information example 255b, respectively, it can be confirmed that the information values of the two items are the same as "romance". can In this case, it may be determined that the information value of the corresponding item matches "100%", and this may be included in the story evaluation index 310 and output.

In the first summary information example 250a and the second summary information example 255b, if the information values of the "event" item are read and compared (300), the information values of the two items are "love" and "love", respectively. It can be confirmed that In this case, it can be seen that the information values of the corresponding items are relatively similar in meaning, although they are different in terms of the word. In this case, the similarity in meaning may be obtained according to a concordance rate between vectors calculated by a mathematical method in a word vector space using Word2Vec or the like by the natural language processing unit 248 as described above. According to the similarity determination result, when the agreement rate between the two words is determined to be "70%", this may be included in the story evaluation index 310 and output.

On the other hand, when the information values of the "background" item are compared in the same way as above, it can be seen that the meanings of "mountain" and "sea" are relatively different. In this case, the matching rate in the above word vector space may be determined to be relatively low, eg "30%", and this may be included in the story evaluation index 310 and output.

In the example of FIG. 3 , values constituting the story evaluation index 310 are presented in the form of "consistent/unmatched" and percentage. However, the present invention is not limited to this form, and various techniques capable of measuring the degree of agreement or similarity between two pieces of summary information can be applied. For example, in another embodiment of the present invention, the story evaluation index may be expressed as an integer score out of 5 points for each item.

As in the above example, the index calculation unit 272 may generate a story evaluation index by comparing the first summary information 250 and the second summary information 255 with each other. Based on the story evaluation index, the matching calculation unit 274 may calculate a matching rate between the first summary information 250 and the second summary information 255 . The matching rate is a score for the correlation between the two pieces of information, and when the matching rate is high, it can be determined that the correlation between the work for sale and the purchase request is high. Based on the matching rate, the seller and the buyer can identify information on the counterparty that is closest to information on works for sale or purchase request information uploaded by them.

According to an embodiment of the present invention, the method of calculating the matching rate in the matching calculation unit 274 is the matching score, popularity score, and price corresponding to one sales work information 220 and purchase request information 225. It may be to calculate three points, such as a score, and then add them up in weight.

The match score may be a combination of values representing the degree of match for each item of the story evaluation index calculated by the index calculation unit 272 . Preferably, the matching score may be a total score obtained by summing all individual item values of the story evaluation index. However, the merging method is not necessarily limited to this, and in addition to simple summation, summation with weights applied to each item is applied, arithmetic or geometric averages are applied, weights are applied for each item and average is applied, or other values are merged to form a single product. Various mathematical means for converting into scores may be used, and the mathematical means may be arbitrarily selected according to the needs of those who wish to practice the present invention.

The popularity score is a score calculated based on information that can be provided through the sales work information 220 . The calculation process of the popularity score may include various indices indicating the public popularity of the work for sale. For example, if a work for sale is a literary work published online, the number of online views of the work for sale may be included. If the sales work is a published literary work, the sales volume of the publication or popularity ranking of the seller may be included. Ratings given by viewers or readers to sales works may be included. Unstructured popularity data such as ratings, impressions, reviews, and user-specified hashtags written by viewers or readers for sales works may also be included, and such unstructured popularity data may be additionally processed by the natural language processing unit 248. , and may be processed in a manner compared with the first summary information 250 or the second summary information 255 .

The popularity score may be normalized and/or merged with the above-listed or other indexes related to the popularity of sales works. According to an embodiment of the present invention, the popularity score may be a score obtained by normalizing at least one popularity index value based on a percentage and then averaging the at least one normalized popularity index value. The normalization method may be to obtain an upper percentage of the value of the selling work information 220 currently being processed among the same popularity index values of all the selling writing information stored in the selling writing database 260, but is not necessarily limited thereto. . The merging method is not limited to the above, and simple summation or summation with weights applied for each item is applied, arithmetic or geometric average is applied, weight is applied for each item and averaged, or other values are merged by item. A variety of mathematical means can be used to convert to a single score. The mathematical means for processing the normalization and the merging can be arbitrarily selected according to the needs of those who wish to practice the present invention.

The price score is the desired price information set by the seller 210 for the work for sale and described in the product for sale information 220, and the amount the buyer 215 is willing to pay for the IP purchase of any work for sale. It can be determined based on the degree to which the price ranges of each other match each other by comparing the expected price information described in the purchase request information 225 by determining .

According to an embodiment of the present invention, the desired price information and the expected price information are each defined as a range consisting of a minimum value and a maximum value, and the price score is mutually consistent when the respective ranges overlap in whole or in part. If the respective ranges do not overlap, the degree of matching may be determined based on the difference between the maximum value of the lower amount range and the minimum value of the higher amount range.

According to another embodiment of the present invention, the price score is calculated in the form of recommended price information by currency unit, and the matching calculation unit based on the desired price information, the expected price information, and the average transaction price 274) may be the work IP transaction recommendation price.

According to another embodiment of the present invention, the price score is the average transaction price derived from IP transaction records of literary works previously concluded by the seller 210 of the work for sale and the buyer 215 who entered the purchase request, respectively. It can be adjusted up or down with reference to. In particular, when the price score is transaction recommendation price information as in the above-described embodiment, the transaction recommendation price may be adjusted by referring to a conventional transaction record. The adjustment may be performed by an information processing algorithm, and preferably, may be performed by machine learning artificial intelligence that has learned by receiving the IP transaction record as an input.

As described above, when the matching score, popularity score, and price score are determined, the matching calculation unit 274 may derive a final matching score by merging the respective scores. Preferably, the matching score is obtained by giving a predetermined additional value to each of the matching score, the popularity score, and the price score, and adding a weighted sum of the matching score, wherein a first percentage weight is applied to the matching score and a second percentage weight is applied to the popularity score. The weight may be obtained by multiplying the price score by the third percentage weight and then summing the results.

According to one embodiment of the present invention, the first percentage ratio can be defined as 40%, the second percentage ratio is 30%, and the third percentage ratio is 30%. According to another embodiment of the present invention, the first, second, and third percentage ratios are applied variably, and based on records of IP transactions of literary works that have been compared between the seller 210 and the buyer 215 in the past can be adjusted The adjustment may be performed by an information processing algorithm, and preferably, may be performed by machine learning artificial intelligence that has learned by receiving the IP transaction record as an input. For example, if a number of transaction records are learned from conventional IP transaction records even though there is some discrepancy in terms of price, the artificial intelligence may operate such as to lower the third percentage weight assigned to the price score. . The method of adjustment in the present invention is not limited to the above-described embodiment, and may be variously modified and implemented by the practitioner of the present invention within the scope corresponding to the spirit of the present invention.

When all of the above procedures are completed, transaction brokerage information 280 is generated. The transaction information 295 is an information set including the matching score calculated by the matching unit 270, and preferably, the matching score, the profile of the seller 210, the profile of the buyer 215, and the work for sale. information 220, purchase request information 225, and recommended price information calculated according to the above embodiment.

According to various embodiments of the present invention, the transaction brokerage information 280 may correspond to one purchase request information 225 for one sales work information 220, and one sales work information 220 ) may correspond to a plurality of purchase request information 225, and may correspond to one purchase request information 225 to a plurality of sales work information 220, and the example shown in FIG. Likewise, a plurality of purchase request information 225 may correspond to a plurality of sales work information 220 . In order to generate the transaction brokerage information including the plurality of corresponding relationships, the entire procedure of converting and comparing the sales copyright information 220 and the purchase request information 250, respectively, is repeatedly performed as many times as necessary, and the result can be synthesized as a single piece of information.

The transaction brokerage information 280 may be transmitted to one or both of the seller 210 and the buyer 215 by the transmission unit 290 . Thereafter, the seller 210 and the buyer 210 may determine whether to trade literary works IP based on information such as matching scores shown in the transaction brokerage information 280, and through this, the automation of literary works to be achieved by the present invention The purpose of IP trading is achieved.

According to an embodiment of the present invention, the system 200 enables the seller 210 and the buyer 215 receiving the transaction brokerage information 280 to determine whether an actual literary work IP transaction has been concluded and related to its price conditions. Information may be requested to be entered into the system 200 as an IP transaction record. As described in some of the above-described embodiments, the input IP transaction records can be used as reference data in some algorithms operating based on conventional transaction records.

5 shows a computer device including a processor for implementing a system according to an embodiment of the present invention. Referring to FIG. 5 , a computer device 500 includes a processor 510, a memory 520, and a transceiver 530. Processor 510 may be configured to implement the functions, processes and/or methods described herein. Each function described as a functional unit in the present invention may be implemented in the processor 510 .

The memory 520 is connected to the processor 510 and stores various information for driving the processor 510 . The communication unit 530 is connected to the processor 510 and transmits a communication signal to the communication network 540 or receives a communication signal from the communication network 540 .

The processor 510 may include an application-specific integrated circuit (ASIC), other chipsets, logic circuits, and/or data processing devices. The memory 520 may include read-only memory (ROM), random access memory (RAM), flash memory, a memory card, a storage medium, and/or other storage devices. The communication unit 530 may include a TCP-IP based network connection device for receiving predetermined information through an external communication signal and transmitting or receiving a predetermined signal through an external communication signal.

When the exemplary system 200 of the present invention as shown in FIG. 2 is implemented as a computer device 500 as shown in FIG. 5, preferably, the communication unit 530 is a seller 210 and a buyer 215 ), to receive the sales work information 220 and the purchase request information 225, and to transmit the transaction brokerage information 290 to one or both of the seller and the buyer, The processor 510 may be configured to perform all processes until the transaction brokerage information 290 is derived by receiving the sales work information 220 and the purchase request information 225 as inputs. In addition, the memory 520 may include a mass storage device, and may be configured to store and operate the databases 260 and 265 by utilizing the memory 520 .

When an embodiment of the present invention is implemented as software, the above-described technique may be implemented as a module (process, function, etc.) that performs the above-described functions. The module may be stored in memory 520 and executed by processor 510 . The memory 520 may be internal or external to the processor 510 and may be connected to the processor 510 by various well-known means.

Although it has been described with reference to the drawings and examples, it does not mean that the scope of protection of the present invention is limited by the drawings or examples, and those skilled in the art can understand the spirit of the present invention described in the following claims. And it will be understood that the present invention can be variously modified and changed without departing from the scope.

Claims (21)

In the transaction mediation method of intellectual property (IP) using a computer device including a processor,
receiving, by the processor, at least one piece of work for sale information from a seller of the work IP;
receiving, by the processor, at least one piece of purchase request information from a buyer of the work IP;
extracting, by the processor, first summary information from the sales work information;
extracting, by the processor, second summary information from the purchase request information;
calculating, by the processor, a story evaluation index based on a correlation between the first summary information and the second summary information;
calculating, by the processor, a matching score between the sales work information and the purchase request information based on the story evaluation index; and
transmitting, by the processor, information about the matching score to one or both of the IP seller and the IP buyer; Including, a transaction brokerage method of literary works IP using a computer device.
According to claim 1,
The first summary information and the second summary information,
at least one of content type, genre, subject, character, event, and background; and
A story information value corresponding to each of the items; a transaction mediation method of literary works IP using a computer device, which is work metadata including.
According to claim 2,
In the step of calculating the story evaluation index,
calculating, by the processor, at least one matching rate by comparing information values assigned to each item for each of at least one item constituting the first summary information and the second summary information;
configuring, by the processor, a story evaluation index corresponding to the matching rate for each of the items; Including, a transaction brokerage method of literary works IP using a computer device.
According to claim 3,
The method in which the processor compares the information values assigned to the respective items with each other,
read a first information value specified in a first item of the first summary information;
reading a second information value specified in the first item from the second summary information;
inputting the first information value and the second information value into a natural language processing device to calculate a matching rate in a word vector space;
determining a match rate in the word vector space as a match rate for the first item; and
A transaction mediation method of literary works IP using a computer device, wherein the processor performs the above procedure for each of the items.
According to claim 1,
The sales work information includes desired price information set by the seller for the sales work;
The purchase request information includes expected price information that the buyer is willing to pay for any sales work,
The matching score is
a matching score obtained by merging the story evaluation indices using at least one calculation method of sum, average, weighted sum, and weighted average;
a popularity score based on at least one of the number of user views, user evaluation scores, user review content, and volume or playing time of the sold work; and
a price score determined based on a comparison between the desired price information and the amount of the expected price information; A transaction mediation method of literary works IP using a computer device, in which the processor weighted sums.
According to claim 5,
The method of the processor weighted summing the matching scores,
Transaction brokerage of literary works IP using a computer device, multiplying the match score by a first percentage weight, the popularity score by a second percentage weight, and the price score by a third percentage weight, respectively, and then summing the results. method.
According to claim 5,
Transmitting, by the processor, information about the matching score,
generating, by the processor, recommended price information of a work for sale based on the desired price information and the expected price information; and
transmitting, by the processor, the recommended price information together with the matching score; Further comprising, a transaction brokerage method of literary works IP using a computer device.
According to claim 1,
Transmitting, by the processor, information about the matching score,
transmitting, by the processor, a plurality of matching scores calculated between one piece of sales work information and a plurality of pieces of purchase request information; and
transmitting, by the processor, a plurality of matching scores calculated between a plurality of pieces of sale information and one piece of purchase request information; Further comprising at least one of, a transaction brokerage method of literary works IP using a computer device.
According to claim 1,
At least one of the first summary information and the second summary information is extracted by a natural language processing device, a transaction brokering method of literary works IP using a computer device.
According to claim 1,
reading, by the processor, at least one of the sales work information and the first summary information from a database; Further comprising, a transaction brokerage method of literary works IP using a computer device.
In a computer device including a processor having a transaction mediation function of literary works IP (Intellectual Property),
A first input unit that receives at least one sales work information from a seller of the work IP;
a second input unit that receives at least one piece of purchase request information from a purchaser of the work IP;
a first extracting unit extracting first summary information from the selling work information;
a second extraction unit extracting second summary information from the purchase request information;
an index calculation unit that calculates a story evaluation index based on the correlation between the first summary information and the second summary information;
a matching calculation unit calculating a matching score between the sales work and the purchase request information based on the story evaluation index; and
a transmitter for transmitting information about the matching score to one or both of the IP seller and the IP buyer; A computer device including a processor that brokers transactions of literary works IP, including a.
According to claim 11,
The first summary information and the second summary information,
at least one of content type, genre, subject, character, event, and background; and
Story information value corresponding to each of the items; A computer device including a processor that mediates transactions of literary works IP, which is work metadata including.
According to claim 12,
The index calculation unit,
A story in which the matching rate for each item is matched based on at least one matching rate calculated by comparing the information values assigned to each item for each of the at least one items constituting the first summary information and the second summary information. A computer device including a processor for brokering a transaction of a literary work IP, configured to output an evaluation index.
According to claim 13,
The method in which the index calculation unit compares the information values assigned to each item with each other,
read a first information value specified in a first item of the first summary information;
reading a second information value specified in the first item from the second summary information;
inputting the first information value and the second information value into a natural language processing device to calculate a matching rate in a word vector space;
determining a match rate in the word vector space as a match rate for the first item; and
A computer device including a processor that mediates transactions of literary works IP, wherein the processor performs the above procedure for each of the items.
According to claim 11,
The sales work information includes desired price information set by the seller for the sales work;
The purchase request information includes expected price information that the buyer is willing to pay for any sales work,
The matching score is
a matching score obtained by merging the story evaluation indices using at least one calculation method of sum, average, weighted sum, and weighted average;
a popularity score based on at least one of the number of user views, user evaluation scores, user review content, and volume or playing time of the sold work; and
a price score determined based on a comparison between the desired price information and the amount of the expected price information; A computer device including a processor for brokering a transaction of a literary work IP, which is a weighted sum of the matching calculation unit.
According to claim 15,
The matching calculator,
The matching score is multiplied by a first percentage weight, the popularity score is multiplied by a second percentage weight, and the price score is multiplied by a third percentage weight, and the results are summed to output the matching score. A computer device containing a processor that brokers transactions.
According to claim 15,
The matching calculator,
Based on the desired price information and the expected price information, generating recommended price information for a work for sale;
The transmission unit,
A computer device including a processor for brokering a transaction of a literary work IP, configured to transmit the recommended price information together with the matching score.
According to claim 11,
the transmission unit,
a plurality of matching scores each calculated between one piece of sales work information and a plurality of pieces of purchase request information; and
a plurality of matching scores each calculated between a plurality of pieces of sale information and one piece of purchase request information; A computer device including a processor for brokering a transaction of a literary work IP, configured to further transmit at least one of the above.
According to claim 11,
At least one of the first extractor and the second extractor is operated by a natural language processing device, and a computer device including a processor for brokering a transaction of a literary work IP.
According to claim 11,
A computer device including a processor for brokering a transaction of a literary work IP, further comprising a database storing at least one of the work for sale information and the first summary information.
A computer readable recording medium containing instructions,
The instructions, when executed in a processor, cause the processor to:
Receive at least one sales work information from the seller of the work IP;
receiving at least one piece of purchase request information from a buyer of the work IP;
extracting first summary information from the selling work information;
extracting second summary information from the purchase request information;
Calculate a story evaluation index based on the correlation between the first summary information and the second summary information;
calculating a matching score between the sales work information and the purchase request information based on the story evaluation index; and
A computer-readable recording medium configured to transmit information about the matching score to one or both of the IP seller and the IP buyer.

Images