KR20230029738A - Method for judging technology similarity by using neural network - Google Patents

Abstract

According to an embodiment of the present disclosure, a computer program stored in a computer-readable storage medium is disclosed. The computer program performs, when run by one or more processors, the following operations for extracting technical fields by using a neural network. The operations may include: an operation of inputting at least a portion of a first partial document or at least one of documents to be compared into a technology similarity determination model trained to determine the similarity between the first partial document and a second partial document or the at least one of the documents to be compared; and an operation of obtaining the technology similarity between the at least a portion of the first partial document and the documents to be compared by using the technology similarity determination model. Therefore, a computing device for determining technology similarity by using a neural network can be provided.

Description

Technology similarity judgment method using neural network {METHOD FOR JUDGING TECHNOLOGY SIMILARITY BY USING NEURAL NETWORK}

The present disclosure relates to data processing using a computing device, and more particularly, to a method for determining technology similarity using a neural network.

As computer and Internet technologies develop, a large amount of data is being accumulated. The importance of technology for efficiently storing large amounts of data and finding data easily and quickly with desired search conditions is increasing. Current storage media allow storing and retrieving a great deal of information.

Many of the large amounts of data may contain technology-related content. As computer and Internet technologies develop, documents containing technical contents, such as papers containing technical contents, internal technical documents of companies, patent documents, etc., are electronically accumulated.

When the storage medium contains a large amount of technical content, there may be difficulties in obtaining the technical content of interest to the user. In addition, it may take a relatively long time for the computing device to search for technical content that the user is interested in.

Even in artificial intelligence technology, especially in the field of deep learning, remarkable progress has been made in data automation processing. In addition, technology for comparing similarity between documents using deep learning is also developing. Therefore, as the technology for judging the similarity between documents develops, there is a trend to use deep learning for technical document retrieval.

Korean Patent Publication No. 2012-0108124 discloses a method for generating a patent dispute prediction model, a system for implementing the method, and a recording medium on which the method is recorded.

The present disclosure has been made in response to the aforementioned background art, and is intended to provide a method for determining technology similarity using a neural network.

According to one embodiment of the present disclosure for realizing the above object, a computer program including instructions stored in a computer readable storage medium and causing a computer to perform the following operations is disclosed. The above operations include inputting at least a part of the first partial document or at least one of the comparison target documents into a technology similarity judgment model trained to determine the degree of similarity between the first partial document and the second partial document or at least one of the comparison target documents. movement; and obtaining a technology similarity between at least a part of the first partial document and the comparison target document by using a technology similarity determination model.

In an alternative embodiment, the document may include at least one of a first part document for claiming a right or a second part document for describing content of a claim.

In an alternative embodiment, the technology similarity determination model inputs the first partial document and the second partial document to the technology similarity determination model to learn a correlation between the contents of the first partial document and the contents of the second partial document. 1 model can be included.

In an alternative embodiment, the technology similarity determination model includes a second model obtained by inputting the first partial document and the comparison target document to the first model and learning a correlation between the contents of the first partial document and the comparison target document. can do.

In an alternative embodiment, the comparison target document may include at least one of a first comparison target document that is a raw comparison target document and a second comparison target document that is a processed document of the original comparison target document.

In an alternative embodiment, the first comparison target document is a document including technology-related content, and may include at least one of an intellectual property-related document, a dissertation, or a technical document.

In an alternative embodiment, the second comparison target document includes at least some of the plurality of sentences included in the first comparison target document, and at least some of the plurality of sentences included in the first comparison target document is a key sentence calculation algorithm. It may include a sentence determined based on.

In an alternative embodiment, the core sentence calculation algorithm may include: calculating a technical term importance based on a technical term appearance frequency for each technical term in a sentence included in the first comparison target document; obtaining an inter-sentence weight matrix by calculating a first matrix having each calculated technical term importance as an element of the matrix and a transposed matrix of the first matrix; calculating a rank of a sentence included in the first comparison target document based on an inter-sentence weight matrix; and determining a sentence having a calculated rank equal to or higher than a predetermined rank as a core sentence.

In an alternative embodiment, the comparison target documents may include updated documents at predetermined intervals based on a specific time point in order to determine the degree of technological similarity with the comparison target documents that are released over time.

In an alternative embodiment, the technology similarity determination model includes a first sub-model for extracting features of the first partial document; a second sub-model for extracting a feature of a second partial document or a feature of the comparison target document; a third sub-model for deriving a correlation between the first partial document and the comparison target document based on the output of the first sub-model and the output of the second sub-model; and a fourth sub-model for acquiring the technology similarity based on correlation.

In an alternative embodiment, the degree of technical similarity may be determined based on a score determined based on the number of sentences included in the compared documents having a correlation with sentences included in the first partial document equal to or greater than a predetermined criterion, or a sentence included in the first partial document. It may include at least one of the scores determined based on the correlation scores of each of the at least one sentence included in the comparison target document.

In an alternative embodiment, performing technology monitoring; and determining a document based on dispute resolution for resolving technology-related disputes.

In an alternative embodiment, the operation of performing technology monitoring may include an operation of determining a comparison target document as a rights infringing document based on a predetermined rights infringement criterion.

In an alternative embodiment, the operation of performing technical monitoring may include sending information related to the determined infringing document to at least one of a terminal of the right holder having a right to the first partial document or an agent terminal of the right holder to the first partial document. It may include the operation of transmitting.

In an alternative embodiment, a technology-related dispute is a technology-related dispute between individuals, between individuals and entities, or between entities and entities, including intellectual property-related judgments, intellectual property-related civil litigation, intellectual property-related criminal litigation, or intellectual property-related administrative disputes. According to another embodiment, including at least one of the lawsuits, a method for determining technology similarity using a neural network is disclosed.

In an alternative embodiment, the dispute resolution basis document is a document helpful in the process of resolving the dispute by sending it to at least one of the parties to the dispute or the dispute resolution agency, such as a warning letter related to intellectual property rights, a document of evidence of invalidity of intellectual property rights, knowledge It may include at least one of a document of evidence of a judgment for revocation of property rights or a document of evidence of litigation related to intellectual property rights.

According to another embodiment of the present disclosure, a method for determining technology similarity using a neural network is disclosed. The method includes inputting at least a part of the first partial document or at least one of the comparison target documents into a technology similarity judgment model trained to determine the degree of similarity between the first partial document and the second partial document or at least one of the comparison target documents. step; and obtaining a technological similarity between at least a part of the first partial document and the comparison target document by using the technology similarity determination model.

A computing device is disclosed according to another embodiment of the present disclosure. one or more processors; and a memory storing instructions executable by the one or more processors; wherein the one or more processors determine the degree of similarity between a first partial document and a second partial document or at least one of a comparison target document by using a technology similarity judgment model trained to determine a degree of similarity between at least a portion of the first partial document or a comparison target; At least one of the documents may be input, and a technology similarity between at least a part of the first partial document and the comparison target document may be obtained using the technology similarity determination model.

According to another embodiment of the present disclosure, a computer readable recording medium storing a data structure is disclosed. A computer-readable recording medium storing a data structure corresponding to parameters of a neural network, at least partially updated in a learning process, wherein the operation of the neural network may be based at least in part on the parameters. The learning process may include: learning a correlation between contents of the first partial document and contents of the second partial document; and learning a correlation between the contents of the first partial document and the contents of the comparison target document.

According to an embodiment of the present disclosure, a computing device for determining technology similarity using a neural network may be provided.

In order that the above-mentioned features of the present disclosure may be understood in detail and with reference to the more detailed description and the following embodiments, some of the embodiments are shown in the accompanying drawings. Also, like reference numbers in the drawings are intended to refer to the same or similar function throughout the various aspects. However, it is to be understood that the accompanying drawings merely illustrate certain exemplary embodiments of the present disclosure and are not considered to limit the scope of the present disclosure, and that other embodiments having the same effect may be fully appreciated. Be careful.
1 is a block diagram of a computing device for data processing according to an embodiment of the present disclosure.
2 is a schematic diagram illustrating a network function for data processing according to an embodiment of the present disclosure.
3 is an exemplary diagram illustrating a technology similarity determination model obtained by learning a correlation between a first partial document and a second partial document.
4 is an exemplary view illustrating a technology similarity determination model obtained by learning a correlation between a first partial document and a comparison target document.
5 is an exemplary diagram illustrating a technology similarity determination model by way of example.
6 is an exemplary diagram for explaining a process of technology monitoring and determining a document based on dispute resolution by way of example.
7 is a flowchart of a method for determining technology similarity according to an embodiment of the present disclosure.
8 is a block diagram illustrating a module for determining technology similarity according to an embodiment of the present disclosure.
9 is a simplified and general schematic diagram of an exemplary computing environment in which embodiments of the present disclosure may be implemented.

Various embodiments are now described with reference to the drawings. In this specification, various descriptions are presented to provide an understanding of the present disclosure. However, it is apparent that these embodiments may be practiced without these specific details.

The terms “component,” “module,” “system,” and the like, as used herein, refer to a computer-related entity, hardware, firmware, software, a combination of software and hardware, or an execution of software. For example, a component may be, but is not limited to, a procedure, processor, object, thread of execution, program, and/or computer running on a processor. For example, both an application running on a computing device and a computing device may be components. One or more components may reside within a processor and/or thread of execution. A component can be localized within a single computer. A component may be distributed between two or more computers. Also, these components can execute from various computer readable media having various data structures stored thereon. Components may be connected, for example, via signals with one or more packets of data (e.g., data and/or signals from one component interacting with another component in a local system, distributed system) to other systems and over a network such as the Internet. data being transmitted) may communicate via local and/or remote processes.

In addition, the term “or” is intended to mean an inclusive “or” rather than an exclusive “or”. That is, unless otherwise specified or clear from the context, “X employs A or B” is intended to mean one of the natural inclusive substitutions. That is, X uses A; X uses B; Or, if X uses both A and B, "X uses either A or B" may apply to either of these cases. Also, the term “and/or” as used herein should be understood to refer to and include all possible combinations of one or more of the listed related items.

Also, the terms "comprises" and/or "comprising" should be understood to mean that the features and/or components are present. However, it should be understood that the terms "comprises" and/or "comprising" do not exclude the presence or addition of one or more other features, elements, and/or groups thereof. Also, unless otherwise specified or where the context clearly indicates that a singular form is indicated, the singular in this specification and claims should generally be construed to mean "one or more".

In addition, the term “at least one of A or B” should be interpreted as meaning “when only A is included”, “when only B is included”, and “when A and B are combined”.

Those skilled in the art will further understand that the various illustrative logical blocks, components, modules, circuits, means, logics, and algorithm steps described in connection with the embodiments disclosed herein may be implemented using electronic hardware, computer software, or combinations of both. It should be recognized that it can be implemented as To clearly illustrate the interchangeability of hardware and software, various illustrative components, blocks, configurations, means, logics, modules, circuits, and steps have been described above generally in terms of their functionality. Whether such functionality is implemented in hardware or as software depends on the particular application and design constraints imposed on the overall system. Skilled artisans may implement the described functionality in varying ways for each particular application. However, such implementation decisions should not be interpreted as causing a departure from the scope of this disclosure.

The description of the presented embodiments is provided to enable any person skilled in the art to use or practice the present invention. Various modifications to these embodiments will be apparent to those skilled in the art of this disclosure. The general principles defined herein may be applied to other embodiments without departing from the scope of this disclosure. Thus, the present invention is not limited to the embodiments presented herein. The present invention is to be accorded the widest scope consistent with the principles and novel features set forth herein.

1 is a block diagram of a computing device for data processing according to an embodiment of the present disclosure.

1 shows a configuration of a computing device 100 according to an embodiment of the present invention. The computing device 100 shown in FIG. 1 is only a simplified example of a computing device 100 according to one embodiment of the present invention. In one embodiment of the present disclosure, the computing device 100 may include other components for performing a computing environment of the computing device 100, and only some of the disclosed components may constitute the computing device 100.

The computing device 100 may include a processor 110 , a memory 130 , and a network unit 150 .

The processor 110 may be composed of one or more cores, and may include a central processing unit (CPU), a general purpose graphics processing unit (GPGPU), and a tensor processing unit (TPU) of a computing device. unit), data analysis, and processors for deep learning. The processor 110 may read a computer program stored in the memory 130 and process data for machine learning according to an embodiment of the present disclosure. According to an embodiment of the present disclosure, the processor 110 may perform an operation for learning a neural network. The processor 110 performs neural network learning, such as processing input data for learning in deep learning (DL), extracting features from input data, calculating errors, and updating neural network weights using backpropagation. calculations can be performed for At least one of the CPU, GPGPU, and TPU of the processor 110 may process learning of the network function. For example, the CPU and GPGPU can process learning of network functions and data classification using network functions. In addition, in an embodiment of the present disclosure, the learning of a network function and data classification using a network function may be processed by using processors of a plurality of computing devices together. In addition, a computer program executed in a computing device according to an embodiment of the present disclosure may be a CPU, GPGPU or TPU executable program.

According to an embodiment of the present disclosure, the processor 110 may determine technology similarity using a neural network. In order to determine technology similarity using a neural network, the processor 110 may perform the following operations. The processor 110 assigns at least a part of the first partial document or at least one of the comparison target documents to a technical similarity judgment model learned to determine the degree of similarity between the first partial document and the second partial document or at least one of the comparison target documents. can be entered.

According to an embodiment of the present disclosure, the document may include at least one of a first part document for claiming a right or a second part document for explaining content of claiming a right.

According to one embodiment of the present disclosure, a document may include a physically incorporated set of information designed to communicate. A document may include a textual representation of a person's thoughts by providing information using symbolic symbols. Documents may include digital files in a specific format according to the development of computers. Digital files of a specific type may include, for example, music files, image files, text files, and the like. The foregoing document is only an example, and the present disclosure is not limited thereto.

A document according to an embodiment of the present invention may include two or more arbitrary partial documents. Any union of two or more may collectively include all documents according to an embodiment of the present invention. In another embodiment, the union of any two or more partial documents may contain only a portion of a document according to an embodiment of the present invention. Two or more arbitrary partial documents may share common parts. The foregoing partial document is only an example, and the present disclosure is not limited thereto.

In one embodiment of the present invention, two or more arbitrary partial documents included in a document may be defined as a first partial document and a second partial document. The first part document and the second part document may have a correlation based on the content of the document. For example, at least a portion of the first partial document may be the same as or substantially correspond to any portion of the second partial document. In one embodiment of the present invention, the first partial document may be prepared based on all or part of the document or the second partial document. For example, the first part document may be the claims of a patent specification, an abstract of a paper, a table of contents of a book, a book review of a book, a synopsis of a book, and the like.

In one embodiment of the present invention, a document may include a first partial document including core information of the document and a second partial document having a substantial correspondence with the first partial document including core information of the document. The core information of the document may include information that compressively expresses the content to be conveyed by the document. Key information in a document may include information that allows a person or computing device to know what the document is intended to convey without having to read the document in its entirety. The key information of a document is, for example, a part document containing the contents of an abstract or introduction in a paper, a part document including a decision, a summary of a judgment in a judgment, a part document including the content of claims in a patent specification, and a design registration. This may include drawings in applications, trademarks and designated products in trademark registration applications, tables of contents or plot summaries in books (e.g., magazines, poems, novels, workbooks, textbooks).

In one embodiment of the present disclosure, the second partial document may substantially correspond to the first partial document. The second partial document may have a correlation with the first partial document. The second partial document may have a certain level or higher correlation with the first partial document. The second part document may include contents supporting or specifically explaining the first part document. The second part document, for example, the experimental method and the result of the experiment in the paper, the reason for the judgment in the judgment, the specific description for practicing the invention in the patent specification, the part document explaining the design in the design registration application, and the trademark in the trademark registration application. It may include part of a document, contents corresponding to the table of contents in a book, or all or part of a book. The aforementioned second part document is only an example, and the present disclosure is not limited thereto.

According to an embodiment of the present disclosure, the first part document for claiming rights may include at least a part of a document describing a part to be legally protected according to each country's laws and regulations. Rights may include patent rights, utility model rights, design rights, trademark rights, and copyrights. For example, if the right is a patent right, the first part document for claiming the right may include the matter set forth in the claim. The aforementioned first part document is only an example, and the present disclosure is not limited thereto.

According to one embodiment of the present disclosure, the second part of the document for explaining the contents of the claim may include at least a part of the document for specifically explaining the part to be legally protected according to the laws and regulations of each country. For example, if the right is a patent right, the second part of the document to explain the content of the claim is the name of the invention, technical field, background technology of the invention, content of the invention, a brief description of the drawings, and a description of the invention. It may include detailed descriptions, code descriptions, drawings, summaries, and representative diagrams. The title of the invention may include a name for distinguishing categories of the invention in consideration of the contents of the invention. The technical field may include a clear and concise description of the technical field of the invention for which a patent is sought. The background technology of the invention may include the description of the prior art that is considered useful for understanding and researching the invention. The content of the invention may include a problem to be solved, a solution to the problem, and an effect of the invention. A brief description of the drawings may include a description of the city state, type, and illustrated portion of the accompanying drawings so that it is easy to understand what the drawings are specifically intended to represent. Specific details for carrying out the invention may include at least one specific description for carrying out the invention so that a person skilled in the art can easily know how the invention is practiced. The description of the symbols may include descriptions of the symbols representing the main parts of the drawings. The drawings may include images for explaining specific details for carrying out the invention. The summary may include content briefly explaining the invention within a certain number of characters (eg, within 300 characters). Representative drawings may include reference numerals representing the contents of the invention in at least one or more drawings. The aforementioned second part document is only an example, and the present disclosure is not limited thereto.

Hereinafter, a technology similarity judgment model will be described in detail with reference to FIG. 3 .

According to an embodiment of the present disclosure, the technology similarity determination model 250 may include a model learned to determine a degree of similarity between the first partial document 210 and the second partial document 230 . The model learned to determine the degree of similarity between the first partial document 210 and the second partial document 230 is a machine learning model, for example, a convolutional neural network (CNN), a recurrent neural network (RNN), or an LSTM ( Long Short Term Memory), GRU (Gated Recurrent Unit), FCN (Fully Connected Layer), Activation Layer, and the like may be included. The technology similarity determination model described above is only an example, and the present disclosure is not limited thereto.

According to an embodiment of the present disclosure, the technology similarity determination model 250 inputs the first partial document 210 and the second partial document 230 to the technology similarity determination model 250 to determine the first partial document ( 210) and the first model for learning the correlation between the contents of the second partial document 230. The processor 110 may perform preprocessing to convert the first partial document 210 and the second partial document 230 into a form to be input to the technology similarity determination model 250 . For example, the processor 110 may perform a tokenization process of dividing sentences included in the first partial document 210 and the second partial document 230 into tokens according to a predetermined criterion. there is. As another example, the processor 110 converts sentences included in the first partial document and the second partial document into a table form (eg, each word in the sentence is a table row, and each word vector is represented as a table column). table) can be performed. The foregoing is only an example, and the present disclosure is not limited thereto.

According to an embodiment of the present disclosure, the first model may include a model obtained by learning a correlation between the contents of the first partial document 210 and the contents of the second partial document 230 . Accordingly, the processor 110 may obtain a degree of similarity by inputting the first partial document 210 and the second partial document to the first model. For example, the first model may include a model learned using a patent specification as a learning data set. The training data set includes a first sub-learning data set comprising a first part document (eg claims) and a second sub-learning data set comprising a second part document (eg specific description of an invention). can do. Accordingly, the first model may include a model learned using the first sub-training data set and the second sub-training data set. According to an embodiment of the present disclosure, the technology similarity determination model may include a first model. Accordingly, the processor 110 may obtain the degree of similarity 270 based on the correlation between the contents of the first part document and the second part document by using the technology similarity determination model. For example, when the correlation between the contents of the first partial document and the second partial document is high, the degree of similarity 270 obtained by the processor 110 may be high. The aforementioned first model is only an example, and the present disclosure is not limited thereto.

According to another embodiment of the present disclosure, the first model is for summarizing documents input to the first model based on a correlation between the contents of the first partial document 210 and the contents of the second partial document 230. models can be included. And, the third model may include a model that performs description classification on the summary document, which is the output data of the first model. The third model uses classification target data (eg, patent specification) as learning input data, and learning data that labels a classification (eg, artificial intelligence, semiconductor, steel, communication, etc.) corresponding to the classification target data. It can include models trained using sets. Accordingly, the technology classification model may include the first model and/or the third model. For example, the technology classification model may include a model in which the first model and the third model are connected. The processor 110 may obtain a summary of the document by inputting the document into the first model, and then obtain a classification result of the document by inputting the obtained summary of the document into the third model. According to an embodiment of the present disclosure, the processor 110 may obtain a technology classification result of a document input to the technology classification model by using the technology classification model. For example, the processor 110 may acquire a technology classification of 'artificial intelligence' by inputting a patent specification including artificial intelligence technology into a technology classification model. As described above, the technology classification model may include a first model and/or a third model. The processor 110 may perform end-to-end learning on the technology classification model. That is, the first model may include a first model obtained by end-to-end training of a technology classification model. End-to-end learning may refer to learning that optimizes weights of the neural network in consideration of input data and output data in a process of receiving input data at one end of the neural network and generating output data at the other end. The processor 110 may perform end-to-end training of the technical classification model as the accuracy of the technical classification of the document increases, thereby increasing the probability of obtaining a first model capable of summarizing the document including only the core content of the document. Accordingly, a first model having excellent ability to summarize documents may be obtained through a learning process that increases the accuracy of the skill classification of the skill classification model. In the present disclosure, the meaning of having excellent summary ability of a document means that a second part document (eg, a detailed description of an invention supporting the content of the claim) having a high correlation with a first part document (eg, claim) It may mean that it can be extracted with high accuracy. Accordingly, the processor 110 may obtain a second partial document having a high correlation with the first partial document by using the first model acquired through end-to-end learning. In other words, the processor 110 may use the first model to find content highly correlated with a specific part of a document from other documents with high accuracy. More specifically, the processor 110 may use the first model to find the same or similar technical content as the technical idea included in the document in another document with high accuracy. The foregoing is only an example, and the present disclosure is not limited thereto.

According to an embodiment of the present disclosure, the processor 110 may obtain a similarity 270 by inputting the first partial document 210 and the second partial document 230 into the technology similarity determination model 250 . The similarity 270 may include a similarity between the first partial document and the second partial document or at least one of the comparison target documents. For example, if the first part document is the claims of a patent application and the second part document is a detailed description of an invention contained in that patent application, between the first part document and at least one of the second part document or the comparative document The degree of similarity may be high. Since the detailed description of the invention included in the patent application contains content to support the content included in the claims, the degree of similarity between the first part document and the second part document may be high. According to an embodiment of the present disclosure, the degree of similarity 270 may be expressed in a form of probability. For example, when the technical similarity between the first partial document 210 and the second partial document is high, the similarity 270 may be 0.96. The foregoing is only an example and the present disclosure is not limited thereto.

Hereinafter, another example of a technology similarity determination model will be described in detail with reference to FIG. 4 .

According to an embodiment of the present disclosure, the technology similarity determination model 250 inputs the first partial document 210 and the comparison target document 310 to the first model and compares them with the contents of the first partial document 210. A second model obtained by learning the correlation between contents of the target document 310 may be included.

According to an embodiment of the present disclosure, the second model may include a model obtained by learning a correlation between the contents of the first partial document 210 and the contents of the comparison target document 310 . The second model may include a model obtained by re-training the first model obtained by learning the correlation between the contents of the first partial document and the contents of the second partial document. The processor 110 learns the first model by inputting the first partial document 210 and the comparison target document 310 into the first model, thereby establishing a relationship between the contents of the first partial document 210 and the contents of the second partial document. In addition to the correlation, a learned model may be obtained for both the correlation between the contents of the first partial document 210 and the contents of the comparison target document 310 . The processor 110 is configured to learn a correlation between the contents of the first part document 210 and the comparison target document 310 by a learning method of fixing at least some of the parameters included in the first model and updating the remaining parameters. can The aforementioned second model is only an example, and the present disclosure is not limited thereto.

According to another embodiment of the present disclosure, the technology similarity determination model 250 may include a second model. Accordingly, the processor 110 may obtain the similarity 270 based on the correlation between the contents of the first partial document 210 and the comparison target document 310 using the technology similarity determination model 250 . For example, when the correlation between the contents of the first partial document 210 and the comparison target document 310 is high, the degree of similarity 270 may be high. That is, the processor 110 uses the technology similarity determination model 250 to determine a higher similarity 270 as the technical ideas included in the first partial document 210 and the technical ideas included in the comparison target document 310 are similar. can be obtained The foregoing is only an example and the present disclosure is not limited thereto.

According to an embodiment of the present disclosure, the comparison target document 310 may include a first partial document and a comparison target document. The comparison target document 310 may include a document to be judged in order to check whether the technical idea included in the first partial document has the same or similar technical idea.

According to an embodiment of the present disclosure, the comparison target document 310 may include at least one of a first comparison target document, which is a raw comparison target document, and a second comparison target document, which is a document obtained by processing the original comparison target document. .

According to an embodiment of the present disclosure, the first comparison target document may include an original comparison target document. Raw documents to be compared may include all documents converted to data form. Accordingly, the processor 110 may obtain a raw comparison target document from the Internet through web crawling. Web crawling may include an operation in which the processor 110 visits a web page on the Internet and collects various types of information. The original comparison target document may include a document obtained by the processor 110 through web crawling and not processed (eg, document summary, key word extraction). The foregoing first comparison target document is merely an example, and the present disclosure is not limited thereto.

According to an embodiment of the present disclosure, the first comparison target document may include a document including technology-related content. The first comparison target document may include at least one of an intellectual property-related document, a thesis, or a technical document. According to an embodiment of the present disclosure, documents related to intellectual property rights may include documents necessary to legally protect intellectual property rights. Documents necessary for legal protection of intellectual property rights are documents exchanged between private individuals and administrative agencies in order to register intellectual property rights with administrative agencies. . According to an embodiment of the present disclosure, a thesis may include a document systematically prepared by an author in accordance with a consistent and consistent format by interpreting his/her academic research results, opinions, and arguments on a certain subject in accordance with logic. The paper may include at least one of the paper's title, abstract, experimental method, results, discussion, current affairs, or references. According to an embodiment of the present disclosure, a technical document may include any document describing the structure, use, function, creation, etc., of a technology. For example, technical documentation may include a technical brief prepared by a company, an invention declaration prepared by an inventor, and the like. Other examples of technical documents include technical reports, technology development business plans, technology evaluation reports, technology development business application plans, and job descriptions. It can include technology status analysis tables, new technology proposals, technology-related documents written on Internet blogs, and technology-related documents written on Internet homepages. The foregoing first comparison target document is merely an example, and the present disclosure is not limited thereto.

According to an embodiment of the present disclosure, the second comparison target document may include at least a part of the first comparison target document. The second comparison target document may include at least some of the plurality of sentences included in the first comparison target document. A sentence may consist of at least one word. Accordingly, the second comparison target document may include at least some of the plurality of words included in the first comparison target document. The foregoing is only an example and the present disclosure is not limited thereto.

According to an embodiment of the present disclosure, the second comparison target document may include sentences determined based on a key sentence calculation algorithm, at least some of a plurality of sentences included in the first comparison target document. The processor 110 may obtain a second comparison target document from the first comparison target document by using a core sentence calculation algorithm. The processor 110 may obtain a degree of similarity with the first partial document with high accuracy by obtaining the second comparison target document. That is, the processor 110 may acquire a similarity with high accuracy by extracting only the part including the core content from the comparison target document and acquiring the similarity with the first part document. The reason why a similarity with high accuracy can be obtained is that the traditional similarity judgment method extracts only the core contents of each document and determines the similarity by comparing the core contents of each document, and it has been proven that the method has high accuracy as a rule of thumb. . In addition, the processor 110 inputs the second comparison target document and the first partial document to the technology similarity acquisition model by using the second comparison target document including only the core content of the first comparison target document, and calculates the degree of similarity. It could be lighter and faster. The foregoing is only an example and the present disclosure is not limited thereto.

According to an embodiment of the present disclosure, the key sentence calculation algorithm may include a method of calculating key sentences and/or key words as a method of summarizing a document. The core sentence calculation algorithm may include a TextRank algorithm. The TextRank algorithm may calculate each key sentence and/or key word by using PageRank, a graph ranking algorithm, after constructing a word graph or sentence graph. The foregoing key sentence calculation algorithm is merely an example, and the present disclosure is not limited thereto.

According to an embodiment of the present disclosure, a key sentence calculation algorithm may include calculating a technical term importance based on a technical term appearance frequency for each technical term in a sentence included in a first comparison target document; obtaining an inter-sentence weight matrix by calculating a first matrix having each calculated technical term importance as an element of the matrix and a transposed matrix of the first matrix; calculating a rank of a sentence included in a first comparison target document based on an inter-sentence weight matrix; and determining a sentence having a calculated rank equal to or higher than a predetermined rank as a core sentence.

According to an embodiment of the present disclosure, the processor 110 may calculate a technical term importance based on a technical term appearance frequency for each technical term in a sentence included in the first comparison target document. The processor 110 may calculate the technical term importance using a term frequency-inverse document frequency (TF-IDF) algorithm. TF (Term Frequency) may include a word frequency and a scale indicating the frequency in which a specific word appears in a document. Inverse document frequency (IDF) may include dividing the total number of documents by the number of documents including a specific word. The foregoing is only an example and the present disclosure is not limited thereto.

According to an embodiment of the present disclosure, the processor 110 may obtain an inter-sentence weight matrix by calculating a first matrix having each calculated technical term importance as a matrix element and a transposed matrix of the first matrix. Rows of the first matrix may correspond to sentences, and columns of the first matrix may correspond to words. The processor 110 may obtain an inter-sentence weight matrix by calculating (multiplying) the first matrix and the transposed matrix of the first matrix. In this case, each row and column of the inter-sentence weight matrix may correspond to a sentence. Accordingly, the processor 110 may generate an inter-sentence weight graph based on the inter-sentence weight matrix. In the inter-sentence weight graph, vertices are sentences and edges may include information about correlations between sentences. The foregoing is only an example, and the present disclosure is not limited thereto.

According to an embodiment of the present disclosure, the processor 110 may calculate a rank of a sentence included in the first comparison target document based on an inter-sentence weight matrix. The rank may be determined based on the weight of citing and referencing different web pages by a plurality of web pages in the PageRank algorithm. For example, the ranking of a particular website may be higher if other websites refer to the particular website. A high rank may mean that there is a high probability that information desired by a user exists on a specific website. In the present disclosure, the rank of a sentence may mean the importance of a sentence based on a relationship between a plurality of sentences. Accordingly, the higher the rank of the specific sentence, the higher the importance of the specific sentence. The foregoing is only an example and the present disclosure is not limited thereto.

According to an embodiment of the present disclosure, the processor 110 may determine a sentence having a calculated rank equal to or higher than a predetermined rank as a core sentence. The predetermined rank may include criteria for calculating key sentences in a document with high accuracy. A predetermined rank may be set by an operator. Through this, the processor 110 can extract key sentences with high accuracy by performing filtering by applying criteria including qualitative and/or quantitative analysis of the operator. The foregoing is only an example and the present disclosure is not limited thereto.

A description of the specific content related to the key sentence calculation algorithm is specifically discussed in the paper TextRank: Bringing Order into Texts (published date: 2004, authors: Rada Mihalcea, Paul Tarau), which is incorporated by reference in its entirety in this application.

According to an embodiment of the present disclosure, documents to be compared may include updated documents at predetermined intervals based on a specific point in time in order to determine a similarity of technology to documents to be compared that are disclosed over time. The predetermined period may include a period determined by an operator and/or a period determined based on a data analysis result of a computing device. The predetermined period may be, for example, one week, one month, or one year. The predetermined cycle may include, for example, an average cycle at which papers are presented at conferences. Through this, a person who has an intellectual property right for a specific technology can continuously monitor whether a technology similar to her own technology is disclosed by updating the latest data. Documents included in the comparison target documents are automatically updated, so that a person who has intellectual property rights for a specific technology can find others who imitate their technology or infringe their intellectual property rights. The foregoing is only an example and the present disclosure is not limited thereto.

Hereinafter, a specific configuration of the technology similarity determination model will be described in detail with reference to FIG. 5 . Some of the specific configurations of the technology similarity determination model may be omitted or additional configurations may be included.

According to an embodiment of the present disclosure, the first submodel 410 may include a recurrent neural network. A recurrent neural network may include a long short-term memory network. The processor 110 may extract features of the first partial document 210 by inputting the first partial document to the long and short term storage network. A feature of the first part document may include a first submodel encoding matrix. The first submodel encoding matrix may include a vector obtained by inputting a document into a long-term and short-term storage network. The obtained vector may include a vector having a probability value of each word as a vector element. The processor 110 may obtain the first submodel encoding matrix using the vector obtained from the first submodel 410 . For example, if a set of words included in the first sub-document is (x1, x2, x3, x4, ... , xn) and a vector corresponding to each word (the processor assigns words to the first submodel 410) When the vector obtained by input) is (v1, v2, v3, v4, ... ,vn), the first submodel encoding matrix will be expressed as [v1, v2, v3, v4, ... ,vn] can The aforementioned first submodel is only an example, and the present disclosure is not limited thereto.

According to an embodiment of the present disclosure, the second submodel 430 may include a recurrent neural network. A recurrent neural network may include a long short-term memory network. The long and short term storage network of the second submodel 430 may include the same network as the short and long term storage network of the first submodel 410 . That is, the parameters of the long and short term storage network of the second submodel 430 may be the same as the parameters of the short and long term storage network of the first submodel 410 . Since the first sub-model 410 and the second sub-model 430 share the parameters of the network, the processor 110 can convert the same word sequence (eg, sentence) to the first sub-model 410 and the second sub-model 430. Inputs to the model 430 can obtain the same or similar vector representations. The aforementioned second submodel is only an example, and the present disclosure is not limited thereto.

According to an embodiment of the present disclosure, the processor 110 may extract features of the second partial document 230 by inputting the second partial document 230 to a short-term storage network. A feature of the second part document 230 may include a second submodel encoding matrix. The second submodel encoding matrix may include a vector obtained by inputting a document into a long-term and short-term storage network. The obtained vector may include a vector having a probability value of each word as a component of the vector. The processor 110 may obtain a second sub-model encoding matrix using a vector obtained from the second sub-model. For example, if the set of words included in the second sub-document is (y1, y2, y3, y4, ... , yn), and a vector corresponding to each word (obtained by the processor inputting words to the second submodel) If one vector) is (v'1, v'2, v'3, v'4, ... , v'n), the second submodel encoding matrix is [v'1, v'2, v' 3, v'4, ... , v'n]. The aforementioned first sub-model is only an example, and the present disclosure is not limited thereto.

According to an embodiment of the present disclosure, the third sub-model 450 may include a model capable of deriving a correlation between a first partial document and a second partial document. The processor 110 derives a correlation between the first partial document and the second partial document based on the outputs of the first sub-model 410 and the second sub-model 430 using the third sub-model 450. can do. The processor 110 may derive a correlation between the first partial document and the second partial document based on the first submodel encoding matrix and the second submodel encoding matrix. The processor 110 inputs a result obtained by calculating the transpose of the second sub-model encoding matrix and the first sub-model encoding matrix to the softmax function, so that words included in the first partial document and the second partial document It is possible to obtain correlations of words included in . The aforementioned third sub-model is only an example, and the present disclosure is not limited thereto.

Description of the specific configuration of the technology similarity judgment model is discussed in detail in the patent "Data processing method using neural network" (application number: KR10-2019-0103499), which is incorporated as a reference in its entirety in this application.

According to an embodiment of the present disclosure, the fourth sub-model 470 may include a model for obtaining a similarity between the first partial document 210 and the second partial document 230 . The fourth sub-model may include a machine learning model and/or a rule-based model for obtaining a degree of similarity. According to an embodiment of the present disclosure, the processor 110 uses the fourth sub-model 470 to process sentences and/or sentences included in the second partial document 230 having a high correlation with the first partial document 210 . words can be obtained. In another embodiment of the present disclosure, the processor 110 may obtain a similarity based on a correlation score for each sentence and/or word having a correlation with the first partial document 210 equal to or greater than a predetermined criterion. The degree of similarity is described in detail below. The aforementioned fourth sub-model is only an example, and the present disclosure is not limited thereto.

According to an embodiment of the present disclosure, the degree of similarity 270 may include a degree of similarity between a specific document and another document. The similarity 270 may include a similarity between technical ideas included in a specific document and technical ideas included in other documents. Therefore, the degree of similarity may be high if the technical idea is similar and the technical idea included in a specific document is imitated or infringed. The similarity 270 may include a similarity between the first partial document and the second partial document or at least one of the comparison target documents. Also, the similarity 270 may include a similarity between the first partial document and the comparison target document. The above-described degree of similarity is only an example, and the present disclosure is not limited thereto.

According to another embodiment of the present disclosure, the degree of similarity 270 is a word included in the comparison target document 310 having a correlation with at least one of words or sentences included in the first partial document 210 equal to or greater than a predetermined criterion. Alternatively, it may include a score determined based on the number of at least one of the sentences. The predetermined criterion may be a criterion determined based on whether the contents of sentences included in the first partial document 210 and sentences included in the comparison target document are similar to each other. Accordingly, the processor 110 may determine the degree of similarity based on the number of at least one of words or sentences included in the comparison target document 310 having a correlation equal to or greater than a predetermined criterion. For example, if the number of sentences having a correlation higher than a predetermined criterion (for example, to the extent that technical ideas are judged to be identical or similar to each other) is 10, the similarity between the first part document and the comparison target document is 10 days. may be The aforementioned degree of similarity is only an example, and the present disclosure is not limited thereto.

According to another embodiment of the present disclosure, the degree of similarity 270 is the degree of at least one of words or sentences included in the first partial document 210 and at least one word or sentence included in the comparison target document 310. It may include a score determined based on each correlation score. The correlation score may be a score obtained by the processor 110 using the third sub-model 450 . The correlation score may include a score determined based on a correlation between a specific sentence and/or word and another sentence and/or word. Therefore, the higher the correlation score, the higher the semantic similarity between a specific sentence and/or word and another sentence and/or word. For example, the correlation score may be a probability obtained by the processor 110 using the third sub-model 450 . The probability may indicate a probability that a word or sentence included in the first partial document is semantically similar to a word or sentence included in the comparison target document. For example, when a word included in the first partial document is an RNN and a word included in the comparison target document is an RNN, the probability that the meaning is similar may be 100%. As another example, if a word included in the first partial document is an RNN and a word included in the comparison target document is a CNN, the probability that the meaning is similar may be 80%. As another example, if the word included in the first partial document is RNN and the word included in the comparison target document is melting pot, the probability that the meaning is similar may be 1%. Accordingly, the processor 110 may obtain correlation scores between core sentences and/or words in the first partial document and sentences and/or words included in the comparison target document by using the third submodel. The processor 110 may calculate the similarity 270 by summing up correlation scores for sentences and/or words included in comparison target documents. According to another embodiment of the present disclosure, the processor 110 may add weights to sentences and/or words included in comparison target documents to add up correlation scores. For example, if the document to be compared is a dissertation, high weights are given to sentences and/or words included in the introduction and abstract of the paper, and low weights are given to sentences and/or words included in previous studies. Correlation scores can be summed. The foregoing is only an example and the present disclosure is not limited thereto.

According to another embodiment of the present disclosure, the degree of similarity 270 is a word or sentence included in the comparison target document 310 having a correlation with at least one of words or sentences included in the first partial document 210 equal to or greater than a predetermined criterion. It may include a score determined based on a correlation score of at least one of the sentences. That is, instead of calculating and summing the correlation scores for all sentences and/or words included in the comparison target documents, correlation scores are calculated and summing up for some sentences and/or words included in the comparison target documents. could be a way to do it. Documents to be compared may include a number of parts that are not core technical ideas, and when correlation scores are derived and summed for the relevant parts, the similarity of documents to be compared that are actually included in similar technical ideas may be calculated low. Accordingly, the processor 110 may increase the accuracy of determining similarity by summing only correlation scores equal to or greater than a predetermined standard. The foregoing is only an example and the present disclosure is not limited thereto.

Hereinafter, a method of utilizing the calculated degree of similarity will be described in detail with reference to FIG. 6 .

According to one embodiment of the present disclosure, processor 110 may perform technology monitoring 510 . The technology monitoring 510 may include an operation of monitoring whether or not a subsequent technology similar to the technology possessed by the user is disclosed. Users may not have enough time and money to determine who imitates and uses their technology without permission through Internet searches every day. Therefore, the processor 110 automatically judges the published documents to be compared with the first part document including the user's technology using the technology similarity judgment model 250 and transmits the similarity judgment result to the user, so that the user can conveniently You can easily spot others who infringe on your rights. The foregoing technical monitoring is only an example, and the present disclosure is not limited thereto.

According to an embodiment of the present disclosure, the processor 110 processes documents published after the legal protection start date of the first part document (eg, the patent application date if a patent is registered) of the first part document. Technical monitoring can be performed for Documents opened after the legal protection start date of the first part document may include a comparison target document D 314 , a comparison target document E 315 , and a comparison target document F 316 . The processor 110 may calculate a similarity between the first partial document 210 and the comparison target document D 314 , the comparison target document E 315 , or the comparison target document F 316 . For example, the processor 110 determines that the similarity between the first partial document 210 and the comparison target document D 314 is 0.97 (640) and the similarity between the first partial document 210 and the comparison target document E 315 is 0.97 (640). 0.13 (650), and the similarity between the first partial document 210 and the comparison target document F 316 is 0.92 (660). In this case, since the comparison target document D 314 and the comparison target document F 316 have a high degree of similarity, they may include the same or similar technical ideas as the first partial document 210 . The foregoing is only an example and the present disclosure is not limited thereto.

According to an embodiment of the present disclosure, the operation of technology monitoring may include an operation of determining a comparison target document as a rights infringing document based on a predetermined rights infringement criterion. The predetermined criterion for determining infringement of rights may include a criterion determined based on whether the comparison target document includes the same or similar technical idea as the first part document. For example, the predetermined criterion for determining infringement of rights may include a criterion for determining a document as infringing upon a right if the degree of similarity is 0.7 or higher. The predetermined rights infringement judgment criterion may include an operator or a criterion determined based on data analysis for determining similarity. Accordingly, the predetermined rights infringement judgment criterion may be a criterion for determining a comparison target document having a similarity of 0.7 or higher as a rights infringement document. In this case, the processor 110 may determine the comparison target document D 314 and the comparison target document F 316 as documents infringing on rights. The foregoing is only an example and the present disclosure is not limited thereto.

According to an embodiment of the present disclosure, the operation of performing technical monitoring may include transmitting information related to the determined infringing document to at least one of a terminal of a right holder having a right to the document of the first part or an agent terminal of the holder of the right to the document of the first part. It may include an operation to transmit to one. The processor 110 may transmit the document to be compared, which is determined to be a document infringing on rights, to at least one of a terminal of a right holder having a right to the first partial document or an agent terminal of the right holder to the first partial document. Through this, the holder of the right or the agent for the document of the first part can continuously receive information on the document infringing on rights through the user terminal. Therefore, the right holder or agent for the first part document can selectively determine whether the technology is infringed based on the information received without having to continuously search for infringing technology through Internet searches, saving both time and money. there is. The foregoing is only an example and the present disclosure is not limited thereto.

According to an embodiment of the present disclosure, the processor 110 may determine a dispute resolution basis document for resolving a technology-related dispute. An owner with rights to a Part 1 document may send a notice of infringement or initiate an infringement action against a person practicing a particular technology. In this case, the person practicing the particular technology may invalidate the rights to the Part 1 document or make a defense that the technology contained in the Part 1 document is publicly available technology in order to conduct and defend its business. . In this case, a person practicing a particular technology may have a need to find a document that was published prior to the first part document 210 legal protection commencement date. If the technical idea contained in the document of the first part exists in a document disclosed before the legal protection start date of the first part document 210, the right to the first part document may be invalidated, and furthermore, as a publicly available technology, the right holder You may not be able to claim your rights to others. Through this, disputes between right holders and implementers may be resolved. The foregoing is only an example and the present disclosure is not limited thereto.

According to an embodiment of the present disclosure, a technology-related dispute may include a technology-related dispute between individuals, between individuals and groups, or between groups and groups. Specifically, a technology-related dispute may include at least one of intellectual property-related judgment, intellectual property-related civil litigation, intellectual property-related criminal litigation, or intellectual property-related administrative litigation. Intellectual property-related trials may include intellectual property-related trials against rejection decisions, cancellation trials, or invalidation trials. For example, trials related to intellectual property rights may include trials against patent rejection decisions, trials against trademark rejection decisions, patent cancellation applications, trademark cancellation trials, patent invalid trials, trademark invalid trials, and the like. Civil actions related to intellectual property rights may include infringement actions related to intellectual property rights. Right holders may claim damages from the infringer through intellectual property-related infringement lawsuits. Civil actions related to intellectual property rights may include, for example, patent infringement actions, trademark infringement actions, design infringement actions, and copyright infringement actions. Criminal litigation related to intellectual property rights includes intellectual property infringement litigation, and right holders can sue infringers, and prosecutors can file public charges and seek criminal penalties against violators. Administrative litigation related to intellectual property rights may include litigation for administrative dispositions related to intellectual property rights. Administrative litigation related to intellectual property rights may include, for example, a trial decision cancellation lawsuit filed with a court against an administrative disposition called a trial decision. The foregoing is only an example, and the present disclosure is not limited thereto.

According to an embodiment of the present disclosure, the dispute resolution basis document may include documents that may be helpful in resolving technology-related disputes. The document based on dispute resolution may include a document that is helpful in the process of resolving the dispute by sending it to at least one of the parties to the dispute or the dispute resolution institution. The document based on dispute resolution may include at least one of a warning letter related to intellectual property rights, a document proving invalidity of intellectual property rights, a document proving evidence of a judgment to revoke intellectual property rights, or a document proving evidence of litigation related to intellectual property rights. Warning letters related to intellectual property rights may include warning letters sent to others when others infringe or are likely to infringe on intellectual property rights. A warning notice related to intellectual property rights may include, for example, a warning notice based on patent rights, a warning notice based on trademark rights, a warning notice based on design protection laws, a warning notice based on copyright, and the like. Evidence documents for invalidation of intellectual property rights may include documents necessary to invalidate intellectual property rights. Evidence documents for invalidation of intellectual property rights may include documents published prior to the commencement date of legal protection of specific intellectual property rights. Evidence documents for invalidation of intellectual property rights may include, for example, documents published prior to the patent filing date. Evidence of an intellectual property right revocation trial may include documents necessary to revoke an intellectual property right. Evidence of an intellectual property right cancellation trial may include, for example, documents published prior to the patent application date. Evidence documents in litigation related to intellectual property rights may include documents that may restrict the exercise of certain intellectual property rights in infringement proceedings, invalidation trials, and trial decision cancellation proceedings. Evidence documents in intellectual property litigation may include, for example, documents that can be used for invalidity defenses and free technology defenses. The foregoing is only an example, and the present disclosure is not limited thereto.

According to one embodiment of the present disclosure, processor 110 may determine a dispute resolution basis document. Comparative Document A 311, Comparative Document B 312, Comparable Document C ( 313) may exist. The processor 110 determines the degree of similarity between the first partial document 210 and the comparison target document A 311, the comparison target document B 312, and the comparison target document C 313 by using the technology similarity determination model 250. can be obtained. For example, the processor 110 uses the technology similarity determination model 250 to determine the similarity between the first partial document 210 and the comparison target document A 311 as 0.98 (610) and the first partial document 210 and the first partial document 210. The similarity between the comparison target document B 312 is 0.95 (620) and the similarity between the first partial document 210 and the comparison target document C 313 is 0.31 (630). In this case, the processor 110 may determine a comparison target document A 311 and a comparison target document B 312 having a high similarity to the first partial document 210 as dispute resolution basis documents. The processor 110 may transmit the document determined as the dispute resolution basis document to the user terminal. Accordingly, the user may use the comparison target document A 311 and the comparison target document B 312 as a basis document for dispute resolution. For example, the user may invalidate the patent right of the right holder of the first part document 210 by filing a patent invalidation trial based on the comparison target document A 311 and the comparison target document B 312. . In addition, the user may apply for patent cancellation by submitting the comparison target document A 311 and the comparison target document B 312 to the Korean Intellectual Property Office with respect to the right holder of the first partial document 210 . As another example, in an infringement action filed by the right holder who has obtained a patent registration for the first part document 210, the user can file a defense of abuse or invalidity based on the document A 311 and the document B 312 compared to the document B 312. You can protest. Therefore, users can quickly and easily find documents based on dispute resolution using deep learning technology, protect their rights and interests, and terminate disputes related to intellectual property rights early. The foregoing is only an example, and the present disclosure is not limited thereto.

2 is a schematic diagram illustrating a network function for data processing according to an embodiment of the present disclosure.

Throughout this specification, computational model, neural network, network function, and neural network may be used interchangeably. A neural network may consist of a set of interconnected computational units, which may generally be referred to as nodes. These nodes may also be referred to as neurons. A neural network includes at least one or more nodes. Nodes (or neurons) constituting neural networks may be interconnected by one or more links.

In a neural network, one or more nodes connected through a link may form a relative relationship of an input node and an output node. The concept of an input node and an output node is relative, and any node in an output node relationship to one node may be in an input node relationship to another node, and vice versa. As described above, an input node to output node relationship may be created around a link. More than one output node can be connected to one input node through a link, and vice versa.

In a relationship between an input node and an output node connected through one link, the value of data of the output node may be determined based on data input to the input node. Here, a link interconnecting an input node and an output node may have a weight. The weight may be variable, and may be changed by a user or an algorithm to perform a function desired by the neural network. For example, when one or more input nodes are connected to one output node by respective links, the output node is set to a link corresponding to values input to input nodes connected to the output node and respective input nodes. An output node value may be determined based on the weight.

As described above, in the neural network, one or more nodes are interconnected through one or more links to form an input node and output node relationship in the neural network. Characteristics of the neural network may be determined according to the number of nodes and links in the neural network, an association between the nodes and links, and a weight value assigned to each link. For example, when there are two neural networks having the same number of nodes and links and different weight values of the links, the two neural networks may be recognized as different from each other.

A neural network may be composed of a set of one or more nodes. A subset of nodes constituting a neural network may constitute a layer. Some of the nodes constituting the neural network may form one layer based on distances from the first input node. For example, a set of nodes having a distance of n from the first input node may constitute n layers. The distance from the first input node may be defined by the minimum number of links that must be passed through to reach the corresponding node from the first input node. However, the definition of such a layer is arbitrary for explanation, and the order of a layer in a neural network may be defined in a method different from that described above. For example, a layer of nodes may be defined by a distance from a final output node.

An initial input node may refer to one or more nodes to which data is directly input without going through a link in relation to other nodes among nodes in the neural network. Alternatively, in a relationship between nodes based on a link in a neural network, it may mean nodes that do not have other input nodes connected by a link. Similarly, the final output node may refer to one or more nodes that do not have an output node in relation to other nodes among nodes in the neural network. In addition, the hidden node may refer to nodes constituting the neural network other than the first input node and the last output node.

In the neural network according to an embodiment of the present disclosure, the number of nodes in the input layer may be the same as the number of nodes in the output layer, and the number of nodes decreases and then increases again as the number of nodes progresses from the input layer to the hidden layer. can In addition, the neural network according to another embodiment of the present disclosure may be a neural network in which the number of nodes of the input layer may be less than the number of nodes of the output layer and the number of nodes decreases as the number of nodes increases from the input layer to the hidden layer. there is. In addition, the neural network according to another embodiment of the present disclosure is a neural network in which the number of nodes of the input layer may be greater than the number of nodes of the output layer, and the number of nodes increases as the number of nodes increases from the input layer to the hidden layer. can A neural network according to another embodiment of the present disclosure may be a neural network in the form of a combination of the aforementioned neural networks.

A deep neural network (DNN) may refer to a neural network including a plurality of hidden layers in addition to an input layer and an output layer. Deep neural networks can reveal latent structures in data. In other words, it can identify the latent structure of a photo, text, video, sound, or music (e.g., what objects are in the photo, what the content and emotion of the text are, what the content and emotion of the audio are, etc.). . Deep neural networks include convolutional neural networks (CNNs), recurrent neural networks (RNNs), auto encoders, generative adversarial networks (GANs), and restricted boltzmann machines (RBMs). machine), a deep belief network (DBN), a Q network, a U network, a Siamese network, a Generative Adversarial Network (GAN), and the like. The description of the deep neural network described above is only an example, and the present disclosure is not limited thereto.

In one embodiment of the present disclosure, the network function may include an autoencoder. An autoencoder may be a type of artificial neural network for outputting output data similar to input data. An auto-encoder may include at least one hidden layer, and an odd number of hidden layers may be disposed between input and output layers. The number of nodes of each layer may be reduced from the number of nodes of the input layer to an intermediate layer called the bottleneck layer (encoding), and then expanded symmetrically with the reduction from the bottleneck layer to the output layer (symmetrical to the input layer). Autoencoders can perform non-linear dimensionality reduction. The number of input layers and output layers may correspond to dimensions after preprocessing of input data. In the auto-encoder structure, the number of hidden layer nodes included in the encoder may decrease as the distance from the input layer increases. If the number of nodes in the bottleneck layer (the layer with the fewest nodes located between the encoder and the decoder) is too small, a sufficient amount of information may not be conveyed, so more than a certain number (e.g., more than half of the input layer, etc.) ) may be maintained.

The neural network may be trained using at least one of supervised learning, unsupervised learning, semi-supervised learning, and reinforcement learning. Learning of the neural network may be a process of applying knowledge for the neural network to perform a specific operation to the neural network.

A neural network can be trained in a way that minimizes output errors. In learning the neural network, the training data is repeatedly input into the neural network, the output of the neural network for the training data and the error of the target are calculated, and the error of the neural network is transferred from the output layer of the neural network to the input layer in the direction of reducing the error. It is a process of updating the weight of each node of the neural network by backpropagating in the same direction. In the case of teacher learning, the learning data in which the correct answer is labeled is used for each learning data (ie, the labeled learning data), and in the case of comparative teacher learning, the correct answer may not be labeled in each learning data. That is, for example, learning data in the case of teacher learning regarding data classification may be data in which each learning data is labeled with a category. Labeled training data is input to the neural network, and an error may be calculated by comparing an output (category) of the neural network and a label of the training data. As another example, in the case of comparative history learning for data classification, an error may be calculated by comparing input learning data with a neural network output. The calculated error is back-propagated in a reverse direction (ie, from the output layer to the input layer) in the neural network, and the connection weight of each node of each layer of the neural network may be updated according to the back-propagation. The amount of change in the connection weight of each updated node may be determined according to a learning rate. The neural network's calculation of input data and backpropagation of errors can constitute a learning cycle (epoch). The learning rate may be applied differently according to the number of iterations of the learning cycle of the neural network. For example, a high learning rate is used in the early stage of neural network training to increase efficiency by allowing the neural network to quickly achieve a certain level of performance, and a low learning rate can be used in the late stage to increase accuracy.

In neural network learning, generally, training data can be a subset of real data (ie, data to be processed using the trained neural network), and therefore, errors for training data are reduced, but errors for real data are reduced. There may be incremental learning cycles. Overfitting is a phenomenon in which errors for actual data increase due to excessive learning on training data. For example, a phenomenon in which a neural network that has learned a cat by showing a yellow cat does not recognize that it is a cat when it sees a cat other than yellow may be a type of overfitting. Overfitting can act as a cause of increasing errors in machine learning algorithms. Various optimization methods can be used to prevent such overfitting. To prevent overfitting, methods such as increasing the training data, regularization, inactivating some nodes in the network during learning, and using a batch normalization layer should be applied. can

According to an embodiment of the present disclosure, a computer readable medium storing a data structure is disclosed.

Data structure can refer to the organization, management, and storage of data that enables efficient access and modification of data. Data structure may refer to the organization of data to solve a specific problem (eg, data retrieval, data storage, data modification in the shortest time). A data structure may be defined as a physical or logical relationship between data elements designed to support a specific data processing function. A logical relationship between data elements may include a connection relationship between data elements that a user thinks. A physical relationship between data elements may include an actual relationship between data elements physically stored on a computer-readable storage medium (eg, a hard disk). The data structure may specifically include a set of data, a relationship between data, and a function or command applicable to the data. Through an effectively designed data structure, a computing device can perform calculations while using minimal resources of the computing device. Specifically, the computing device can increase the efficiency of operation, reading, insertion, deletion, comparison, exchange, and search through an effectively designed data structure.

The data structure can be divided into a linear data structure and a non-linear data structure according to the shape of the data structure. A linear data structure may be a structure in which only one data is connected after one data. Linear data structures may include lists, stacks, queues, and decks. A list may refer to a series of data sets in which order exists internally. The list may include a linked list. A linked list may be a data structure in which data are connected in such a way that each data is connected in a line with a pointer. In a linked list, a pointer can contain information about connection to the next or previous data. A linked list can be expressed as a singly linked list, a doubly linked list, or a circular linked list depending on the form. A stack can be a data enumeration structure that allows limited access to data. A stack can be a linear data structure in which data can be processed (eg, inserted or deleted) at only one end of the data structure. The data stored in the stack may be a LIFO-Last in First Out (Last in First Out) data structure. A queue is a data listing structure that allows limited access to data, and unlike a stack, it can be a data structure (FIFO-First in First Out) in which data stored later comes out later. A deck can be a data structure that can handle data from either end of the data structure.

The nonlinear data structure may be a structure in which a plurality of data are connected after one data. The non-linear data structure may include a graph data structure. A graph data structure can be defined as a vertex and an edge, and an edge can include a line connecting two different vertices. A graph data structure may include a tree data structure. The tree data structure may be a data structure in which one path connects two different vertices among a plurality of vertices included in the tree. That is, it may be a data structure that does not form a loop in a graph data structure.

Throughout this specification, computational model, neural network, network function, and neural network may be used interchangeably. (Hereinafter, it is unified and described as a neural network.) The data structure may include a neural network. And the data structure including the neural network may be stored in a computer readable medium. The data structure including the neural network may also include data input to the neural network, weights of the neural network, hyperparameters of the neural network, data obtained from the neural network, an activation function associated with each node or layer of the neural network, and a loss function for learning the neural network. there is. A data structure including a neural network may include any of the components described above. In other words, the data structure including the neural network includes all or all of the data input to the neural network, weights of the neural network, hyperparameters of the neural network, data obtained from the neural network, activation function associated with each node or layer of the neural network, and loss function for training the neural network. It may be configured to include any combination of. In addition to the foregoing configurations, the data structure comprising the neural network may include any other information that determines the characteristics of the neural network. In addition, the data structure may include all types of data used or generated in the computational process of the neural network, but is not limited to the above. A computer readable medium may include a computer readable recording medium and/or a computer readable transmission medium. A neural network may consist of a set of interconnected computational units, which may generally be referred to as nodes. These nodes may also be referred to as neurons. A neural network includes one or more nodes.

The data structure may include data input to the neural network. A data structure including data input to the neural network may be stored in a computer readable medium. Data input to the neural network may include training data input in a neural network learning process and/or input data input to the neural network after learning has been completed. Data input to the neural network may include data that has undergone pre-processing and/or data to be pre-processed. Pre-processing may include a data processing process for inputting data to a neural network. Accordingly, the data structure may include data subject to preprocessing and data generated by preprocessing. The foregoing data structure is only an example, and the present disclosure is not limited thereto.

The data structure may include data input to or output from the neural network. A data structure including data input or output to the neural network may be stored in a computer readable medium. A data structure stored in a computer readable medium may include data input during a neural network inference process or output data output as a result of neural network inference. Also, since the data structure may include data processed by a specific data processing method, it may include data before and after processing. Accordingly, the data structure may include data subject to processing and data processed through a data processing method.

The data structure may include the weights of the neural network. (In this specification, weights and parameters may be used in the same meaning.) Also, a data structure including weights of a neural network may be stored in a computer readable medium. A neural network may include a plurality of weights. The weight may be variable, and may be changed by a user or an algorithm to perform a function desired by the neural network. For example, when one or more input nodes are connected to one output node by respective links, the output node is set to a link corresponding to values input to input nodes connected to the output node and respective input nodes. An output node value can be determined based on the parameters. The foregoing data structure is only an example, and the present disclosure is not limited thereto.

As a non-limiting example, the weights may include weights that are varied during neural network training and/or weights for which neural network training is completed. The variable weight in the neural network learning process may include a weight at the time when the learning cycle starts and/or a variable weight during the learning cycle. The weights for which neural network learning has been completed may include weights for which learning cycles have been completed. Accordingly, the data structure including the weights of the neural network may include a data structure including weights that are variable during the neural network learning process and/or weights for which neural network learning is completed. Therefore, it is assumed that the above-described weights and/or combinations of weights are included in the data structure including the weights of the neural network. The foregoing data structure is only an example, and the present disclosure is not limited thereto.

The data structure including the weights of the neural network may be stored in a computer readable storage medium (eg, a memory or a hard disk) after going through a serialization process. Serialization can be the process of converting a data structure into a form that can be stored on the same or another computing device and later reconstructed and used. A computing device may serialize data structures to transmit and receive data over a network. The data structure including the weights of the serialized neural network may be reconstructed on the same computing device or another computing device through deserialization. The data structure including the weights of the neural network is not limited to serialization. Furthermore, the data structure including the weights of the neural network is a data structure for increasing the efficiency of operation while minimizing the resource of the computing device (for example, B-Tree, Trie, m-way search tree, AVL tree, Red-Black Tree). The foregoing is only an example and the present disclosure is not limited thereto.

The data structure may include hyper-parameters of the neural network. Also, the data structure including the hyperparameters of the neural network may be stored in a computer readable medium. A hyperparameter may be a variable variable by a user. Hyperparameters include, for example, learning rate, cost function, number of learning cycle iterations, weight initialization (eg, setting the range of weight values to be targeted for weight initialization), hidden unit number (eg, the number of hidden layers and the number of nodes in the hidden layer). The foregoing data structure is only an example, and the present disclosure is not limited thereto.

3 is an exemplary diagram illustrating a technology similarity determination model obtained by learning a correlation between a first partial document and a second partial document.

In FIG. 3 , a first partial document 210 , a second partial document 230 , a technology similarity determination model 250 , and a similarity 270 are illustrated.

According to an embodiment of the present disclosure, the technology similarity determination model 250 may include a model learned to determine a degree of similarity between the first partial document 210 and the second partial document 230 . The model learned to determine the degree of similarity between the first partial document 210 and the second partial document 230 is a machine learning model, for example, a convolutional neural network (CNN), a recurrent neural network (RNN), or an LSTM ( Long Short Term Memory), GRU (Gated Recurrent Unit), FCN (Fully Connected Layer), Activation Layer, and the like may be included. The technology similarity determination model described above is only an example, and the present disclosure is not limited thereto.

According to an embodiment of the present disclosure, the technology similarity determination model 250 inputs the first partial document 210 and the second partial document 230 to the technology similarity determination model 250 to determine the first partial document ( 210) and a first model obtained by learning the correlation between the contents of the second partial document 230. The computing device 100 may perform preprocessing to convert the first partial document 210 and the second partial document 230 into a form to be input to the technology similarity determination model 250 . For example, the computing device 100 may perform a tokenization process of dividing sentences included in the first partial document 210 and the second partial document 230 into tokens according to a predetermined criterion. can As another example, the computing device 100 converts sentences included in the first partial document and the second partial document into a table form (eg, each word in the sentence is a table row, and each word vector is a table column). expressed table) can be preprocessed. The foregoing is only an example and the present disclosure is not limited thereto.

According to an embodiment of the present disclosure, the first model may include a model obtained by learning a correlation between the contents of the first partial document 210 and the contents of the second partial document 230 . Accordingly, the computing device 100 may acquire a degree of similarity by inputting the first partial document 210 and the second partial document to the first model. For example, the first model may include a model learned using a patent specification as a learning data set. The training data set includes a first sub-learning data set comprising a first part document (eg claims) and a second sub-learning data set comprising a second part document (eg specific description of an invention). can do. Accordingly, the first model may include a model learned using the first sub-training data set and the second sub-training data set. According to an embodiment of the present disclosure, the technology similarity determination model may include a first model. Accordingly, the computing device 100 may obtain the degree of similarity 270 based on the correlation between the contents of the first partial document and the second partial document by using the technology similarity determination model. For example, when the correlation between the contents of the first partial document and the second partial document is high, the degree of similarity 270 obtained by the computing device 100 may be high. The aforementioned first model is only an example, and the present disclosure is not limited thereto.

According to an embodiment of the present disclosure, the computing device 100 may obtain a similarity 270 by inputting the first partial document 210 and the second partial document 230 into the technology similarity determination model 250. . The similarity 270 may include a similarity between the first partial document and the second partial document or at least one of the comparison target documents. For example, if the first part document is the claims of a patent application and the second part document is a detailed description of an invention contained in that patent application, between the first part document and at least one of the second part document or the comparative document The degree of similarity may be high. Since the detailed description of the invention included in the patent application contains content to support the content included in the claims, the degree of similarity between the first part document and the second part document may be high. According to an embodiment of the present disclosure, the degree of similarity 270 may be expressed in a form of probability. For example, when the technical similarity between the first partial document 210 and the second partial document is high, the similarity 270 may be 0.96. The foregoing is only an example and the present disclosure is not limited thereto.

4 is an exemplary diagram illustrating a technology similarity judgment model obtained by learning a correlation between a first partial document and a comparison target document.

In FIG. 4 , a first partial document 210 , a comparison target document 310 , a technology similarity determination model 250 , and a similarity 270 are illustrated.

According to an embodiment of the present disclosure, the technology similarity determination model 250 inputs the first partial document 210 and the comparison target document 310 to the first model and compares them with the contents of the first partial document 210. A second model obtained by learning the correlation between contents of the target document 310 may be included.

According to an embodiment of the present disclosure, the second model may include a model obtained by learning a correlation between the contents of the first partial document 210 and the contents of the comparison target document 310 . The second model may include a model obtained by re-training the first model obtained by learning the correlation between the content of the first partial document and the content of the second partial document. The computing device 100 inputs the first partial document 210 and the comparison target document 310 to the first model and learns the first model, thereby obtaining the contents of the first partial document 210 and the contents of the second partial document. In addition to the correlation between the contents of the first partial document 210 and the contents of the comparison target document 310, a learned model may be acquired. The computing device 100 learns the correlation between the contents of the first part document 210 and the comparison target document 310 by a learning method of fixing at least some of the parameters included in the first model and updating the remaining parameters. can make it The aforementioned second model is only an example, and the present disclosure is not limited thereto.

According to another embodiment of the present disclosure, the technology similarity determination model 250 may include a second model. Accordingly, the computing device 100 may obtain the similarity 270 based on the correlation between the contents of the first partial document 210 and the comparison target document 310 using the technology similarity determination model 250 . For example, when the correlation between the contents of the first partial document 210 and the comparison target document 310 is high, the degree of similarity 270 may be high. That is, the computing device 100 uses the technology similarity determination model 250 to determine the similarity 270 as the technical idea included in the first partial document 210 and the technical idea included in the comparison target document 310 are similar. can be obtained. The foregoing is only an example, and the present disclosure is not limited thereto.

5 is an exemplary diagram illustrating a technology similarity determination model by way of example.

5, a first partial document 210, a second partial document 230, a first sub-model 410, a second sub-model 430, a third sub-model 450, and a fourth sub-model 470 , similarity 270 is shown.

According to an embodiment of the present disclosure, the first submodel 410 may include a recurrent neural network. A recurrent neural network may include a long short-term memory network. The computing device 100 may extract features of the first partial document 210 by inputting the first partial document to the long and short term storage network. A feature of the first part document may include a first submodel encoding matrix. The first submodel encoding matrix may include a vector obtained by inputting a document into a long and short term storage network. The obtained vector may include a vector having a probability value of each word as a vector element. The computing device 100 may obtain the first submodel encoding matrix by using the vector obtained from the first submodel 410 . For example, if a set of words included in the first sub-document is (x1, x2, x3, x4, ... , xn) and a vector corresponding to each word (the processor assigns words to the first submodel 410), When the vector obtained by input) is (v1, v2, v3, v4, ... ,vn), the first submodel encoding matrix will be expressed as [v1, v2, v3, v4, ... ,vn] can The aforementioned first submodel is only an example, and the present disclosure is not limited thereto.

According to an embodiment of the present disclosure, the second submodel 430 may include a recurrent neural network. A recurrent neural network may include a long short-term memory network. The long and short term storage network of the second submodel 430 may include the same network as the short and long term storage network of the first submodel 410 . That is, parameters of the long and short term storage network of the second submodel 430 may be the same as parameters of the short and long term storage network of the first submodel 410 . Since the first sub-model 410 and the second sub-model 430 share parameters of the network, the computing device 100 can transmit the same word sequence (eg, sentence) to the first sub-model 410 and the second sub-model 410 . Input to submodel 430 to obtain the same or similar vector representation. The aforementioned second submodel is only an example, and the present disclosure is not limited thereto.

According to one embodiment of the present disclosure, the computing device 100 may input the second partial document 230 to a short-term storage network to extract features of the second partial document 230 . A feature of the second part document 230 may include a second submodel encoding matrix. The second submodel encoding matrix may include a vector obtained by inputting a document into a long-term and short-term storage network. The obtained vector may include a vector having a probability value of each word as a component of the vector. The computing device 100 may obtain the second sub-model encoding matrix by using the vector obtained from the second sub-model. For example, if the set of words included in the second sub-document is (y1, y2, y3, y4, ... , yn), and a vector corresponding to each word (obtained by the processor inputting words to the second submodel) If one vector) is (v'1, v'2, v'3, v'4, ... , v'n), the second submodel encoding matrix is [v'1, v'2, v' 3, v'4, ... , v'n]. The aforementioned first sub-model is only an example, and the present disclosure is not limited thereto.

According to an embodiment of the present disclosure, the third sub-model 450 may include a model capable of deriving a correlation between a first partial document and a second partial document. The computing device 100 determines the correlation between the first partial document and the second partial document based on the outputs of the first sub-model 410 and the second sub-model 430 using the third sub-model 450. can be derived The computing device 100 may derive a correlation between the first partial document and the second partial document based on the first submodel encoding matrix and the second submodel encoding matrix. The computing device 100 inputs a result obtained by calculating the transpose of the second submodel encoding matrix and the first submodel encoding matrix to the softmax function, so that words included in the first part document and the second part Correlation of words included in a document can be obtained. The aforementioned third sub-model is only an example, and the present disclosure is not limited thereto.

Description of the specific configuration of the technology similarity judgment model is discussed in detail in the patent "Data processing method using neural network" (application number: KR10-2019-0103499), which is incorporated as a reference in its entirety in this application.

According to an embodiment of the present disclosure, the fourth sub-model 470 may include a model for obtaining a similarity between the first partial document 210 and the second partial document 230 . The fourth sub-model may include a machine learning model and/or a rule-based model for obtaining a degree of similarity. According to an embodiment of the present disclosure, the computing device 100 uses the fourth sub-model 470 to generate sentences and/or sentences included in the second partial document 230 having a high correlation with the first partial document 210 . Or you can get a word. In another embodiment of the present disclosure, the computing device 100 may obtain a degree of similarity based on a correlation score for each sentence and/or word having a correlation with the first partial document 210 equal to or greater than a predetermined criterion. . The degree of similarity is described in detail below. The aforementioned fourth sub-model is only an example, and the present disclosure is not limited thereto.

According to an embodiment of the present disclosure, the degree of similarity 270 may include a degree of similarity between a specific document and another document. The similarity 270 may include a similarity between technical ideas included in a specific document and technical ideas included in other documents. Therefore, the degree of similarity may be high if the technical idea is similar and the technical idea included in a specific document is imitated or infringed. The similarity 270 may include a similarity between a first partial document and a second partial document or at least one of a comparison target document. Also, the similarity 270 may include a similarity between the first partial document and the comparison target document. The above-described degree of similarity is only an example, and the present disclosure is not limited thereto.

6 is an exemplary diagram for explaining a process of technology monitoring and determining a document based on dispute resolution by way of example.

In FIG. 6, the first part document 210, the determination of the document based on dispute resolution 520, the technology monitoring 510, and the similarity between the comparison target document and the first part document 610, 620, 630, 640, 650, 660 this is shown

According to one embodiment of the present disclosure, computing device 100 may perform technology monitoring 510 . The technology monitoring 510 may include an operation of monitoring whether or not a subsequent technology similar to the technology possessed by the user is disclosed. Users may not have enough time and money to determine who imitates and uses their technology without permission through Internet searches every day. Therefore, the computing device 100 automatically determines the comparison target documents that are disclosed using the technology similarity determination model 250 with the first part document including the user's technology and transmits the similarity determination result to the user, so that the user can conveniently You can easily spot others who are infringing on your rights. The foregoing technical monitoring is only an example, and the present disclosure is not limited thereto.

According to an embodiment of the present disclosure, the computing device 100 provides a disclosure after the legal protection start date of the first part document (eg, the patent application date if the patent is registered) of the first part document. Technical monitoring may be performed on the documentation. Documents disclosed after the legal protection start date of the first part document may include a comparison target document D 314 , a comparison target document E 315 , and a comparison target document F 316 . The computing device 100 may calculate a similarity between the first partial document 210 and a comparison target document D 314 , a comparison target document E 315 , or a comparison target document F 316 . For example, the computing device 100 calculates that the similarity between the first partial document 210 and the comparison target document D 314 is 0.97 (640) and the similarity between the first partial document 210 and the comparison target document E 315. is 0.13 (650), and the similarity between the first partial document 210 and the comparison target document F 316 is 0.92 (660). In this case, since the comparison target document D 314 and the comparison target document F 316 have a high degree of similarity, they may include the same or similar technical ideas as the first partial document 210 . The foregoing is only an example, and the present disclosure is not limited thereto.

According to an embodiment of the present disclosure, the operation of monitoring the technology may include an operation of determining a comparison target document as a rights infringing document based on a predetermined rights infringement criterion. The predetermined criterion for determining infringement of rights may include a criterion determined based on whether the comparison target document includes the same or similar technical idea as the first part document. For example, the predetermined criterion for determining infringement of rights may include a criterion for determining a document as infringing upon a right if the degree of similarity is 0.7 or higher. The predetermined rights infringement judgment criterion may include an operator or a criterion determined based on data analysis for determining similarity. Accordingly, the predetermined rights infringement judgment criterion may be a criterion for determining a comparison target document having a similarity of 0.7 or higher as a rights infringement document. In this case, the computing device 100 may determine the comparison target document D 314 and the comparison target document F 316 as documents infringing on rights. The foregoing is only an example, and the present disclosure is not limited thereto.

According to an embodiment of the present disclosure, the operation of performing technology monitoring may include transmitting information related to the determined infringing document to at least one of a terminal of a right holder having a right to the document of the first part or an agent terminal of the holder of the right to the document of the first part. It may include an operation to transmit to one. The computing device 100 may transmit the document to be compared, which is determined to be a document infringing on rights, to at least one of a terminal of a right holder having a right to the first partial document or an agent terminal of the right holder to the first partial document. Through this, the holder of the right or the agent for the document of the first part can continuously receive information on the document infringing on rights through the user terminal. Therefore, the right holder or agent for the first part document can selectively determine whether the technology is infringed based on the information received without having to continuously search for infringing technology through Internet searches, saving both time and money. there is. The foregoing is only an example, and the present disclosure is not limited thereto.

According to an embodiment of the present disclosure, the computing device 100 may determine a dispute resolution basis document for resolving a technology-related dispute. An owner with rights to a Part 1 document may send a notice of infringement or initiate an infringement action against a person practicing a particular technology. In this case, the person practicing the particular technology may invalidate the rights to the Part 1 document or make a defense that the technology contained in the Part 1 document is publicly available technology in order to conduct and defend its business. . In this case, a person practicing a particular technology may have a need to find a document that was published prior to the first part document 210 legal protection commencement date. If the technical idea contained in the document of the first part exists in a document disclosed before the legal protection start date of the first part document 210, the right to the document of the first part may be invalidated. You may not be able to claim your rights to others. Through this, disputes between rightholders and implementers may be resolved. The foregoing is only an example, and the present disclosure is not limited thereto.

According to an embodiment of the present disclosure, a technology-related dispute may include a technology-related dispute between individuals, between individuals and groups, or between groups and groups. Specifically, a technology-related dispute may include at least one of intellectual property-related judgment, intellectual property-related civil litigation, intellectual property-related criminal litigation, or intellectual property-related administrative litigation. Intellectual property-related trials may include intellectual property-related trials against rejection decisions, cancellation trials, or invalidation trials. For example, trials related to intellectual property rights may include trials against patent rejection decisions, trials against trademark rejection decisions, patent cancellation applications, trademark cancellation trials, patent invalid trials, trademark invalid trials, and the like. Civil actions related to intellectual property rights may include infringement actions related to intellectual property rights. Right holders may claim damages from the infringer through intellectual property-related infringement lawsuits. Civil actions related to intellectual property rights may include, for example, patent infringement actions, trademark infringement actions, design infringement actions, and copyright infringement actions. Criminal litigation related to intellectual property rights includes intellectual property infringement litigation, and right holders can sue infringers, and prosecutors can file public charges and seek criminal penalties against violators. Administrative litigation related to intellectual property rights may include litigation for administrative dispositions related to intellectual property rights. Administrative litigation related to intellectual property rights may include, for example, a trial decision cancellation lawsuit filed with a court against an administrative disposition called a trial decision. The foregoing is only an example, and the present disclosure is not limited thereto.

According to an embodiment of the present disclosure, the dispute resolution basis document may include documents that may be helpful in resolving technology-related disputes. The document based on dispute resolution may include a document that is helpful in the process of resolving the dispute by sending it to at least one of the parties to the dispute or the dispute resolution institution. The document based on dispute resolution may include at least one of a warning letter related to intellectual property rights, a document proving invalidity of intellectual property rights, a document proving evidence of a judgment to revoke intellectual property rights, or a document proving evidence of litigation related to intellectual property rights. Warning letters related to intellectual property rights may include warning letters sent to others when others infringe or are likely to infringe on intellectual property rights. A warning notice related to intellectual property rights may include, for example, a warning notice based on patent rights, a warning notice based on trademark rights, a warning notice based on design protection laws, a warning notice based on copyright, and the like. Evidence documents for invalidation of intellectual property rights may include documents necessary to invalidate intellectual property rights. Evidence documents for invalidation of intellectual property rights may include documents published prior to the commencement date of legal protection of specific intellectual property rights. Evidence documents for invalidation of intellectual property rights may include, for example, documents published prior to the patent filing date. Evidence of an intellectual property right revocation trial may include documents necessary to revoke an intellectual property right. Evidence of an intellectual property right cancellation trial may include, for example, documents published prior to the patent application date. Evidence documents in litigation related to intellectual property rights may include documents that may restrict the exercise of certain intellectual property rights in infringement proceedings, invalidation trials, and trial decision cancellation proceedings. Evidence documents in intellectual property litigation may include, for example, documents that can be used for invalidity defenses and free technology defenses. The foregoing is only an example, and the present disclosure is not limited thereto.

According to one embodiment of the present disclosure, computing device 100 may determine a dispute resolution basis document. Comparative Document A 311, Comparative Document B 312, Comparable Document C ( 313) may exist. The computing device 100 uses the technology similarity determination model 250 for the first partial document 210 and the comparison target document A 311, comparison target document B 312, and comparison target document C 313, respectively. Similarity can be obtained. For example, the computing device 100 uses the technology similarity determination model 250 to determine the similarity between the first partial document 210 and the comparison target document A 311 as 0.98 (610) and the first partial document 210 The similarity between the document B 312 and the document B 312 to be compared is 0.95 (620), and the similarity between the first partial document 210 and the document C 313 to be compared is 0.31 (630). In this case, the computing device 100 may determine a comparison target document A 311 and a comparison target document B 312 having a high similarity to the first partial document 210 as dispute resolution basis documents. The computing device 100 may transmit the document determined as the dispute resolution basis document to the user terminal. Accordingly, the user may use the comparison target document A 311 and the comparison target document B 312 as a basis document for dispute resolution. For example, the user may invalidate the patent right of the right holder of the first part document 210 by filing a patent invalidation trial based on the comparison target document A 311 and the comparison target document B 312. . In addition, the user may apply for patent cancellation by submitting the comparison target document A 311 and the comparison target document B 312 to the Korean Intellectual Property Office with respect to the right holder of the first partial document 210 . As another example, in an infringement action filed by the right holder who has obtained a patent registration for the first part document 210, the user can file a defense of abuse or invalidity based on the document A 311 and the document B 312 compared to the document B 312. You can protest. Therefore, users can quickly and easily find documents based on dispute resolution using deep learning technology, protect their rights and interests, and terminate disputes related to intellectual property rights early. The foregoing is only an example, and the present disclosure is not limited thereto.

7 is a flowchart of a method for determining technology similarity according to an embodiment of the present disclosure.

According to an embodiment of the present disclosure, the computing device 100 determines the degree of similarity between the first partial document and the second partial document or at least one of the comparison target document using a technology similarity determination model learned to determine the degree of similarity between the first partial document and the second partial document. At least one part or at least one of the documents to be compared may be input (710).

According to an embodiment of the present disclosure, the document may include at least one of a first part document for claiming a right or a second part document for explaining content of claiming a right.

According to an embodiment of the present disclosure, the technology similarity determination model inputs the first partial document and the second partial document to the technology similarity determination model, and determines the correlation between the contents of the first partial document and the contents of the second partial document. It may include a first model that has been trained.

According to an embodiment of the present disclosure, the technology similarity determination model is configured by inputting the first partial document and the comparison target document to the first model and learning the correlation between the contents of the first partial document and the comparison target document. models can be included.

According to an embodiment of the present disclosure, the comparison target document may include at least one of a first comparison target document that is an original comparison target document and a second comparison target document that is a document obtained by processing the original comparison target document.

According to an embodiment of the present disclosure, the first comparison target document is a document including technology-related content, and may include at least one of an intellectual property-related document, a thesis, or a technical document.

According to an embodiment of the present disclosure, the second comparison target document includes at least some of the plurality of sentences included in the first comparison target document, and at least some of the plurality of sentences included in the first comparison target document are key words. It may include a sentence determined based on a sentence calculation algorithm.

According to an embodiment of the present disclosure, a core sentence calculation algorithm may include calculating a technical term importance based on a technical term appearance frequency for each technical term in a sentence included in a first comparison target document; obtaining an inter-sentence weight matrix by calculating a first matrix having each calculated technical term importance as an element of the matrix and a transposed matrix of the first matrix; calculating a rank of a sentence included in the first comparison target document based on an inter-sentence weight matrix; and determining a sentence having a calculated rank equal to or higher than a predetermined rank as a core sentence.

According to an embodiment of the present disclosure, documents to be compared may include updated documents at predetermined intervals based on a specific point in time in order to determine a similarity of technology with respect to documents to be compared that are disclosed over time.

According to an embodiment of the present disclosure, the technology similarity determination model may include a first sub-model for extracting features of a first partial document; a second sub-model for extracting a feature of a second partial document or a feature of the comparison target document; a third sub-model for deriving a correlation between the first partial document and the comparison target document based on the output of the first sub-model and the output of the second sub-model; and a fourth sub-model for acquiring the technology similarity based on correlation.

According to an embodiment of the present disclosure, the technical similarity is a score determined based on the number of sentences included in comparison target documents having a correlation with sentences included in the first partial document or higher than a predetermined criterion or included in the first partial document. and at least one of scores determined based on correlation scores of each of the corresponding sentence and at least one sentence included in the comparison target document.

According to an embodiment of the present disclosure, the computing device 100 may acquire ( 720 ) a degree of technological similarity between at least a part of the first partial document and the comparison target document by using a technology similarity determination model.

According to an embodiment of the present disclosure, technology-related disputes are disputes between individuals, between individuals and organizations, or between organizations and organizations related to technology, such as intellectual property-related judgments, intellectual property-related civil lawsuits, intellectual property-related criminal proceedings, or intellectual property-related disputes. According to another embodiment, including at least one of administrative litigation related to property rights, a method for determining technology similarity using a neural network is disclosed.

According to an embodiment of the present disclosure, the dispute resolution basis document is a document that is helpful in the process of resolving a dispute by sending it to at least one of the other party to the dispute or the dispute resolution institution, and is a warning letter related to intellectual property rights, evidence of invalidity of intellectual property rights It may include at least one of documents, intellectual property cancellation judgment evidence documents, or intellectual property related litigation evidence documents.

8 is a block diagram illustrating a module for determining technology similarity according to an embodiment of the present disclosure.

A method for determining technology similarity according to an embodiment of the present disclosure may be implemented by the following modules. A module 810 for inputting at least a part of the first partial document or at least one of the comparison target documents into a technical similarity judgment model trained to determine the degree of similarity between the first partial document and the second partial document or at least one of the comparison target documents (810). ); and a module 820 for obtaining a technology similarity between at least a part of the first partial document and a comparison target document by using a technology similarity determination model.

In an alternative embodiment for determining technology similarity, a module for performing technology monitoring; and a module for determining a document based on dispute resolution for resolving technology-related disputes.

In an alternative embodiment for determining technology similarity, the module for performing technology monitoring may include a module for determining a document to be compared as an infringing document based on a predetermined criterion for determining infringing rights.

In an alternative embodiment for determining technology similarity, the module for performing technology monitoring transmits information related to the determined infringing document to a terminal of the right holder having a right to the document of the first part or to the holder of the right to the document of the first part. It may include a module for transmitting to at least one of the agent terminals.

According to an embodiment of the present disclosure, a module for determining technology similarity may be implemented by a means, circuit, or logic for implementing a computing device. Those skilled in the art will further understand that the various illustrative logical blocks, components, modules, circuits, means, logics, and algorithm steps described in connection with the embodiments disclosed herein may be implemented in electronic hardware, computer software, or combinations of both. It should be recognized that it can be implemented as To clearly illustrate the interchangeability of hardware and software, various illustrative components, blocks, configurations, means, logics, modules, circuits, and steps have been described above generally in terms of their functionality. Whether such functionality is implemented in hardware or as software depends on the particular application and design constraints imposed on the overall system. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present disclosure.

9 is a simplified and general schematic diagram of an exemplary computing environment in which embodiments of the present disclosure may be implemented.

Although the present disclosure has been described above as being generally embodied by a computing device, those skilled in the art will understand that the present disclosure may be combined with computer-executable instructions and/or other program modules that may be executed on one or more computers and/or may be implemented in hardware and software. It will be appreciated that it can be implemented as a combination.

Generally, program modules include routines, programs, components, data structures, etc. that perform particular tasks or implement particular abstract data types. It will also be appreciated by those skilled in the art that the methods of the present disclosure may be used in single-processor or multiprocessor computer systems, minicomputers, mainframe computers, as well as personal computers, handheld computing devices, microprocessor-based or programmable consumer electronics, and the like (each of which is It will be appreciated that other computer system configurations may be implemented, including those that may be operative in connection with one or more associated devices.

The described embodiments of the present disclosure may also be practiced in distributed computing environments where certain tasks are performed by remote processing devices that are linked through a communications network. In a distributed computing environment, program modules may be located in both local and remote memory storage devices.

Computers typically include a variety of computer readable media. Computer readable media can be any medium that can be accessed by a computer, including volatile and nonvolatile media, transitory and non-transitory media, removable and non-transitory media. Includes removable media. By way of example, and not limitation, computer readable media may include computer readable storage media and computer readable transmission media. Computer readable storage media are volatile and nonvolatile media, transitory and non-transitory, removable and non-removable media implemented in any method or technology for storage of information such as computer readable instructions, data structures, program modules or other data. includes media Computer readable storage media may include RAM, ROM, EEPROM, flash memory or other memory technology, CD-ROM, digital video disk (DVD) or other optical disk storage device, magnetic cassette, magnetic tape, magnetic disk storage device or other magnetic storage device. device, or any other medium that can be accessed by a computer and used to store desired information.

A computer readable transmission medium typically embodies computer readable instructions, data structures, program modules or other data in a modulated data signal such as a carrier wave or other transport mechanism. Including all information delivery media. The term modulated data signal means a signal that has one or more of its characteristics set or changed so as to encode information within the signal. By way of example, and not limitation, computer readable transmission media includes wired media such as a wired network or direct-wired connection, and wireless media such as acoustic, RF, infrared, and other wireless media. Combinations of any of the above are also intended to be included within the scope of computer readable transmission media.

An exemplary environment 1100 implementing various aspects of the present disclosure is shown including a computer 1102, which includes a processing unit 1104, a system memory 1106, and a system bus 1108. do. System bus 1108 couples system components, including but not limited to system memory 1106 , to processing unit 1104 . Processing unit 1104 may be any of a variety of commercially available processors. Dual processor and other multiprocessor architectures may also be used as the processing unit 1104.

System bus 1108 may be any of several types of bus structures that may additionally be interconnected to a memory bus, a peripheral bus, and a local bus using any of a variety of commercial bus architectures. System memory 1106 includes read only memory (ROM) 1110 and random access memory (RAM) 1112 . A basic input/output system (BIOS) is stored in non-volatile memory 1110, such as ROM, EPROM, or EEPROM, and is a basic set of information that helps transfer information between components within computer 1102, such as during startup. contains routines. RAM 1112 may also include high-speed RAM, such as static RAM, for caching data.

The computer 1102 may also include an internal hard disk drive (HDD) 1114 (eg, EIDE, SATA) - the internal hard disk drive 1114 may also be configured for external use within a suitable chassis (not shown). Yes—a magnetic floppy disk drive (FDD) 1116 (e.g., for reading from or writing to a removable diskette 1118), and an optical disk drive 1120 (e.g., a CD-ROM) for reading disc 1122 or reading from or writing to other high capacity optical media such as DVDs). The hard disk drive 1114, magnetic disk drive 1116, and optical disk drive 1120 are connected to the system bus 1108 by a hard disk drive interface 1124, magnetic disk drive interface 1126, and optical drive interface 1128, respectively. ) can be connected to The interface 1124 for external drive implementation includes at least one or both of USB (Universal Serial Bus) and IEEE 1394 interface technologies.

These drives and their associated computer readable media provide non-volatile storage of data, data structures, computer executable instructions, and the like. In the case of computer 1102, drives and media correspond to storing any data in a suitable digital format. Although the description of computer readable media above refers to HDDs, removable magnetic disks, and removable optical media such as CDs or DVDs, those skilled in the art can use zip drives, magnetic cassettes, flash memory cards, cartridges, etc. It will be appreciated that other tangible media readable by the computer, such as the like, may also be used in the exemplary operating environment and any such media may include computer executable instructions for performing the methods of the present disclosure.

A number of program modules may be stored on the drive and RAM 1112, including an operating system 1130, one or more application programs 1132, other program modules 1134, and program data 1136. All or portions of the operating system, applications, modules and/or data may also be cached in RAM 1112. It will be appreciated that the present disclosure may be implemented in a variety of commercially available operating systems or combinations of operating systems.

A user may enter commands and information into the computer 1102 through one or more wired/wireless input devices, such as a keyboard 1138 and a pointing device such as a mouse 1140. Other input devices (not shown) may include a microphone, IR remote control, joystick, game pad, stylus pen, touch screen, and the like. Although these and other input devices are often connected to the processing unit 1104 through an input device interface 1142 that is connected to the system bus 1108, a parallel port, IEEE 1394 serial port, game port, USB port, IR interface, may be connected by other interfaces such as the like.

A monitor 1144 or other type of display device is also connected to the system bus 1108 through an interface such as a video adapter 1146. In addition to the monitor 1144, computers typically include other peripheral output devices (not shown) such as speakers, printers, and the like.

Computer 1102 may operate in a networked environment using logical connections to one or more remote computers, such as remote computer(s) 1148 via wired and/or wireless communications. Remote computer(s) 1148 may be a workstation, computing device computer, router, personal computer, handheld computer, microprocessor-based entertainment device, peer device, or other common network node, and generally includes It includes many or all of the components described for, but for simplicity, only memory storage device 1150 is shown. The logical connections shown include wired/wireless connections to a local area network (LAN) 1152 and/or a larger network, such as a wide area network (WAN) 1154 . Such LAN and WAN networking environments are common in offices and corporations and facilitate enterprise-wide computer networks, such as intranets, all of which can be connected to worldwide computer networks, such as the Internet.

When used in a LAN networking environment, computer 1102 connects to local network 1152 through wired and/or wireless communication network interfaces or adapters 1156. Adapter 1156 may facilitate wired or wireless communications to LAN 1152, which also includes a wireless access point installed therein to communicate with wireless adapter 1156. When used in a WAN networking environment, computer 1102 may include a modem 1158, be connected to a communicating computing device on WAN 1154, or establish communications over WAN 1154, such as over the Internet. have other means. A modem 1158, which may be internal or external and a wired or wireless device, is connected to the system bus 1108 through a serial port interface 1142. In a networked environment, program modules described for computer 1102, or portions thereof, may be stored on remote memory/storage device 1150. It will be appreciated that the network connections shown are exemplary and other means of establishing a communication link between computers may be used.

Computer 1102 is any wireless device or entity that is deployed and operating in wireless communication, eg, printers, scanners, desktop and/or portable computers, portable data assistants (PDAs), communication satellites, wireless detectable tags associated with It operates to communicate with arbitrary equipment or places and telephones. This includes at least Wi-Fi and Bluetooth wireless technologies. Thus, the communication may be a predefined structure as in conventional networks or simply an ad hoc communication between at least two devices.

Wi-Fi (Wireless Fidelity) makes it possible to connect to the Internet without wires. Wi-Fi is a wireless technology, such as a cell phone, that allows such devices, eg, computers, to transmit and receive data both indoors and outdoors, i.e. anywhere within coverage of a base station. Wi-Fi networks use a radio technology called IEEE 802.11 (a, b, g, etc.) to provide secure, reliable, and high-speed wireless connections. Wi-Fi can be used to connect computers to each other, to the Internet, and to wired networks (using IEEE 802.3 or Ethernet). Wi-Fi networks can operate in the unlicensed 2.4 and 5 GHz radio bands, for example, at 11 Mbps (802.11a) or 54 Mbps (802.11b) data rates, or in products that include both bands (dual band) .

Those skilled in the art will understand that information and signals may be represented using any of a variety of different technologies and techniques. For example, data, instructions, instructions, information, signals, bits, symbols and chips that may be referenced in the above description are voltages, currents, electromagnetic waves, magnetic fields or particles, optical fields s or particles, or any combination thereof.

Those skilled in the art will understand that the various illustrative logical blocks, modules, processors, means, circuits, and algorithm steps described in connection with the embodiments disclosed herein are electronic hardware, (for convenience) , may be implemented by various forms of program or design code (referred to herein as software) or a combination of both. To clearly illustrate this interchangeability of hardware and software, various illustrative components, blocks, modules, circuits, and steps have been described above generally in terms of their functionality. Whether such functionality is implemented as hardware or software depends on the particular application and the design constraints imposed on the overall system. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present disclosure.

Various embodiments presented herein may be implemented as a method, apparatus, or article of manufacture using standard programming and/or engineering techniques. The term article of manufacture includes a computer program, carrier, or media accessible from any computer-readable storage device. For example, computer-readable storage media include magnetic storage devices (eg, hard disks, floppy disks, magnetic strips, etc.), optical disks (eg, CDs, DVDs, etc.), smart cards, and flash memory devices (eg, EEPROM, cards, sticks, key drives, etc.), but are not limited thereto. Additionally, various storage media presented herein include one or more devices and/or other machine-readable media for storing information.

It is to be understood that the specific order or hierarchy of steps in the processes presented is an example of exemplary approaches. Based upon design priorities, it is to be understood that the specific order or hierarchy of steps in the processes may be rearranged within the scope of this disclosure. The accompanying method claims present elements of the various steps in a sample order, but are not meant to be limited to the specific order or hierarchy presented.

The description of the presented embodiments is provided to enable any person skilled in the art to make or use the present disclosure. Various modifications to these embodiments will be apparent to those skilled in the art of this disclosure, and the general principles defined herein may be applied to other embodiments without departing from the scope of this disclosure. Thus, the present disclosure is not to be limited to the embodiments presented herein, but is to be interpreted in the widest scope consistent with the principles and novel features presented herein.

Claims (19)

A computer program stored on a computer-readable storage medium, which, when executed in one or more processors, performs the following operations for determining technology similarity using a neural network, the operations comprising:
inputting at least a part of the first partial document or at least one of the comparison target documents into a technology similarity judgment model learned to determine a degree of similarity between the first partial document and the second partial document or at least one of the comparison target documents; and
obtaining a technology similarity between at least a part of the first partial document and the comparison target document by using the technology similarity determination model;
including,
A computer program stored on a computer-readable storage medium.
According to claim 1,
The document includes at least one of a first part document for claiming the right or a second part document for explaining the content of the right claim.
A computer program stored on a computer-readable storage medium.
According to claim 1,
The technology similarity judgment model,
A first model configured to learn a correlation between contents of the first partial document and contents of the second partial document by inputting the first partial document and the second partial document to a technology similarity determination model,
A computer program stored on a computer-readable storage medium.
According to claim 3,
The technology similarity judgment model
And a second model in which the first partial document and the comparison target document are input to the first model to learn a correlation between the contents of the first partial document and the comparison target document.
A computer program stored on a computer-readable storage medium.
According to claim 1,
The documents to be compared are
Including at least one of a first comparison target document that is a raw comparison target document or a second comparison target document that is a document processed from a raw comparison target document,
A computer program stored on a computer-readable storage medium.
According to claim 5,
The first comparison target document,
As a document containing technical content,
containing at least one of intellectual property-related documents, monographs or technical documents;
A computer program stored on a computer-readable storage medium.
According to claim 5,
The second comparison target document,
It includes at least some of the plurality of sentences included in the first comparison target document,
At least some of the plurality of sentences included in the first comparison target document include sentences determined based on a key sentence calculation algorithm,
A computer program stored on a computer-readable storage medium.
According to claim 7,
The key sentence calculation algorithm,
calculating a technical term importance based on a technical term appearance frequency for each technical term in a sentence included in the first comparison target document;
obtaining an inter-sentence weight matrix by calculating a first matrix having each of the calculated technical term importances as a matrix element and a transposed matrix of the first matrix;
calculating a rank of a sentence included in the first comparison target document based on the inter-sentence weight matrix; and
determining a sentence whose calculated rank is equal to or greater than a predetermined rank as a core sentence;
including,
A computer program stored on a computer-readable storage medium.
According to claim 1,
The documents to be compared are
Including documents updated at predetermined intervals based on a specific point in time in order to determine the degree of technological similarity with respect to documents to be compared that are released over time,
A computer program stored on a computer-readable storage medium.
According to claim 1,
The technology similarity judgment model,
a first sub-model for extracting features of a first partial document;
a second sub-model for extracting a feature of a second partial document or a feature of the comparison target document;
a third sub-model for deriving a correlation between the first partial document and the comparison target document based on the output of the first sub-model and the output of the second sub-model; and
a fourth sub-model for obtaining the degree of similarity based on the correlation;
including,
A computer program stored on a computer-readable storage medium.
According to claim 1,
The degree of similarity is
a score determined based on the number of at least one of words or sentences included in the comparison target document having a correlation with at least one of the words or sentences included in the first partial document equal to or greater than a predetermined criterion;
or
a score determined based on a correlation score between at least one word or sentence included in the first partial document and at least one word or sentence included in the comparison target document;
including at least one of
A computer program stored on a computer-readable storage medium.
According to claim 1,
the operation of performing technical monitoring; and
Determining dispute resolution basis documents for technology-related dispute resolution;
Including more,
A computer program stored on a computer-readable storage medium.
According to claim 12,
The operation of performing the technical monitoring,
determining a comparison target document as a rights infringing document based on a predetermined right infringement criterion;
including,
A computer program stored on a computer-readable storage medium.
According to claim 12,
The operation of performing the technical monitoring,
transmitting information related to the determined infringing document to at least one of a terminal of a right holder having a right to the first partial document or an agent terminal of the right holder to the first partial document;
including,
A computer program stored on a computer-readable storage medium.
According to claim 12,
The above technology-related disputes,
Disputes between individuals, between individuals and organizations, or between organizations and organizations related to technology;
including at least one of intellectual property-related judgments, intellectual property-related civil litigation, intellectual property-related criminal litigation, or intellectual property-related administrative litigation;
A computer program stored on a computer-readable storage medium.
According to claim 12,
The dispute resolution basis document,
A document to assist in the process of resolving a dispute by sending it to at least one of the parties to the dispute or the dispute resolution body;
Including at least one of intellectual property-related warning letters, intellectual property invalidity judgment evidence documents, intellectual property cancellation judgment evidence documents, or intellectual property-related litigation evidence documents,
A computer program stored on a computer-readable storage medium.
A method for determining technology similarity using a neural network,
inputting at least a part of the first partial document or at least one of the comparison target documents into a technology similarity judgment model learned to determine a degree of similarity between the first partial document and the second partial document or at least one of the comparison target documents; and
obtaining a technology similarity between at least a part of the first partial document and the comparison target document by using the technology similarity determination model;
including,
A method for determining technology similarity using a neural network.
A computing device for determining technology similarity using a neural network,
one or more processors; and
a memory storing instructions executable by the one or more processors;
including,
The one or more processors,
inputting at least a part of the first partial document or at least one of the comparison target documents into a technology similarity judgment model learned to determine a degree of similarity between the first partial document and the second partial document or at least one of the comparison target documents; and
Obtaining a technology similarity between at least a part of the first partial document and the comparison target document by using the technology similarity determination model;
A computing device for determining technology similarity using a neural network.
A computer-readable recording medium storing a data structure corresponding to parameters of a neural network at least partially updated in a learning process, wherein the operation of the neural network is based at least in part on the parameters, and the learning process comprises:
learning a correlation between contents of the first partial document and contents of the second partial document; and
learning a correlation between the contents of a first partial document and the contents of the comparison target document;
including,
A computer readable recording medium in which a data structure is stored.

Images