Artificial intelligence-based article generation apparatus and method

Abstract

This invention provides an article generation device and method that automatically generates draft articles using generative AI. [Solution] The article generation device 100 includes a memory 110 in which programs and data for artificial intelligence-based article generation are stored, and at least one processor, and operates by executing the program stored in the memory, and includes a control unit 120 that automatically generates an article draft based on a subject obtained from the user and an article format containing multiple generative AI regions, the control unit executes generative AI and creates an article draft by responding to questions directed to each of the generative AI regions based on AI command words contained in each of the subject and generative AI regions, and inserting the response results into each of the generative AI regions.

Description

Technical Field

[0001] The embodiments disclosed in the present specification relate to an artificial intelligence-based article generation apparatus and method, and more particularly, to an artificial intelligence-based article generation apparatus and method that automatically generates an article draft using a generative AI domain that automatically inserts question-and-answer results for generative AI. Background Art

[0002] Conventionally, most news articles have been produced manually by reporters. Therefore, article production requires a large amount of time, labor, and cost, and there is a problem in that particularly for news articles that require urgent reporting, the promptness of publication is reduced.

[0003] Therefore, in order to quickly produce news, technologies for automatically generating and providing news when the conditions for a specific news event are satisfied, such as the technology disclosed in Korean Registered Patent No. 10-0416953, have been introduced.

[0004] On the other hand, in recent years, many companies have promoted their enterprises and products through press materials, that is, articles. However, since enterprises and advertising agencies consume a large amount of time and cost on article writing, error checking, editing and other processes to produce such corporate advertising articles, the demand for technologies that enable easy article writing and editing is increasing.

[0005] On the other hand, the above-mentioned background art is the technical information that the inventor possessed for deriving the present invention or acquired in the process of deriving the present invention, and it is not necessarily a publicly known art disclosed to the general public before the filing of the present invention. Summary of Invention Problem to be Solved by Invention

[0006] The object of the embodiments disclosed in the present specification is to provide an article generation apparatus and method that automatically generates an article draft using generative AI.

[0007] Furthermore, the embodiments disclosed herein aim to present an article generation apparatus and method for completing article drafts using generative AI. [Means for solving the problem]

[0008] As a technical means for achieving the technical challenges described above, according to one embodiment, an article generation device is disclosed that includes a memory in which a program and data for generating articles based on artificial intelligence are stored, and a control unit which includes at least one processor and operates by executing the program stored in the memory, and automatically generates an article draft based on a subject obtained from a user and an article format which includes multiple generative AI regions, wherein the control unit executes generative AI, responds to questions directed to each of the generative AI regions based on AI command words contained in each of the subject and generative AI regions, and generates an article draft by inserting the response results into each of the generative AI regions.

[0009] According to another embodiment, an article generation method is disclosed which includes the steps of: providing an article format that includes a plurality of generative AI regions, executed by an artificial intelligence-based article generation device; and automatically generating an article draft based on a subject obtained from a user and the article format, wherein the step of generating the article draft includes executing a generative AI, responding to questions directed to each of the generative AI regions based on AI command words contained in the subject and each of the generative AI regions, and generating an article draft by inserting the response results into each of the generative AI regions.

[0010] Further embodiments disclose a computer-readable recording medium on which a program for executing an article generation method is stored, which is executed by an artificial intelligence-based article generation device, and the article generation method includes the steps of providing an article format including a plurality of generative AI regions, and automatically generating an article draft based on a subject obtained from a user and the article format, wherein the step of generating the article draft includes executing a generative AI, responding to questions directed to each of the generative AI regions based on AI command words contained in the subject and each of the generative AI regions, and generating an article draft by inserting the response results into each of the generative AI regions.

[0011] Further embodiments disclose a computer program, which is executed by an artificial intelligence-based article generation device and stored on a computer-readable recording medium for executing an article generation method, the article generation method comprising the steps of: providing an article format including a plurality of generative AI regions; and automatically generating an article draft based on a subject obtained from a user and the article format, the step of generating the article draft comprising executing a generative AI, responding to questions directed to each of the generative AI regions based on AI command words contained in the subject and each of the generative AI regions, and generating an article draft by inserting the response results into each of the generative AI regions. [Effects of the Invention]

[0012] According to any one of the aforementioned problem-solving methods, it is possible to provide an article generation device and method that automatically generates article drafts based on a standardized article format including AI command words and a generative AI.

[0013] Furthermore, using any one of the aforementioned problem-solving methods, it is possible to recommend appropriate keywords related to the subject in conjunction with an article management system (CMS) and conveniently attach reference data. In addition, by analyzing the attached reference data and referencing the analysis results based on a standardized article format and generative AI, it becomes possible to create articles more easily, reducing the time and cost of article creation.

[0014] Furthermore, by using any one of the aforementioned problem-solving methods to complete draft articles using generative AI, the time and cost involved in completing articles can be reduced, and the quality of articles can be easily improved.

[0015] The effects obtained by the disclosed embodiments are not limited to those mentioned above, and other effects not mentioned will be clearly understood by a person with ordinary skill in the art to which the disclosed embodiments belong, as described below. [Brief explanation of the drawing]

[0016] [Figure 1] This block diagram shows the configuration of an article generation device according to one embodiment. [Figure 2] This is a diagram illustrating the article generation process according to one embodiment. [Figure 3] This is a diagram illustrating the article generation process according to one embodiment. [Figure 4] This figure illustrates an example of article proofreading using generative AI. [Figure 5] This figure illustrates an example of article proofreading using generative AI. [Figure 6] This figure illustrates an example of article proofreading using generative AI. [Figure 7] This figure illustrates the generation of corrected text by resetting detailed options according to one embodiment. [Figure 8] This is a flowchart illustrating an article generation method based on one embodiment. DESCRIPTION OF EMBODIMENTS

[0017] Hereinafter, various embodiments will be described in detail with reference to the accompanying drawings. The embodiments described below may be modified and implemented in various different forms. In order to more clearly explain the features of the embodiments, detailed descriptions of matters widely known to those having ordinary knowledge in the technical field to which the following embodiments pertain are omitted. In addition, portions irrelevant to the description of the embodiments in the drawings are omitted, and like reference numerals are given to like portions throughout the specification.

[0018] Throughout the specification, when it is stated that a certain component is "coupled" to another component, this includes not only the case of being "directly coupled" but also the case of being "coupled with another component interposed therebetween". In addition, when it is stated that a certain configuration "includes" another configuration, unless specifically stated to the contrary, it means that other configurations may be further included rather than excluding other configurations.

[0019] Hereinafter, embodiments will be described in detail with reference to the accompanying drawings.

[0020] FIG. 1 is a block diagram showing the configuration of an article generation apparatus according to one embodiment.

[0021] The article generation apparatus 100 can automatically generate an article draft and provide it to a user based on the user's input and an article style including a plurality of generative AI regions. For example, the user's input may be a subject or keyword including the purpose of creating an article.

[0022] Specifically, the article generation device 100 may include a generative AI that generates responses to questions, and each generative AI area may contain AI command words that are questions to be asked of the generative AI. The article generation device 100 can generate responses to questions for each generative AI area based on the AI ​​command words contained in each generative AI area and user input, and can generate a draft article by inserting the responses obtained from the generative AI into the corresponding generative AI areas.

[0023] Furthermore, the article generation device 100 can acquire reference data from the user and, by referring to the reference data, can generate AI command words included in each of the generative AI domains and question-and-answer responses for each of the generative AI domains based on the user's input.

[0024] Furthermore, the article generation device 100 can use generative AI to provide at least one of the following: proofreading suggestions and revised text for the draft article.

[0025] The article generation device described above can be implemented using an electronic terminal or a server-client system.

[0026] Here, an electronic terminal can be embodied by a notebook PC, portable terminal, or wearable device that can connect to a remote server via a network or connect to other electronic terminals and servers. Here, a notebook PC includes, for example, a notebook PC equipped with a web browser or a laptop PC, and a portable terminal is, for example, a wireless communication device that guarantees portability and mobility, and can include all types of handheld-based wireless communication devices such as PCS (Personal Communication System), PDC (Personal Digital Cellular), PHS (Personal Handyphone System), PDA (Personal Digital Assistant), GSM (Global System for Mobile communications), IMT (International Mobile Telecommunication)-2000, CDMA (Code Division Multiple Access)-2000, W-CDMA (W-Code Division Multiple Access), Wibro (Wireless Broadband Internet), smartphone, and mobile WiMAX (Mobile Worldwide Interoperability for Microwave Access). Furthermore, wearable devices are information processing devices that can be worn directly on the human body, such as watches, glasses, accessories, clothing, and footwear, and may connect to a remote server or other terminal via a network, either directly or through other information processing devices.

[0027] The server can be embodied by a computer that can communicate via a network with an electronic terminal that has an application or web browser embedded for user interaction, or by a cloud computer server. The server can also include a storage device capable of storing data, or it can store data via a third server.

[0028] Referring to Figure 2, the article generation device 100 according to one embodiment may include a memory 110, a control unit 120, and an input / output unit 130.

[0029] Memory 110 can store a program for generating articles. Furthermore, memory 110 can store diverse article text format data for article generation, and can store at least one generative AI in program form that outputs content as a response to questions directed to the generative AI domain.

[0030] Generative AI refers to artificial intelligence that generates content such as text, images, videos, and audio as a response to questions posed by AI commands (prompts). For example, generative AI may be an AI model that generates a specific type of response, such as an image-generating AI or a text-generating AI, or it may be a multimodal AI model, like Gemini, that considers various data modalities such as text, images, videos, and audio together to generate responses.

[0031] The control unit 120 includes at least one processor and can execute a program stored in memory 110 to perform the article generation method described below. For example, the control unit 120 can execute a program stored in memory 110 by the processor to perform the article draft generation process described below, and for this purpose, it can operate a generative AI to obtain responses to questions for each generative AI domain.

[0032] Furthermore, the control unit 120 can control other components included in the article generation device 100. For example, the control unit 120 can provide recommended keywords based on the subject input via the input / output unit 130 and article data stored in the article management system.

[0033] The input / output unit 130 may include an interface for obtaining input from the user, a display panel or wearable display device for displaying the generated document draft, and output devices such as speakers, and may also include various types of input devices for receiving input from the user (e.g., keyboard, touchscreen, camera, etc.).

[0034] On the other hand, although not shown in the diagram, the article generation device 100 may further include a communication unit (not shown). The communication unit (not shown) can receive news data or press materials to be used as reference data by performing wired or wireless communication with a server, other device, or network on which the Contents Management System is running, in order to acquire reference data.

[0035] For this purpose, the communications unit (not shown) may include a communications module that supports at least one of a variety of wireless communications methods, and the communications module may be embodied in the form of a chipset. Wireless communications supported by the communications unit (not shown) may include, for example, WiFi (Wireless Fidelity), Wi-Fi Direct, Bluetooth®, UWB (Ultra Wide Band), or NFC (Near Field Communication).

[0036] The following describes in detail an article generation method according to one embodiment that operates by executing a program in the control unit 120. Unless otherwise specified, the process described below is performed by the control unit 120 executing a program stored in memory 110.

[0037] Figures 2 and 3 schematically illustrate the article generation process according to one embodiment.

[0038] The control unit 120 can provide an interface for generating draft articles, as shown in Figure 2, in response to a user request. For example, when a request occurs, such as clicking a specific icon like the AI ​​article generation 21, while a document creator operating as a web editor or application running on the web is being executed, the control unit 120 can provide an interface for generating draft articles.

[0039] Referring to Figure 2, the interface 200 for generating draft articles may include a subject input box 210, a keyword selection box 220, a template selection box 230, a detailed settings box 240, and a generation icon 260 for obtaining a request to create a draft article. The control unit 120 can obtain a request to create a draft article when the user clicks the generation icon 260.

[0040] Furthermore, the interface 200 for generating article drafts may include a file attachment icon 250 for attaching reference data.

[0041] When the control unit 120 receives a request to generate a draft article or a request for keyword recommendation, it can acquire a subject based on user input. The control unit 120 can acquire the text entered by the user in the subject input box 210 as the subject. The subject is the purpose for which the article is created and may be, for example, "Product A publicity," "Product B announcement," or "Service C launch publicity." When the control unit 120 receives a request to generate a draft article, it can generate a draft article based on the acquired subject.

[0042] The control unit 120 can provide recommended keywords based on the acquired subject. When the user clicks on the recommendation button 221, the control unit 120 can acquire a keyword recommendation request, and when it acquires a keyword recommendation request, it can provide recommended keywords based on the acquired subject and article data stored in the article management system (CMS).

[0043] For example, when the control unit 120 receives a topic acquired by the generative AI, it can refer to article data stored in the content management system (CMS) based on the topic acquired by the generative AI and acquire some of the words included in the topic or words with a high degree of relevance to the topic as recommended keywords. As an example, if "A product public relations" is acquired as the topic, and the product group of A product is an artificial intelligence enterprise solution, the control unit 120 can provide "A product," "artificial intelligence," "artificial intelligence solution," and "enterprise solution" as recommended keywords.

[0044] When the control unit 120 receives a request to generate a draft article, it can obtain keywords selected by the user and generate a draft article based on the obtained keywords.

[0045] On the other hand, the template setting box 230 is for selecting an already set article format and may include a template selection button 230 for obtaining a request for an article format. When the user clicks the template selection button 230, the control unit 120 can obtain a request for an article format and, in response to the request, can provide the user with an article format of a type that has already been set. The control unit 120 can provide the user with multiple article formats and, as shown in Figure 2, load the article format 201 selected by the user into the editing window and provide it to the user.

[0046] For example, the article format provided to the user may be at least one of the following: a straight article, a straight article about the launch of a product, a report-style analysis box article, a numerical analysis box article, a Q&A interview article, and an explanatory interview article.

[0047] As shown in Figure 2, the control unit 120 can provide the user with an article format by displaying the selected article format in the editing window. For example, if the article format selected by the user is a straight article, the control unit 120 can provide the article format 201 shown in Figure 2.

[0048] Article format 201 can include multiple generative AI regions 22A, 22B, 22C, 22D, and 22E.

[0049] Each of the generative AI regions 22A, 22B, 22C, 22D, and 22E can contain a name for the generative AI region and AI commands. The name of the generative AI region is an attribute of the response to be inserted into that region within the article format 201, and may be, for example, "Title," "Subtitle," "Image," "Lead Sentence," or "Body Text." The AI ​​commands may be the content of the questions posed to the generative AI in order to obtain the response to be inserted into the generative AI region.

[0050] The AI ​​commands included in the generative AI domain may vary depending on the attributes of the generative AI domain into which the AI ​​commands are inserted. For example, an AI command included in the "image" domain might be, "Please output an image appropriate to the article content, referring to the subject and keywords." Furthermore, the control unit 120 can insert a publisher-specific watermark and explanatory text into the response acquired in the "image" domain, i.e., the image generated by the generative AI.

[0051] When the control unit 120 receives a request to generate an article draft, it can execute a generative AI to obtain AI command words, themes, or responses to questions based on AI command words, themes, and keywords in the generative AI domain, and generate an article draft by inserting the obtained responses corresponding to the generative AI domain.

[0052] On the other hand, the control unit 120 can execute multiple queries simultaneously for each of the generative AI regions 22A, 22B, 22C, 22D, and 22E. For example, the control unit 120 can execute multiple generative AIs corresponding to each generative AI region, and input the AI ​​command words and themes contained in each of the generative AI regions 22A, 22B, 22C, 22D, and 22E into the corresponding generative AIs, thereby executing multiple queries simultaneously for each of the generative AI regions 22A, 22B, 22C, 22D, and 22E.

[0053] On the other hand, when generating an article draft, the control unit 120 can set detailed options, including at least one of the target age group, writing style, and length, based on the user's selection, and generate an article draft based on the set detailed options. To this end, as shown in Figure 2, the control unit 120 can provide selection boxes 241, 242, and 243 in the detailed setting box 240, which can acquire the user's selection for each of the target age group, writing style, and length.

[0054] For example, the target age group that users can select can be all ages, or from their 20s to 60s and over. The writing styles that users can select can be, in principle, a newsletter style, a columnist style, a storytelling style, or an interview style. The length that users can select can be, in principle, a short piece (600-1000 characters) or a long piece (1200-1800 characters).

[0055] When a request is made to generate a draft article, the control unit 120 receives the user's selection for at least one of the target age group, writing style, and length, sets detailed options, and can generate a draft article based on the set detailed options.

[0056] For example, if the topic is "Public Relations for Product A," the article format is "Straight Article," the target age group is "All Ages," the writing style is "Newsletter Style," and the length is set to "Short Article," then a draft of a public relations article for Product A can be generated that is 800 characters long, targeting the entire customer base, and written in a newsletter style with a straight article format.

[0057] To this end, the control unit 120 can enable the generative AI to learn the characteristics of each detailed option. For example, the control unit 120 can generate training data by classifying and labeling articles by writing style, and then enable the generative AI to learn based on the generated training data. Alternatively, the control unit 120 can generate training data by classifying and labeling articles in each field according to age group preference, and then enable the generative AI to learn based on the generated training data.

[0058] On the other hand, when the user clicks the file attachment icon 250 for attaching reference data, the control unit 120 can obtain a request to attach reference data and provide the user with a searcher to obtain the reference data.

[0059] When the control unit 120 receives a request to attach reference data, it can provide a searcher for selecting reference data based on the location of the reference data selected by the user. For example, the location of the reference data may be the user's electronic terminal, an integrated document management system, an article management system (CMS) where the articles of the publishing company are stored, or an external link.

[0060] For example, if the reference data is located on the user's electronic terminal or integrated document management system, the control unit 120 can provide the user with a searcher that can search for files within the user's electronic terminal or integrated document management system. If the reference data is located on an article management system (CMS), the control unit 120 can provide a searcher that allows the user to input the publisher, period, and search terms as conditions, and can search for news based on the input conditions.

[0061] The control unit 120 can attach the reference data selected by the provided searcher to the interface 300 310, as shown in Figure 3.

[0062] Referring to Figure 3, when an article creation request is received, the control unit 120 acquires and analyzes the attached reference data, then generates responses to each question in the generative AI domain by referring to the analysis results, and then generates an article draft by inserting the generated responses into the corresponding generative AI domains.

[0063] For example, the control unit 120 can obtain responses for each of the generative AI regions 22A, 22B, 22C, 22D, and 22E using Retrieval-Augmented Generation (RAG) based on reference data and AI command words.

[0064] Search and generation consists of two stages: search and generation. The control unit 120 searches for information necessary to generate content to be output as a response by querying AI command words included in the generative AI domain and user input (i.e., subject and detailed information) via a searcher, and then transmits the search results to the generative AI.

[0065] For example, the control unit 120 can search for information necessary to generate content to be output as a response to each of the generative AI domains by analyzing reference data based on the AI ​​command words, themes, and keywords contained in each of the generative AI domains via a searcher.

[0066] The searcher embeds AI command words, themes, and keywords contained in each of the generative AI domains 22A, 22B, 22C, 22D, and 22E as vectors, and maps the reference data to be searched into the embedded space, thereby allowing it to search for the information necessary to generate the content to be output as a response.

[0067] For this purpose, the Retriever may include an artificial intelligence model that embeds AI commands, themes, and keywords, and also embeds reference data to be searched. In some embodiments, the artificial intelligence model used to embed AI commands, themes, and keywords and the artificial intelligence model used to embed reference data to be searched may be different from each other.

[0068] Furthermore, the control unit 120 can use a searcher to search for information necessary to generate a response to a question, not only from reference data, but also from databases stored in memory 110 or from other devices connected via the network. The control unit 120 transmits the retrieved information to the generative AI, and based on the retrieved information and general knowledge, the generative AI can obtain a response to the question.

[0069] Here, as shown in Figure 3, the control unit 120 can simultaneously execute multiple queries for each of the generative AI regions 22A, 22B, 22C, 22D, and 22E. In addition, as shown in Figure 3, the control unit 120 can also provide an interface 300 that includes response windows 32A, 32B, 32C, 32D, and 32E, which allow confirmation that responses for each of the generative AI regions are output when a draft generation request is made.

[0070] On the other hand, the control unit 120 can complete the generated article draft. When the control unit 120 receives a request from the user to complete the article draft, it can complete the article draft generated using the generative AI. For example, the control unit 120 can receive a request from the user to complete the article draft by sensing a click of the article completion assistant button 33 in Figure 3. Here, the control unit 120 can complete the article draft using an AI different from the generative AI that generated the article draft.

[0071] When the control unit 120 receives a request to complete a draft article, it can complete the draft article by performing at least one of the following: article spelling check, article clarity check, and article vocabulary refinement check. As a result of completing the draft article, it can provide at least one of the following: proofreading suggestions (revision suggestions) and revised text.

[0072] Figures 4 to 6 illustrate article proofreading using generative AI according to one embodiment. Specifically, Figure 4 shows the results of article spelling inspection by generative AI, Figure 5 shows the results of article clarity inspection by generative AI, and Figure 6 shows the results of article vocabulary refinement inspection by generative AI.

[0073] When a request for completion of a manuscript is received, the control unit 120 performs a spelling check of the article draft by examining its spelling, word spacing, and non-standard language errors, and can provide proofreading suggestions to correct expressions deemed inappropriate to appropriate expressions based on the check results.

[0074] Specifically, the control unit 120 can inspect the draft article for spelling, word spacing, and non-standard language errors, detect inappropriate expressions, and use generative AI to search for appropriate expressions to replace the detected inappropriate expressions. The control unit 120 can then present the found appropriate expression as a revised sentence and, using generative AI, generate and provide the rationale for the proposed revised sentence.

[0075] Referring to Figure 4, the control unit 120 can provide a correction suggestion for the expression "1,000" which was deemed inappropriate, by suggesting "1,000" or "1,000" as the corrected expression. The control unit 120 can also provide the rationale for the correction suggestion as help.

[0076] The control unit 120 can replace the original text with the modified text selected by the user, and can also modify the original text based on the user's input. When the user clicks the Apply button, the control unit 120 can modify the original text with the content that the user directly entered into the modification input window.

[0077] On the other hand, when a request for completion of an article is received, the control unit 120 performs an article clarity check by considering at least one of the following: the relationship between the title and body of the article, paragraph divisions and readability, and parts requiring additional explanation. Based on the check results, it can generate and provide corrected sentences using generative AI for syntax deemed unclear.

[0078] The control unit 120 examines the relationship between the title and body of the generated article draft guide, the paragraph divisions and readability, and whether there are any parts that require additional explanation. As shown in Figure 5, it can provide the number of grammatical errors deemed inappropriate for each item and corrected sentences. For example, for the "1" inappropriate grammatical error in the "Relationship between Title and Body" item 510, the control unit 120 can provide the corrected sentence along with the rationale for the suggested correction (proofreading suggestion) as help.

[0079] Specifically, the control unit 120 can perform article clarity checks on draft articles using a large-scale language model (LLM) that has been pre-trained for natural language processing, which is a type of generative AI. For example, the control unit 120 can separate the title and the body text and embed them, and determine the relationship between the title and the body text by determining the similarity between the embedded title and the body text. If it is determined that the relationship between the title and the body text is low, it can generate and provide a revised text and the rationale for the proposed revision.

[0080] Furthermore, the control unit 120 can analyze the text content using a large-scale language model (LLM) to detect parts of the text that require additional semantic explanation. After classifying the text into semantic units, it can determine the text division and readability by comparing them with the text division of the actual article draft guide. The control unit 120 can generate and provide revised text and justifications for proposed revisions for parts that require additional explanation and parts that are deemed to have inappropriate text division and readability.

[0081] The control unit 120 can replace the original text with the modified text selected by the user, and can also modify the original text based on the user's input. When the user clicks the apply button, the control unit 120 can modify the original text with the content that the user directly entered into the modification input window.

[0082] Furthermore, when a manuscript completion request is received, the control unit 120 performs an article vocabulary refinement check by using generative AI to examine the presence or absence of at least one unnecessary repeated word and sensitive word in the article text, and can propose revised sentences for vocabulary deemed inappropriate as a result of the check.

[0083] Specifically, the control unit 120 can perform article vocabulary refinement checks on article drafts using a large-scale language model (LLM), which is a type of generative AI.

[0084] For example, the control unit 120 can input the draft article into a large-scale language model (LLM) that has been pre-trained for natural language processing, analyze the draft article to examine whether there are any unnecessary repeated words or sensitive words, and generate revised sentences and justifications for the proposed revisions for each inappropriate vocabulary.

[0085] Specifically, the control unit 120 can analyze the article draft using a large-scale language model (LLM), determine if there are any words that are used repeatedly, understand the meaning of those words in context, search for other words with similar meanings, and then generate and provide revised sentences and justifications for the revisions. Furthermore, the control unit 120 can analyze the article draft using a large-scale language model (LLM), check whether the article draft contains socially and culturally sensitive words, and generate and provide revised sentences and justifications for the revisions that either delete the sensitive words or replace them with other expressions.

[0086] Furthermore, the control unit 120 can display the number of vocabulary words deemed inappropriate for each item, and can provide a corrected sentence and the basis for the suggested correction for each inappropriate vocabulary word.

[0087] For example, as shown in Figure 6, the control unit 120 can display seven inappropriate sentence structures in the "Unnecessary Repeated Word Review" item 610, and for each inappropriate sentence structure, the control unit 120 can provide a corrected sentence along with the rationale for the suggested correction (proofreading suggestion) as help. The control unit 120 can replace the original text with a corrected sentence selected by the user, and can also modify the original text based on the user's input.

[0088] On the other hand, the control unit 120 can, at the user's request, provide a revised version of the generated article that reflects the reset detailed options for at least a portion of the generated draft article. For example, the control unit 120 can reset the detailed options for style and tone of articles for at least a portion of the text selected by the user, and then provide a revised version of the text selected by the user that reflects the reset detailed options.

[0089] Figure 7 illustrates the generation of corrected text by resetting detailed options according to one embodiment.

[0090] Referring to Figure 7, the control unit 120 can obtain a request to change detailed options by the user clicking the style change 71. The control unit 120 can provide an interface 700 that can reset the detailed options in response to the obtained request.

[0091] When the user selects a text and changes at least one of the detailed options items in the detailed settings 720, such as the target age group and writing style, and then presses the generate button, the control unit 120 can reflect the changed detailed option and generate and provide a revised text for the selected text 72 that reflects the changed detailed option.

[0092] On the other hand, the control unit 120 can register completed articles in the article management system (CMS) at the user's request.

[0093] For example, when the user clicks CMS registration 73, the control unit 120 can obtain a CMS registration request from the user. The control unit 120 can send the article from the editing window to the article management system (CMS) for registration. Here, the control unit 120 can send the article to the article management system (CMS) of the publication selected by the user and register the article in that publication's article management system (CMS). Here, the control unit 120 can insert the publication's watermark and explanatory text for the images inserted in the article.

[0094] As described above, the article generation device 100 according to one embodiment can reduce the time and cost of article creation by generating content in commonly used, standardized article formats using generative AI. Furthermore, it can link with an article management system (CMS) to recommend appropriate keywords related to the subject and conveniently attach reference data. It can also provide high-quality articles by analyzing the attached reference data and automatically generating article drafts based on the analysis results using the standardized article format and generative AI. Moreover, by using generative AI to complete the article drafts, the time and cost of completion can be reduced, and article quality can be easily improved.

[0095] As used in the above embodiments, the term "~part" refers to software or hardware components such as FPGAs (field programmable gate arrays) or ASICs, and "~part" performs a certain role. However, "~part" is not limited to software or hardware. "~part" can be configured to reside in an addressable storage medium, or to regenerate one or more processors. Thus, as an example, "~part" includes components such as software components, object-oriented software components, class components, and task components, as well as processes, functions, attributes, procedures, subroutines, segments of programmatic code, drivers, firmware, microcode, circuits, data, databases, data structures, tables, arrays, and variables.

[0096] Components and the functions provided within a "~part" can be combined with a smaller number of components and "~parts" or separated from additional components and "~parts".

[0097] Furthermore, the components and "~parts" can also be embodied to regenerate one or more CPUs within the device or security multimedia card.

[0098] On the other hand, Figure 8 is a flowchart illustrating an article generation method according to one embodiment.

[0099] The article generation method shown in Figure 8 includes a step of processing in chronological order in the article generation device 100 shown in Figures 1 to 7. Therefore, even if the details are omitted below, the content described above for the article generation device 100 shown in Figures 1 to 7 can also be used for the article generation method according to the embodiment shown in Figure 8.

[0100] Referring to Figure 8, the article generation device 100 can provide an article text format that includes multiple generative AI regions (S810).

[0101] For example, the article generation device 100 can provide the user with multiple article formatting options, and as shown in Figure 2, load the article formatting option 201 selected by the user into the editing window and provide it to the user.

[0102] Next, the article generation device 100 can create an article draft by responding to questions from the user regarding the subject and the generative AI domain, and inserting the response results into the generative AI domain (S820).

[0103] For this purpose, the article generation device 100 can provide an interface for generating draft articles, as shown in Figure 2, when providing an article text format.

[0104] Specifically, when the article generation device 100 receives a request to generate an article draft, it can execute a generative AI to obtain AI command words, themes, or responses to questions based on AI command words, themes, and keywords in the generative AI domain, and generate an article draft by inserting the obtained responses corresponding to the generative AI domain.

[0105] Additionally, when an article creation request is received, the article generation device 100 can acquire and analyze the attached reference data, then, based on the acquired subject or subject and keywords, refer to the analysis results to generate responses to each question in the generative AI domain, and insert the generated responses into the corresponding generative AI domains to generate a draft article.

[0106] To this end, when the article generation device 100 receives a keyword recommendation request, it can acquire a subject based on user input, and the control unit 120 can provide recommended keywords based on the acquired subject.

[0107] Although not shown in Figure 8, the article generation device 100, after generating an article draft, can complete the article draft when it receives a request to complete the draft, by performing at least one of the following: article spelling check, article clarity check, and article vocabulary refinement check. As a result of completing the draft, it can provide at least one of the following: proofreading suggestions (revision suggestions) and revised text.

[0108] Furthermore, after generating a draft article, the article generation device 100 can, at the user's request, provide a revised version of the generated draft article that reflects the reconfigured detailed options for at least a portion of the generated draft article.

[0109] As described above, the article generation method according to one embodiment can reduce the time and cost of article creation by automatically generating article drafts based on a standardized article format including AI command words and a generative AI. Furthermore, it can be linked with an article management system (CMS) to recommend appropriate keywords related to the subject and conveniently attach reference data. In addition, by analyzing the attached reference data and automatically generating article drafts based on the analysis results using the standardized article format and generative AI, high-quality articles can be provided. Moreover, by using the generative AI to complete the article draft, the time and cost of completion can be reduced, and the quality of articles can be easily improved.

[0110] The article generation method according to one embodiment described with reference to Figure 8 can also be embodied in the form of a computer-readable medium that stores computer-executable instructions and data. Here, the instructions and data can be stored in the form of program code, which, when executed by a processor, can generate a predetermined program module and perform a predetermined operation. The computer-readable medium may be any available medium accessible by a computer, and may include volatile and non-volatile media, and isolated and non-isolated media. The computer-readable medium may also be a computer recording medium. The computer recording medium may include any volatile and non-volatile, isolated and non-isolated media embodied by any method or technique for storing computer-readable information such as instructions, data structures, program modules, or other data. For example, the computer recording medium may be a magnetic storage medium such as an HDD and an SSD, an optical recording medium such as a CD, DVD, and a Blu-ray disc, or memory contained in a server accessible via a network.

[0111] Furthermore, the character trading method described in the embodiment based on Figure 8 can also be embodied in a computer program (or computer program product) that includes computer-executable instructions. A computer program includes programmable machine instructions that are processed by a processor and can be embodied in a high-level programming language, object-oriented programming language, assembly language, or machine language. Computer programs can also be recorded on various computer-readable recording media (e.g., memory, hard disk, magnetic / optical media, or SSD (Solid-State Drive)).

[0112] Therefore, the character trading method according to the embodiment described with reference to Figure 8 can be realized by executing the computer program described above on a computing device. The computing device may include a processor, memory, storage device, a high-speed interface connected to the memory and high-speed expansion port, and at least a portion of a low-speed interface connected to a low-speed bus and storage device. Each of these components is connected to one another using various buses and can be mounted on a common motherboard or in other suitable manner.

[0113] Here, the processor can process instructions within the computing device. Such instructions may include, for example, instructions stored in memory or storage devices for displaying graphic information to provide a GUI (Graphical User Interface) on an external input and output device, such as a display connected to a high-speed interface. In other embodiments, a number of processors and / or a number of buses can be appropriately used together with a number of memories and memory forms. The processor can also be embodied in a chipset consisting of chips containing a number of independent analog and / or digital processors.

[0114] Memory also stores information within a computing device. For example, memory can consist of volatile memory units or a collection thereof. Alternatively, memory can consist of non-volatile memory units or a collection thereof. Furthermore, memory may be other forms of computer-readable media, such as magnetic or optical disks.

[0115] Furthermore, memory can provide computing devices with a large storage space. Memory may be a computer-readable medium or a configuration including such a medium, and may include, for example, devices in a SAN (Storage Area Network) or other configurations, and may be floppy disk drives, hard disk drives, optical disk drives, or tape drives, flash memory, or other similar semiconductor memory devices or device arrays.

[0116] The embodiments described above are illustrative, and those with ordinary skill in the art to which the embodiments belong will understand that they can be easily modified into other specific forms without altering the technical idea or essential features of the embodiments. Therefore, it should be understood that the embodiments described above are illustrative in all respects and not limiting. For example, each component described as a single type can be implemented in a distributed manner, and similarly, components described as distributed can be implemented in a combined manner.

[0117] The scope of protection sought by this specification shall be determined by the claims set forth below rather than by the detailed description above, and shall be interpreted to include all forms of modification or variation derived from the meaning and scope of the claims and their equivalent concepts. [Explanation of symbols]

[0118] 100 Article Generator 110 memory 120 Control Unit 130 Input / output section

Claims

1. A memory containing a program and data for generating articles based on artificial intelligence, It includes at least one processor, a control unit that operates by executing a program stored in the memory, and automatically generates a draft article based on a subject obtained from the user and an article format that includes multiple generative AI fields, The control unit generates a draft article by executing generative AI, responding to questions directed to each of the generative AI regions based on AI command words contained in the subject and each of the generative AI regions, and inserting the response results into each of the generative AI regions.

2. The article generation apparatus according to claim 1, wherein the control unit obtains reference data from the user, responds to questions for each of the generative AI areas by referring to the results of analyzing the reference data, and generates an article draft by inserting the response results into each of the generative AI areas.

3. The article generation apparatus according to claim 1, wherein the control unit provides recommended keywords based on the subject, responds to questions regarding each of the generative AI areas based on the keyword selected by the user from among the recommended keywords, the subject, and the AI ​​command words included in the generative AI area, and generates an article draft by inserting the response results into each of the generative AI areas.

4. The article generation apparatus according to claim 1, wherein the control unit simultaneously performs multiple queries for each of the plurality of generative AI areas.

5. The article generation apparatus according to claim 1, wherein the control unit provides at least one of proofreading suggestions and revised text for the article draft using generative AI.

6. The article generation apparatus according to claim 1, wherein the control unit sets detailed options including at least one of target age group, writing style, and length, according to the user's selection, and generates an article draft based on the set detailed options.

7. The article generation apparatus according to claim 6, wherein the control unit, in response to a user's request, provides a revised text that reflects the reconfigured detail options for at least a portion of the generated draft article.

8. An article generation method performed by an artificial intelligence-based article generation device, The steps include providing an article format that includes multiple generative AI domains, This includes the step of automatically generating a draft article based on the subject matter obtained from the user and the article format, The method for generating an article includes the step of generating an article draft by executing a generative AI, responding to questions directed to each of the generative AI areas based on AI command words contained in the subject and each of the generative AI areas, and inserting the response results into each of the generative AI areas.

9. The article generation method according to claim 8, wherein the step of generating the article draft includes the step of obtaining reference data from a user, referring to the results of analyzing the reference data, responding to questions for each of the generative AI areas, and generating the article draft by inserting the response results into each of the generative AI areas.

10. The article generation method according to claim 8, wherein the step of generating the article draft is to simultaneously perform multiple questioning operations on each of the multiple generative AI domains.

11. The article generation method according to claim 8, further comprising the step of providing at least one of proofreading suggestions and revised text for the draft article using generative AI.

12. The article generation method according to claim 8, wherein the step of generating the article draft includes setting detailed options, at least one of the target age group, writing style, and length, at the user's selection, and generating the article draft based on the set detailed options.

13. The article generation method according to claim 8, further comprising the step of providing a revised text that reflects reconfigured detail options for at least a portion of the generated draft article, at the request of the user.

14. A computer program stored on a computer-readable recording medium, which is executed by an article generation device and performs the method described in claim 8.

15. A computer-readable recording medium on which a computer program that performs the method described in claim 8 is recorded.

Images