Method, system and non-transitory computer-readable recording medium for providing artificial intelligence-based transmission and reception service

Abstract

According to one aspect of the present invention, provided is a method for providing an artificial intelligence-based transmission and reception service, the method comprising the steps of: converting first data including at least one image received from a transmission terminal into second data having a format different from that of the first data; and extracting characters included in the second data and location information of the characters included in the second data with reference to a result output by inputting the second data to an artificial intelligence-based character recognition model, wherein the artificial intelligence-based character recognition model is trained on the basis of third data having the same format as the second data, characters included in the third data, and location information of the characters included in the third data.

Description

Method, system and non-transitory computer-readable recording medium for providing artificial intelligence-based transmission and reception services

[0001] The present invention relates to a method, a system, and a non-transitory computer-readable recording medium for providing an artificial intelligence-based transmission and reception service.

[0002] As we enter the digital age, where massive amounts of text data are generated from diverse sources, the need for efficient information processing and management is increasing. Advances in natural language processing and artificial intelligence technologies are further increasing the demand for summarization technologies. In particular, summarization technologies utilizing deep learning-based natural language processing are significantly helping users quickly access key information from large volumes and effectively consume it.

[0003] Organizations such as companies and institutions must handle a large volume of documents and postings related to the organization's work. Therefore, it is necessary to effectively search for necessary information in a large volume of documents and summarize the documents so that the necessary content can be quickly obtained from the searched documents.

[0004] In particular, call centers and general users using electronic fax machines often need to review and verify the contents of received fax data (e.g., TIF images). For example, insurance companies and banks face the hassle of manually reviewing multiple received faxes, identifying their contents, and then inputting them into robotic process automation (RPA) systems.

[0005] <Prior Art Literature>

[0006] Patent Document

[0007] (Patent Document 1) Patent Publication No. 10-2242590 (April 14, 2021)

[0008] The purpose of the present invention is to solve all of the problems of the above-mentioned prior art.

[0009] In addition, another purpose of the present invention is to provide text recognized through fax data analysis as summary information so that a recipient can quickly understand the contents of received fax data through the summary information without viewing the received fax data.

[0010] In addition, another object of the present invention is to enable an artificial intelligence-based fax data summary service to be provided in a public switched telephone network (PSTN).

[0011] In addition, another object of the present invention is to minimize loss of work time and visually confirmed data input errors in processing received fax data by providing customized summary information with reference to information about a receiving terminal that receives fax data.

[0012] A representative configuration of the present invention to achieve the above purpose is as follows.

[0013] According to one aspect of the present invention, a method for providing an artificial intelligence-based transmission and reception service is provided, comprising the steps of: converting first data including at least one image received from a sending terminal into second data having a format different from that of the first data; and extracting characters included in the second data and positional information of the characters included in the second data by referring to a result output by inputting the second data into an artificial intelligence-based character reading model, wherein the artificial intelligence-based character reading model is trained based on third data having the same format as the second data, characters included in the third data, and positional information of the characters included in the third data.

[0014] According to another aspect of the present invention, a system for providing an artificial intelligence-based transmission and reception service is provided, comprising: a data conversion unit for converting first data including at least one image received from a sending terminal into second data having a different format from the format of the first data; and an extraction unit for extracting characters included in the second data and positional information of the characters included in the second data by referring to a result output by inputting the second data into an artificial intelligence-based character reading model, wherein the artificial intelligence-based character reading model is learned based on third data having the same format as the second data, characters included in the third data, and positional information of the characters included in the third data.

[0015] In addition, a non-transitory computer-readable recording medium on which another method for implementing the present invention, another system, and a computer program for executing the method are recorded is further provided.

[0016] According to the present invention, text recognized through fax data analysis is provided as summary information, so that the recipient can quickly understand the contents of the received fax data through the summary information without viewing the received fax data.

[0017] In addition, according to the present invention, an artificial intelligence-based fax data summary service can be provided even in a public switched telephone network (PSTN).

[0018] In addition, according to the present invention, by providing customized summary information with reference to information about a receiving terminal that receives fax data, it is possible to minimize loss of work time for processing received fax data and data input errors confirmed with the naked eye.

[0019] FIG. 1 is a diagram schematically illustrating the configuration of an entire system for providing an artificial intelligence-based transmission and reception service according to one embodiment of the present invention.

[0020] FIG. 2 is a drawing detailing the internal configuration of a service support system according to one embodiment of the present invention.

[0021] FIG. 3 is a diagram exemplifying a method for providing an artificial intelligence-based transmission and reception service according to one embodiment of the present invention.

[0022] FIG. 4 is a diagram exemplarily showing a case where a summary page is inserted into the first page of fax data in an artificial intelligence-based transmission and reception service according to one embodiment of the present invention.

[0023] FIG. 5 is a diagram exemplarily showing a transmission process and a reception process for providing an artificial intelligence-based transmission and reception service according to one embodiment of the present invention.

[0024] FIG. 6 is a drawing exemplarily showing a character recognition method using an artificial intelligence-based character reading model according to one embodiment of the present invention.

[0025] The following detailed description of the present invention refers to the accompanying drawings, which illustrate specific embodiments in which the present invention may be practiced. These embodiments are described in sufficient detail to enable those skilled in the art to practice the present invention. It should be understood that the various embodiments of the present invention, while different from each other, are not necessarily mutually exclusive. For example, specific shapes, structures, and characteristics described herein may be modified and implemented from one embodiment to another without departing from the spirit and scope of the present invention. Furthermore, it should be understood that the positions or arrangements of individual components within each embodiment may also be modified without departing from the spirit and scope of the present invention. Accordingly, the following detailed description is not to be taken in a limiting sense, and the scope of the present invention is to be construed to encompass the scope of the claims and all equivalents thereof. Like reference numerals in the drawings represent the same or similar elements throughout the several aspects.

[0026] Hereinafter, various preferred embodiments of the present invention will be described in detail with reference to the attached drawings so that a person having ordinary skill in the art to which the present invention pertains can easily practice the present invention.

[0027] Composition of the entire system

[0028] FIG. 1 is a diagram schematically illustrating the configuration of an entire system for providing an artificial intelligence-based transmission and reception service according to one embodiment of the present invention.

[0029] As illustrated in FIG. 1, the entire system according to one embodiment of the present invention may include a communication network (100), a service support system (200), a sending terminal (300), and a receiving terminal (400).

[0030] First, the communication network (100) according to one embodiment of the present invention can be configured regardless of the communication mode, such as wired communication or wireless communication, and can be configured with various communication networks such as a local area network (LAN), a metropolitan area network (MAN), and a wide area network (WAN). Preferably, the communication network (100) referred to herein may be the well-known Internet or the World Wide Web (WWW). However, the communication network (100) is not necessarily limited thereto and may include at least a portion of a well-known wired or wireless data communication network, a well-known telephone network (e.g., a public switched telephone network (PSTN)), or a well-known wired or wireless television communication network.

[0031] For example, the communication network (100) may be a wireless data communication network that implements conventional communication methods such as WiFi communication, WiFi-Direct communication, Long Term Evolution (LTE) communication, 5G communication, Bluetooth communication (including Bluetooth Low Energy (BLE) communication), infrared communication, ultrasonic communication, etc., at least in part.

[0032] Next, the service support system (200) according to one embodiment of the present invention can communicate with a sending terminal (300) and a receiving terminal (400) through a communication network (100), convert first data including at least one image received from the sending terminal (300) into second data having a different format from the format of the first data, input the second data into an artificial intelligence-based character reading model, and extract characters included in the second data and location information of the characters included in the second data with reference to the output result. Here, the artificial intelligence-based character reading model can be trained based on third data having the same format as the second data, characters included in the third data, and location information of the characters included in the third data. This service support system (200) may be a system that is driven by a server equipped with a memory means and a microprocessor and having computational capabilities.

[0033] The configuration and function of the service support system (200) according to the present invention will be described in detail below.

[0034] Next, the sending terminal (300) and the receiving terminal (400) according to one embodiment of the present invention are digital devices that include a function for communicating after connecting to the service support system (200), and any digital device that has a memory means, a microprocessor, and a computing capability, such as a facsimile machine, a smart phone, a tablet, a smart watch, a smart band, smart glasses, a desktop computer, a notebook computer, a workstation, a PDA, a web pad, a mobile phone, etc., can be adopted as the sending terminal (300) and the receiving terminal (400) according to the present invention. Here, the sending terminal (300) and the receiving terminal (400) according to one embodiment of the present invention may include a communication means for transmitting and receiving fax data, an input means (e.g., a keyboard) for inputting recipient information when transmitting fax data or confirming received fax data, a display means (e.g., an LCD, an LED) for displaying the progress status of fax data transmission and reception or displaying received fax data, etc.

[0035] Meanwhile, the sending terminal (300) and the receiving terminal (400) may further include application programs for performing functions according to the present invention. These applications may exist in the form of program modules within the sending terminal (300) and the receiving terminal (400). The nature of these program modules may be generally similar to the components of the service support system (200) described below (i.e., the data conversion unit (210), the extraction unit (220), the summary page insertion unit (230), the communication unit (240), and the control unit (250)). Here, at least a portion of the application may be replaced with a hardware device or firmware device that can perform functions substantially identical to or equivalent thereto, as needed.

[0036] Composition of service support system

[0037] Below, the internal configuration and functions of each component of the service support system (200) that performs important functions for implementing the present invention will be examined.

[0038] FIG. 2 is a drawing showing in detail the internal configuration of a service support system (200) according to one embodiment of the present invention.

[0039] As illustrated in FIG. 2, according to one embodiment of the present invention, the service support system (200) may include a data conversion unit (210), an extraction unit (220), a summary page insertion unit (230), a communication unit (240), and a control unit (250). According to one embodiment of the present invention, at least some of the data conversion unit (210), the extraction unit (220), the summary page insertion unit (230), the communication unit (240), and the control unit (250) of the service support system (200) may be program modules that communicate with an external system (not shown). These program modules may be included in the service support system (200) in the form of an operating system, an application program module, and other program modules, and may be physically stored in various known memory devices. In addition, these program modules may also be stored in a remote memory device that can communicate with the service support system (200). Meanwhile, these program modules include, but are not limited to, routines, subroutines, programs, objects, components, data structures, etc. that perform specific tasks or execute specific abstract data types, as described below, according to the present invention.

[0040] Meanwhile, although the service support system (200) has been described as above, this description is exemplary, and it is obvious to those skilled in the art that at least some of the components or functions of the service support system (200) may be realized within at least one of the sending terminal (300) and the receiving terminal (400) or within a server (not shown) as needed, or may be included within an external system (not shown).

[0041] First, the data conversion unit (210) according to one embodiment of the present invention can perform a function of converting first data including at least one image received from a sending terminal (300) into second data in a format different from the format of the first data.

[0042] Specifically, the data conversion unit (210) can convert first data including at least one image having a first file format (e.g., '.tif' file format) into second data having a second file format (e.g., '.jpg' or '.png' file format) for each image. Here, the reason for converting the file format of the data may be to quickly process and generate summary information about the first data for each image.

[0043] Meanwhile, it should be noted that the fax data conversion according to the present invention is not necessarily limited to the file formats listed above and may be modified in various ways. For example, first data including at least one image having a first file format (e.g., '.tif' file format) may be converted into second data having a second file format (e.g., at least one of '.gif', '.bmp', '.webp', and '.svg' file formats) for each image.

[0044] Next, the extraction unit (220) according to one embodiment of the present invention may input the second data into an artificial intelligence-based character reading model and extract characters included in the second data and location information of the characters included in the second data by referring to the output result. Here, the artificial intelligence-based character reading model may be trained based on third data in the same format as the second data, characters included in the third data, and location information of the characters included in the third data.

[0045] For example, the artificial intelligence-based character reading model may be a deep neural network (DNN) or transformer-based character recognition model, and the extraction unit (220) may input second data into the model and extract characters and their relative (or absolute) position information within the page by referring to the output result.

[0046] In addition, the extraction unit (220) may input the second data into an artificial intelligence-based character reading model and extract characters included in the second data, positional information of characters included in the second data, and style information of characters included in the second data by referring to the output result. Here, the artificial intelligence-based character reading model may be trained based on third data in the same format as the second data, characters included in the third data, positional information of characters included in the third data, and style information of characters included in the third data.

[0047] For example, the artificial intelligence-based character reading model may be a deep neural network (DNN) or transformer-based character recognition model, and the extraction unit (220) may input second data into the model and extract characters, information on the relative (or absolute) position of the characters within the page, and information on the width and height of the characters by referring to the output result.

[0048] Meanwhile, the artificial intelligence-based character reading model according to the present invention is not necessarily limited to the models listed above, and may be modified in various ways within the scope of achieving the purpose of the present invention.

[0049] Meanwhile, if the extraction unit (220) does not utilize the above model, the second data may be analyzed based on optical character recognition (OCR) technology using connected component analysis (CCA), etc., thereby extracting characters included in the second data. In addition, the extraction unit (220) may specify a virtual coordinate system on the second data (specifically, each image or page) and extract position information and style information of characters (specifically, recognized characters) based on the specified virtual coordinate system. For example, the position information of the character may include information on the X-coordinate and Y-coordinate at which the character is located based on the virtual coordinate system specified on the second data (for example, a virtual two-dimensional coordinate system specified with the lower left corner of the page as the origin). Here, the style information of the character may include information on the width and height of the character (furthermore, information on the relative size of the character, the relative thickness, whether to apply underlining or italics, etc.).

[0050] Next, the summary page insertion unit (230) according to one embodiment of the present invention may first perform a hash operation on the first data to generate an identifier for identifying the first data. For example, the hash operation may preferably utilize at least one of the MD5, SHA-1, and SHA-256 hash algorithms.

[0051] In addition, the summary page insertion unit (230) can input characters included in the second data and position information (further, style information of the characters) of the characters included in the second data into an artificial intelligence-based summary model in response to an identifier for identifying the first data, and can generate summary information about the first data based on the output result. Here, the artificial intelligence-based summary model can be trained using characters included in the fourth data of the same format as the second data (e.g., a file of the same format, document properties (e.g., number of characters or paragraphs, nature of the text, etc.), document template, etc.), position information of the characters included in the fourth data (e.g., position information of the characters may be utilized to determine the relationship or distance between characters during training), and summary information determined in response to the characters included in the fourth data and the position information of the characters included in the fourth data as training data. For example, when applying a generative artificial intelligence-based summary model, a model (e.g., a large-scale language model) whose general patterns are pre-trained in a large-scale dataset can be trained (e.g., fine-tuned) using the characters included in the fourth data, the location information of the characters included in the fourth data, and the summary information determined in response to the characters included in the fourth data and the location information of the characters included in the fourth data as learning data.

[0052] In addition, the summary page insertion unit (230) can input characters included in the second data, position information of characters included in the second data, and style information of characters included in the second data into an artificial intelligence-based summary model in response to an identifier for identifying the first data, and can generate summary information about the first data based on the output result. Here, the artificial intelligence-based summary model can be trained using characters included in fourth data of the same format as the second data, position information of characters included in the fourth data, style information of characters included in the fourth data (for example, style information can be used to determine the content importance of characters, such as recognizing them as important and determining them as a summary target or theme when the size of characters is relatively large or underlined, italicized, etc. is applied), and summary information determined in response to characters included in the fourth data, position information of characters included in the fourth data, and style information of characters included in the fourth data as training data. For example, when applying a generative artificial intelligence-based summary model, a model (e.g., a large-scale language model) whose general patterns are pre-trained in a large-scale dataset can be trained (e.g., fine-tuned) using the characters included in the fourth data, the positional information of the characters included in the fourth data, the style information of the characters included in the fourth data, and the summary information determined in response to the characters included in the fourth data, the positional information of the characters included in the fourth data, and the style information of the characters included in the fourth data as learning data.

[0053] For example, it may be desirable for an artificial intelligence-based summary model to use at least one of a deep learning (e.g., transformer)-based language model, specifically a large language model (LLM) and a KoBertSum model, particularly targeting the Korean language.

[0054] Meanwhile, the method of generating summary information on the first data according to the present invention is not necessarily limited to the above natural language processing model, and various analysis methods such as TF-IDF (Term Frequency-Inverse Document Frequency) may be utilized within the scope that can achieve the purpose of the present invention.

[0055] Additionally, the summary page insertion unit (230) can generate customized summary information by referencing information about a receiving terminal associated with the first data (e.g., the information can be obtained by analyzing the first data).

[0056] Specifically, the summary page insertion unit (230) can generate customized summary information about the first data by referencing information about a receiving terminal that includes at least one of information about a fax data reception number of a specific institution (e.g., a specific bank or insurance company) that is the recipient of the first data and information about a unique number of the receiving terminal device.

[0057] For example, the summary page insertion unit (230) may generate summary information composed of 'customer name information', 'contract number information', and 'contact information' as customized summary information regarding the first data by referencing at least one of the fax data reception number information and the receiving terminal device unique number information of 'Insurance Company A', which is the recipient of the first data. In addition, in order to generate the customized summary information as described above, the summary page insertion unit (230) may refer to a database or a lookup table regarding the configuration (or arrangement) of the summary information corresponding to the fax data reception number information or the receiving terminal device unique number information.

[0058] In addition, the summary page insertion unit (230) can learn an artificial intelligence-based summary composition model for each recipient, and use each learned summary composition model to generate customized summary information corresponding to the fax data reception number information or the unique number information of the receiving terminal device. Here, the artificial intelligence-based summary composition model can be configured for each recipient, and can be generated by learning (e.g., fine-tuning) the previously discussed pre-learned summary model (e.g., a model in which common patterns are pre-learned in a large-scale dataset) based on the summary information of each recipient.

[0059] Meanwhile, the summary page insertion unit (230) can provide customized summary information to a robotic process automation (RPA) system using artificial intelligence technology, thereby minimizing waste of human resources, reduction of work processing time, minimizing human mistakes and errors, and improving work quality.

[0060] Next, the summary page insertion unit (230) according to one embodiment of the present invention can insert a summary page generated based on summary information into the first data by referring to an identifier for identifying the first data.

[0061]

[0062] *For example, the summary page insertion unit (230) may determine target first data into which a summary page generated based on summary information among a plurality of first data is to be inserted by referring to an identifier for identifying the first data, and may cause the summary page to be inserted on the first page (specifically, the page before the first page) of the target first data including a plurality of pages (e.g., the first to third pages).

[0063] Next, the communication unit (240) according to one embodiment of the present invention can perform a function that enables data transmission and reception from / to the data conversion unit (210), the extraction unit (220), and the summary page insertion unit (230).

[0064] Finally, the control unit (250) according to one embodiment of the present invention can perform a function of controlling the flow of data between the data conversion unit (210), the extraction unit (220), the summary page insertion unit (230), and the communication unit (240). That is, the control unit (250) according to the present invention can control the data conversion unit (210), the extraction unit (220), the summary page insertion unit (230), and the communication unit (240) to perform their own functions by controlling the flow of data from / to the outside of the service support system (200) or the flow of data between each component of the service support system (200).

[0065] FIG. 3 is a diagram exemplarily illustrating a method for providing an AI-based transmission and reception service according to one embodiment of the present invention. FIG. 5 is a diagram exemplarily illustrating a transmission process (FIG. 5(a)) and a reception process (FIG. 5(b)) for providing an AI-based transmission and reception service according to one embodiment of the present invention. FIG. 6 is a diagram exemplarily illustrating a character recognition method using an AI-based character reading model according to one embodiment of the present invention.

[0066] First, a service support system (200) according to one embodiment of the present invention may include a fax transmission / reception system that enables transmission / reception of first data (e.g., fax data) between a transmission terminal (300) and a reception terminal (400), and a system including an artificial intelligence-based character reading model that converts the first data into second data in a format different from the format of the first data and reads and recognizes characters included in the second data based on artificial intelligence.

[0067] Specifically, the fax transmission / reception system of the service support system (200) according to one embodiment of the present invention can receive (310) first data (e.g., fax data) including a plurality of images from a sending terminal (300).

[0068] Next, a system including an artificial intelligence-based character reading model of a service support system (200) according to one embodiment of the present invention can perform an image preprocessing task by converting (320) first data including a plurality of images in a '.tif' file format into second data in a '.jpg' or '.png' file format for each image.

[0069] Next, a system including an artificial intelligence-based character reading model of a service support system (200) can extract characters included in the second data by analyzing the second data based on optical character recognition (OCR) technology utilizing connected component analysis (CCA), etc., and can specify a virtual coordinate system on the second data, and can extract (330) positional information about the X-coordinate and Y-coordinate at which characters included in the second data are located based on the specified virtual coordinate system (for example, a virtual two-dimensional coordinate system specified with the lower left corner of the page as the origin).

[0070] Meanwhile, the system including the artificial intelligence-based character reading model of the service support system (200) may input the second data converted into a '.jpg' or '.png' file format into the artificial intelligence-based character reading model (this character reading model may be trained based on fifth data in the same format as the second data, characters included in the fifth data, location information of characters included in the fifth data, and pattern inspection results (or character patterns) of characters included in the fifth data (e.g., whether personal information is relevant, whether sensitive information is relevant, etc.)) and obtain characters included in the second data, character location information, and pattern inspection results by referring to the output results. Then, the system including the artificial intelligence-based character reading model of the service support system (200) may provide at least one service among a web fax transmission service, a robotic process automation service, a security solution service, and a search engine service (e.g., provided to a sending terminal (300) or a receiving terminal (400)) based on the characters recognized by the artificial intelligence-based character reading model, character location information, and pattern inspection results. Here, pattern inspection is an inspection that analyzes data included in, for example, a fax or scanned document to identify a specific pattern (for example, a personal information pattern such as a resident registration number, a secret pattern such as a security watermark, etc.), and may further include a process of masking or anonymizing the information to protect sensitive information. In particular, a system including an artificial intelligence-based character reading model of the service support system (200) may perform a function of automatically classifying received documents by type by recognizing the format of characters as well as character recognition during the character recognition process. For example, it may automatically classify them as '000 confirmation', '000 certificate', '000 application', etc. In addition, documents may be automatically classified by referring to the frequency of appearance of specific keywords included in the received document, or documents may be automatically classified by recipient by referring to the recipient.

[0071] Next, a system including an artificial intelligence-based character reading model of a service support system (200) can input characters included in second data and location information of characters included in second data corresponding to an identifier for identifying first data into an artificial intelligence-based summary model and generate summary information about the first data based on the output result (340).

[0072] Next, the system including the artificial intelligence-based character reading model of the service support system (200) can insert (350) a summary page generated based on summary information into the first data by referring to an identifier for identifying the first data. For example, referring to (a) and (b) of FIG. 4, the service support system (200) can insert a summary page (410) (specifically, a summary page (410) generated by arranging the summary information in a predetermined template (such template can be determined by referring to information about a receiving terminal associated with the first data)) into the first page of the first data including the first image to the third image (420, 430, 440) by referring to the identifier for identifying the first data. Accordingly, the user can quickly understand the contents of the received data (e.g., fax data) by checking only the generated summary page (410) without directly viewing the first to third images (420, 430, 440) of the first data.

[0073] The embodiments of the present invention described above may be implemented in the form of program commands that can be executed through various computer components and recorded on a computer-readable recording medium. The computer-readable recording medium may include program commands, data files, data structures, etc., either singly or in combination. The program commands recorded on the computer-readable recording medium may be specially designed and configured for the present invention or may be known and available to those skilled in the art of computer software. Examples of computer-readable recording media include magnetic media such as hard disks, floppy disks, and magnetic tapes, optical recording media such as CD-ROMs and DVDs, magneto-optical media such as floptical disks, and hardware devices specifically configured to store and execute program commands, such as ROMs, RAMs, and flash memories. Examples of program commands include not only machine language codes generated by a compiler, but also high-level language codes that can be executed by a computer using an interpreter, etc. Hardware devices may be changed into one or more software modules to perform processing according to the present invention, and vice versa.

[0074] Although the present invention has been described above with specific details such as specific components and limited examples and drawings, these are provided only to help a more general understanding of the present invention, and the present invention is not limited to the above examples, and those with ordinary knowledge in the technical field to which the present invention pertains can make various modifications and changes based on this description.

[0075] Therefore, the idea of ​​the present invention should not be limited to the embodiments described above, and not only the scope of the patent claims described below but also all scopes equivalent to or equivalently modified from the scope of the patent claims are considered to fall within the scope of the idea of ​​the present invention.

[0076] <Explanation of symbols>

[0077] 100: Communications network

[0078] 200: Service Support System

[0079] 210: Data Conversion Unit

[0080] 220: Extraction section

[0081] 230: Summary page insert

[0082] 240: Communications Department

[0083] 250: Control Unit

[0084] 300: Calling terminal

[0085] 400: Receiving terminal

Claims

1. A method for providing an artificial intelligence-based transmission and reception service, A step of converting first data including at least one image received from a sending terminal into second data in a format different from the format of the first data, and A step of inputting the second data into an artificial intelligence-based character reading model and extracting characters included in the second data and location information of the characters included in the second data by referring to the output result, The above artificial intelligence-based character reading model is learned based on third data of the same format as the second data, characters included in the third data, and location information of characters included in the third data. method.

2. In paragraph 1, The location information of the character included in the second data includes information about the X coordinate and Y coordinate at which the character included in the second data is located based on a virtual coordinate system specified in the second data. method.

3. In paragraph 2, In the above extraction step, further extract style information regarding the width and height of characters included in the second data. The above artificial intelligence-based character reading model is further learned based on the third data, characters included in the third data, location information of characters included in the third data, and style information of characters included in the third data. method.

4. In paragraph 3, A step of inputting characters included in the second data and location information of characters included in the second data corresponding to an identifier for identifying the first data into an artificial intelligence-based summary model and generating summary information about the first data based on the output result, and Further comprising a step of inserting a summary page generated based on the summary information into the first data by referring to the identifier, The artificial intelligence-based summary model is learned using as learning data the characters included in the fourth data of the same format as the second data, the location information of the characters included in the fourth data, and the summary information determined in response to the characters included in the fourth data and the location information of the characters included in the fourth data. method.

5. In paragraph 4, In the step of generating the above summary information, customized summary information is generated by referring to information about the receiving terminal associated with the first data. method.

6. In paragraph 5, The above identifier is generated by hashing the first data. method.

7. In paragraph 6, In the above generating step, the characters included in the second data, the location information of the characters included in the second data, and the style information of the characters included in the second data are input into an artificial intelligence-based summary model, and further summary information about the first data is generated based on the output result. The artificial intelligence-based summary model is further learned by using as learning data the summary information determined in response to the characters included in the fourth data, the position information of the characters included in the fourth data, the style information of the characters included in the fourth data, and the characters included in the fourth data, the position information of the characters included in the fourth data, and the style information of the characters included in the fourth data. method.

8. A non-transitory computer-readable recording medium recording a computer program for executing the method according to paragraph 1.

9. A system for providing artificial intelligence-based transmission and reception services. A data conversion unit that converts first data including at least one image received from a sending terminal into second data in a format different from the format of the first data, and An extraction unit is included that extracts characters included in the second data and location information of characters included in the second data by referring to the output result of inputting the second data into an artificial intelligence-based character reading model, The above artificial intelligence-based character reading model is learned based on third data of the same format as the second data, characters included in the third data, and location information of characters included in the third data. System.

Images