system

Abstract

Provide a system. [Solution] means for receiving natural language text from a user; means for using a pre-trained model to summarize the text; means for translating the summarized text into different languages; means for returning the translation result to the user; A system including:

Description

[Technical Field]

[0001] The technology of the present disclosure relates to a system. [Background technology]

[0002] Patent document 1 discloses a persona chatbot control method performed by at least one processor, the method including the steps of receiving a user utterance, adding the user utterance to a prompt including an instruction sentence related to a description of the chatbot character, encoding the prompt, and inputting the encoded prompt into a language model to generate a chatbot utterance in response to the user utterance. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Publication No. 2022-180282 Summary of the Invention [Problem to be solved by the invention]

[0004] In recent years, the increasing volume of information has created a need for users to efficiently understand large amounts of text data. In particular, the increasing exchange of information between different languages ​​has created a need for rapid and accurate summarization and translation. However, current technology lacks an efficient method for consistently summarizing and translating large amounts of text. To solve this problem, there is a need for a system that can automatically summarize text entered by a user and then translate the summary into a different language. [Means for solving the problem]

[0005] The present invention provides a system that includes means for receiving natural language text from a user, using a pre-trained model to summarize the text, means for translating the summarized text into a different language, and means for returning the translation result to the user, thereby enabling automatic summarization of text entered by a user and translating the summary into a different language for provision, thereby realizing efficient understanding of information and information exchange between multiple languages.

[0006] A "user" is a person or entity that uses the system and inputs natural language text.

[0007] A "natural language" is a language that humans use on a daily basis and is not bound to a specific form.

[0008] "Text" is a collection of sentences or paragraphs written in a natural language.

[0009] A "pre-trained model" is an algorithm or computational model that has been pre-trained on a large dataset and is used to perform a specific task (e.g., summarization or translation).

[0010] "Summarization" is the process or result of extracting important information from a longer text and presenting it in a shortened form.

[0011] "Translation" is the process of converting text written in one language into another language.

[0012] "Different language" means a language other than the language in which the original text is written.

[0013] A "system" is an entire structure consisting of multiple hardware and software components designed to perform a specific function or task. [Brief explanation of the drawings]

[0014] [Figure 1]1 is a conceptual diagram showing an example of the configuration of a data processing system according to a first embodiment. [Figure 2] 1 is a conceptual diagram showing an example of main functions of a data processing device and a smart device according to a first embodiment. [Figure 3] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a second embodiment. [Figure 4] FIG. 10 is a conceptual diagram showing an example of main functions of a data processing device and smart glasses according to a second embodiment. [Figure 5] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a third embodiment. [Figure 6] FIG. 11 is a conceptual diagram showing an example of main functions of a data processing device and a headset-type terminal according to a third embodiment. [Figure 7] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a fourth embodiment. [Figure 8] FIG. 10 is a conceptual diagram showing an example of main functions of a data processing device and a robot according to a fourth embodiment. [Figure 9] 1 shows an emotion map onto which multiple emotions are mapped. [Figure 10] 1 shows an emotion map onto which multiple emotions are mapped. [Figure 11] FIG. 3 is a sequence diagram showing a processing flow of the data processing system according to the first embodiment. [Figure 12] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 1. [Figure 13] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system according to the second embodiment when an emotion engine is combined. [Figure 14] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 2 when an emotion engine is combined. DETAILED DESCRIPTION OF THE INVENTION

[0015] An example of an embodiment of a system according to the technology of the present disclosure will be described below with reference to the accompanying drawings.

[0016] First, the terms used in the following description will be explained.

[0017] In the following embodiments, a coded processor (hereinafter simply referred to as a "processor") may be a single arithmetic device or a combination of multiple arithmetic devices. Furthermore, a processor may be a single type of arithmetic device or a combination of multiple types of arithmetic devices. Examples of arithmetic devices include a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), a GPGPU (General-Purpose computing on Graphics Processing Units), and an APU (Accelerated Processing Unit).

[0018] In the following embodiments, a coded RAM (Random Access Memory) is a memory in which information is temporarily stored and is used as a working memory by a processor.

[0019] In the following embodiments, the coded storage is one or more non-volatile storage devices that store various programs, various parameters, etc. Examples of non-volatile storage devices include flash memory (SSD (Solid State Drive)), magnetic disks (e.g., hard disks), and magnetic tapes.

[0020] In the following embodiments, a communication I / F (Interface) with a symbol is an interface including a communication processor, an antenna, etc. The communication I / F controls communication between multiple computers. Examples of communication standards applied to the communication I / F include wireless communication standards including 5G (5th Generation Mobile Communication System), Wi-Fi (registered trademark), Bluetooth (registered trademark), etc.

[0021] In the following embodiments, "A and / or B" is synonymous with "at least one of A and B." In other words, "A and / or B" means that it may be only A, only B, or a combination of A and B. Furthermore, in this specification, the same concept as "A and / or B" is also applied when three or more things are expressed connected by "and / or."

[0022] [First embodiment]

[0023] FIG. 1 shows an example of the configuration of a data processing system 10 according to the first embodiment.

[0024] 1, a data processing system 10 includes a data processing device 12 and a smart device 14. An example of the data processing device 12 is a server.

[0025] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0026] The smart device 14 includes a computer 36, a reception device 38, an output device 40, a camera 42, and a communication I / F 44. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The reception device 38, the output device 40, and the camera 42 are also connected to the bus 52.

[0027] The reception device 38 includes a touch panel 38A, a microphone 38B, and the like, and receives user input. The touch panel 38A detects contact with an indicator (for example, a pen or a finger) to receive user input by the touch of the indicator. The microphone 38B detects the user's voice to receive user input by voice. The control unit 46A transmits data indicating the user input received by the touch panel 38A and the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the data indicating the user input.

[0028] The output device 40 includes a display 40A and a speaker 40B, and presents data to the user 20 by outputting the data in a form of expression that the user 20 can perceive (for example, audio and / or text). The display 40A displays visible information such as text and images in accordance with instructions from the processor 46. The speaker 40B outputs audio in accordance with instructions from the processor 46. The camera 42 is a compact digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor.

[0029] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 control the exchange of various information between the processor 46 and the processor 28 via the network 54.

[0030] FIG. 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0031] 2, in the data processing device 12, a specific process is performed by the processor 28. A specific processing program 56 is stored in the storage 32. The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific process is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0032] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0033] In the smart device 14, the processor 46 performs the reception output process. The storage 50 stores a reception output program 60. The reception output program 60 is used in conjunction with the specific processing program 56 by the data processing system 10. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0034] Next, a description will be given of the specific processing performed by the specific processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0035] The present invention provides a system that summarizes text written in a natural language by a user and then translates the summary into a different language. This system receives text entered by a user, summarizes it briefly, and automatically translates it into other languages, such as Japanese, thereby supporting efficient understanding of information and communication between multiple languages.

[0036] Program processing

[0037] The program for this system is as follows: The server summarizes text using a pre-trained natural language processing model (e.g., the BART model), and uses a translation API provided by a third party (e.g., Google Translate) for translation.

[0038] First, a user accesses the server using a terminal and inputs a long piece of text. This text is sent to the server as a POST request. The server receives the text and inputs it into a pre-trained model to automatically generate a summary.

[0039] The server then inputs the generated summary into a third-party translation API, which translates the summarized text into the desired language (e.g., Japanese). Once the translation is complete, the server returns the summary and translation results to the device as a JSON response. The user can view the summary and translation results on the device screen.

[0040] Specific examples

[0041] Here's a concrete example: Suppose the user types the following long sentence:

[0042] “Artificial intelligence has seen rapid advancements in recent years. These advancements are evident in various fields such as healthcare, finance, and education. The implementation of AI systems has led to more efficient processes, better decision-making, and significant cost savings. However, ethical concerns remain, particularly in terms of data privacy and algorithmic bias.”

[0043] When the terminal sends this text to the server, the server generates a summary like this:

[0044] “Artificial intelligence has seen rapid advancements in recent years, improving efficiency and decision-making in various fields such as healthcare, finance, and education. However, ethical concerns like data privacy and bias remain.”

[0045] The server then translates this summary into a different language, for example Japanese, to produce the final result:

[0046] "Artificial intelligence has advanced rapidly in recent years, improving efficiency and decision-making in sectors such as healthcare, finance and education. However, ethical concerns such as data privacy and bias remain."

[0047] In this way, the present invention realizes a series of processes for summarizing text entered by a user and then translating the summary into a different language. The embodiment of the invention embodies a series of processes including multiple steps, such as data exchange between a server and a terminal, summarizing the text, and translating it.

[0048] The processing flow will be explained below.

[0049] Step 1:

[0050] A user uses a terminal to input a long text in a natural language into a text input form.

[0051] Example: "Artificial intelligence has seen rapid advancements in recent years. ..."

[0052] Step 2:

[0053] The device sends the entered text to the server as a POST request, with the request body containing the text in JSON format.

[0054] Example: {"text": "Artificial intelligence has seen rapid advancements in recent years. ..."}

[0055] Step 3:

[0056] The server receives the POST request, extracts the JSON data from the request body, and uses request.json to get the text field.

[0057] Example: text = data['text']

[0058] Step 4:

[0059] The server tokenizes the received text using a pre-trained natural language processing model (BART model). The tokenizer.encode method is used to convert the text into tokens for summarization.

[0060] Example: inputs = tokenizer.encode("summarize: " + text, return_tensors='pt')

[0061] Step 5:

[0062] The server passes the tokenized input to a natural language processing model to generate a summary of the text. The summary result is obtained using the model.generate method.

[0063] Example: summary_ids = model.generate(inputs, max_length=150, min_length=40, ...)

[0064] Step 6:

[0065] The server decodes the generated summary and converts it into a natural language form that the user can understand. Use the tokenizer.decode method to obtain the summary text.

[0066] Example: summary = tokenizer.decode(summary_ids[0], skip_special_tokens=True)

[0067] Step 7:

[0068] The server inputs the summarized text into a third-party translation API (Google Translate API) to translate it into a different language, and retrieves the translation result using the translator.translate method.

[0069] Example: translation = translator.translate(summary, src='en', dest='ja').text

[0070] Step 8:

[0071] The server compiles the summary and translation results into JSON format and generates a response. The jsonify function is used to create the response.

[0072] Example: return jsonify({'summary': summary, 'translation': translation})

[0073] Step 9:

[0074] The server sends a JSON response back to the device, which receives the response and displays the summary and translation results to the user.

[0075] Example: {"summary": "Artificial intelligence has seen rapid advancements...", "translation": "Artificial intelligence has seen rapid advancements..."}

[0076] This completes the process of efficiently summarizing user-entered text and translating the summary into different languages.

[0077] Example 1

[0078] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0079] In today's world, the exchange and understanding of information across multiple languages ​​is extremely important. However, it takes a great deal of time and effort for users to understand long texts written in languages ​​other than their native language. Furthermore, there are limited technologies that can automatically summarize texts and appropriately translate them without specialized knowledge. This makes it difficult for users to easily exchange information across multiple languages, hindering efficient communication.

[0080] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[0081] In this invention, the server includes means for receiving text written in a natural language from a user, means for using a pre-trained generative model to automatically summarize the text, means for using a translation service to translate the summarized text into a different language, and means for returning the translation result to the user's terminal, thereby enabling the user to easily summarize long texts and further translate them into different languages, thereby enabling efficient information comprehension and smooth communication between multiple languages.

[0082] "Text written in a natural language" refers to text written in a language that humans use on a daily basis.

[0083] "User" refers to a person or organization that uses the system.

[0084] A "server" refers to a computer that provides services to other computers over a network.

[0085] "Terminal" refers to the computer or device through which a user interacts with a server.

[0086] A "summary" is a concise description of the main information extracted from a longer original text.

[0087] A "pre-trained generative model" refers to a model that has been trained in advance on a large amount of data and can automatically perform a specific task (in this case, summarization).

[0088] "Translation" refers to the conversion of text written in one language into a different language.

[0089] "Translation Service" means an online application or API that provides text translation.

[0090] "Translation Result" means text that has been converted into a different language by a Translation Service.

[0091] A "pre-trained natural language processing model" refers to a model that has been trained on a large amount of text data in advance to perform natural language processing tasks.

[0092] "Translation API" refers to an application programming interface provided to enable translation functionality to be used from any program.

[0093] "JSON format" refers to a data exchange format called JavaScript (registered trademark) Object Notation.

[0094] This invention is a system that automatically summarizes natural language text entered by a user and translates the summary into a different language. The system has the following main functions:

[0095] 1. User enters text

[0096] A user inputs long pieces of text through a web browser or application.

[0097] 2. Sending text from the device to the server

[0098] The device sends the text entered by the user to the server using an HTTP POST request.

[0099] 3. Server-generated text summarization

[0100] The server uses a pre-trained generative model (e.g., the BART model) for text summarization, which summarizes long texts into short, concise summaries.

[0101] 4. Server-Based Abstract Translation

[0102] The summarized text is sent via the server to a translation service (e.g., Google Translate API) and translated into the desired language.

[0103] 5. The server returns the translation results

[0104] The server returns the translated text in JSON format to the device, where the user can view a summary and translation results.

[0105] As a concrete example, consider the case where a user enters the following long sentence:

[0106] Artificial intelligence has seen rapid advancements in recent years. These advancements are evident in various fields such as healthcare, finance, and education. The implementation of AI systems has led to more efficient processes, better decision-making, and significant cost savings. However, ethical concerns remain, particularly in terms of data privacy and algorithmic bias.

[0107] When the device sends this text to the server, the server generates a summary like this:

[0108] Artificial intelligence has seen rapid advancements in recent years, improving efficiency and decision-making in various fields such as healthcare, finance, and education. However, ethical concerns like data privacy and bias remain.

[0109] The server then translates this summary into a different language, for example Japanese:

[0110] Artificial intelligence has advanced rapidly in recent years, improving efficiency and decision-making in fields such as medicine, finance, and education, but ethical concerns such as data privacy and bias remain.

[0111] In this way, users can easily summarize long texts they input and then translate the summaries into different languages. This system allows users to efficiently exchange information between multiple languages.

[0112] Prompt Sentence Examples

[0113] Summarize the following long passage and translate it into Japanese:

[0114] "Artificial intelligence has seen rapid advancements in recent years. These advancements are evident in various fields such as healthcare, finance, and education. The implementation of AI systems has led to more efficient processes, better decision-making, and significant cost savings. However, ethical concerns remain, particularly in terms of data privacy and algorithmic bias."

[0115] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0116] System program processing flow

[0117] Step 1: User Enters Text

[0118] Specific description:

[0119] A user inputs a long piece of natural language text into an input field in a web browser or application on his or her device.

[0120] Specific behavior:

[0121] A user enters the following text into a browser input field:

[0122] Artificial intelligence has seen rapid advancements in recent years. These advancements are evident in various fields such as healthcare, finance, and education. The implementation of AI systems has led to more efficient processes, better decision-making, and significant cost savings. However, ethical concerns remain, particularly in terms of data privacy and algorithmic bias.

[0123] Step 2: Send text to the server using the device

[0124] Specific description:

[0125] The terminal receives the text entered by the user and sends it to the server as a POST request.

[0126] Input and Output:

[0127] Input: Long text entered by the user

[0128] Output: HTTP POST request sent to the server

[0129] Specific behavior:

[0130] The device sends the following HTTP POST request to the server:

[0131] POST / summarize_translate

[0132] Content-Type: application / json

[0133] {

[0134] "text": "Artificial intelligence has seen rapid advancements in recent years. These advancements are evident in various fields such as healthcare, finance, and education. The implementation of AI systems has led to more efficient processes, better decision-making, and significant cost savings. However, ethical concerns remain, particularly in terms of data privacy and algorithmic bias."

[0135] }

[0136] Step 3: Server-generated text summarization

[0137] Specific description:

[0138] The server inputs the received text into a pre-trained generative model (the BART model) and automatically generates a summary.

[0139] Input and Output:

[0140] Input: Long text sent from the terminal

[0141] Output: A short summary of the text

[0142] Specific behavior:

[0143] The server invokes the BART model and generates a summary like this:

[0144] Artificial intelligence has seen rapid advancements in recent years, improving efficiency and decision-making in various fields such as healthcare, finance, and education. However, ethical concerns like data privacy and bias remain.

[0145] Step 4: Server translation of the summary

[0146] Specific description:

[0147] The server sends the generated summary to a translation service (Google Translate API) for translation into the desired language.

[0148] Input and Output:

[0149] Input: A short text to be summarized

[0150] Output: The translated text

[0151] Specific behavior:

[0152] The server calls the Google Translate API and obtains the translation result as follows:

[0153] Artificial intelligence has advanced rapidly in recent years, improving efficiency and decision-making in fields such as medicine, finance, and education, but ethical concerns such as data privacy and bias remain.

[0154] Step 5: The server returns the translation results to the device

[0155] Specific description:

[0156] The server returns the translated text and a summary to the device in JSON format.

[0157] Input and Output:

[0158] Input: translated text and summary

[0159] Output: JSON response sent back to the device

[0160] Specific behavior:

[0161] The server sends the following JSON response to the device:

[0162] {

[0163] "summary": "Artificial intelligence has seen rapid advancements in recent years, improving efficiency and decision-making in various fields such as healthcare, finance, and education. However, ethical concerns like data privacy and bias remain.",

[0164] "translation": "Artificial intelligence has advanced rapidly in recent years, improving efficiency and decision-making in sectors like healthcare, finance, and education. However, ethical concerns such as data privacy and bias remain."

[0165] }

[0166] Step 6: User summarization and translation confirmation

[0167] Specific description:

[0168] The user checks the returned summary and translation results on the screen of the terminal.

[0169] Input and Output:

[0170] Input: JSON response returned from the server

[0171] Output: Summary and translation results displayed on the terminal

[0172] Specific behavior:

[0173] The user sees the summary and translation results in the browser or application screen, as shown below.

[0174] summary:

[0175] Artificial intelligence has seen rapid advancements in recent years, improving efficiency and decision-making in various fields such as healthcare, finance, and education. However, ethical concerns like data privacy and bias remain.

[0176] translation:

[0177] Artificial intelligence has advanced rapidly in recent years, improving efficiency and decision-making in fields such as medicine, finance, and education, but ethical concerns such as data privacy and bias remain.

[0178] (Application example 1)

[0179] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0180] On modern e-commerce sites, reviews posted by users are written in a wide variety of languages, making it difficult for multilingual users to understand the content of the reviews. Furthermore, reading the entire lengthy review is inefficient, and summaries are required. The present invention addresses this issue.

[0181] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[0182] In this invention, the server includes means for receiving natural language text from a user, means for using a pre-trained model to summarize the text, means for translating the summarized text into a different language, means for applying the summarized text and the translation result to user reviews on an e-commerce site, and means for returning the translation result to the user, thereby enabling multilingual users to efficiently understand the review content.

[0183] A "user" is a user of the system, and is an individual or organization that inputs text in a natural language.

[0184] "Natural language" refers to a language that humans use on a daily basis and that does not have a specific structure or form.

[0185] "Text" is character string information written in a natural language.

[0186] "Pre-trained model" refers to a machine learning model that has been pre-trained with a large amount of reference data for the purpose of natural language processing.

[0187] A "summary" is a short sentence that succinctly expresses the main points of a long text.

[0188] "Translation" is the act of converting text written in one language into another language.

[0189] "Different language" refers to a language that is different from the natural language input by the user.

[0190] "Translation result" refers to the output of the translated text.

[0191] An "e-commerce site" is a website that sells goods and services over the Internet.

[0192] "User reviews" refer to texts in which users express their opinions or evaluations of products or services they have purchased.

[0193] "Replying" refers to the act of providing an appropriate output in response to an input from a user.

[0194] "Means" refers to elements necessary to achieve a specific function.

[0195] The present invention is a system that automatically summarizes natural language text entered by a user, translates the summary into different languages, and applies it to user reviews on e-commerce sites.

[0196] The system includes the following means:

[0197] 1. A means for users to input natural language text

[0198] 2. Using a pre-trained model to summarize input text

[0199] 3. A means of translating the summarized text into different languages

[0200] 4. A method for applying summarized text and translation results to user reviews on e-commerce sites

[0201] 5. Means of returning translation results to the user

[0202] Hardware and software used

[0203] Hardware: Server, user devices (smartphones, PCs)

[0204] Software: BART model (summarization), Google Translate API (translation), web service (RESTful API), front-end (React, Angular, etc.)

[0205] Processing procedures and examples

[0206] 1. Processing when posting a review

[0207] A user submits a review on an e-commerce site, and the text is sent to the server as a POST request.

[0208] 2. Summary Generation

[0209] The server inputs the received reviews into a pre-trained BART model, a machine learning model pre-trained on large amounts of data for the purpose of natural language processing, to generate summaries.

[0210] 3. Translation

[0211] The server inputs the summarized text into the Google Translate API, which is a third-party web API that provides translation services, and translates it into a different specified language (e.g., Japanese).

[0212] 4. Application

[0213] The server applies the summarized text and translation results to user reviews on an e-commerce site, where the translated summaries are displayed in the review section.

[0214] 5. Response Generation

[0215] The server stores the summary and translation results in a cloud database and returns them to the user's device in JSON format, where the user can view the summary and translation results in real time.

[0216] Specific examples

[0217] For example, if a user posts a review like this:

[0218] “Artificial intelligence has seen rapid advancements in recent years. These advancements are evident in various fields such as healthcare, finance, and education. The implementation of AI systems has led to more efficient processes, better decision-making, and significant cost savings. However, ethical concerns remain, particularly in terms of data privacy and algorithmic bias.”

[0219] The system generates a summary like this, which is then translated into Japanese:

[0220] Abstract: "Artificial intelligence has seen rapid advancements in recent years, improving efficiency and decision-making in various fields such as healthcare, finance, and education. However, concerns like data privacy and ethical bias remain."

[0221] Translation result: "Artificial intelligence has advanced rapidly in recent years, improving efficiency and decision-making in fields such as medicine, finance, and education. However, ethical concerns such as data privacy and bias remain."

[0222] In this way, multilingual users can efficiently understand the review content.

[0223] Prompt Sentence Examples

[0224] "Please summarize the long review below to a maximum of 150 characters and then translate it into Japanese. Review: {Review content}"

[0225] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0226] Step 1:

[0227] A user submits a review by entering natural language text into an e-commerce site. This text is sent from the user's device to the server as a POST request. The input text is a string of natural language characters entered by the user as a review. When the user presses the submit button for the text they entered, the text data is sent to the server in JSON format.

[0228] Step 2:

[0229] The server detects the review text received from the user and inputs it into a pre-trained generative AI model (e.g., the BART model) to summarize it. The input text data is the natural language review text received from the user. The server inputs this text data into the BART model, which summarizes the text based on important information. This results in a shorter summary being output.

[0230] Step 3:

[0231] The server inputs the generated summary into a translation API (for example, Google Translate API) to translate it into a different specified language. The input data is the summary and the language information for the translation. The server sends the summary to the translation API and receives the translation result. As a result, the summary translated into the specified language is output.

[0232] Step 4:

[0233] The server applies the translated summary to user reviews on the e-commerce site. The input data is the translated summary. The server inserts the summary into the review section, making it visible to other users. This step allows the summarized translation to be displayed in the review section.

[0234] Step 5:

[0235] The server stores the summary and translation results in a cloud database and returns them to the user's device in JSON format. The input data is the summary and translation. The server stores this data in a cloud database and sends it to the user's device as a response. The user can check the summary and translation results on their device.

[0236] The above processing steps enable users who speak different languages ​​to efficiently understand the review content.

[0237] Furthermore, an emotion engine that estimates the user's emotion may be combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59 and perform identification processing using the user's emotion.

[0238] The present invention combines a system that summarizes text written by a user in a natural language and then translates the summary into a different language with an emotion engine that recognizes emotions from the user's text. The system receives text entered by a user, summarizes it, translates the summary into another language, and recognizes emotions from the text and reflects them in the results.

[0239] Program processing

[0240] The program for this system is as follows. The server summarizes text using a pre-trained natural language processing model (e.g., the BART model). It also uses a translation API provided by a third party (e.g., Google Translate) for translation and an emotion engine for emotion recognition.

[0241] First, a user accesses the server using a terminal and inputs a long piece of text. This text is sent to the server as a POST request. The server receives the text and recognizes emotions using an emotion engine. The recognized emotions influence the subsequent summarization and translation processes.

[0242] The server then inputs the text into a pre-trained model to automatically generate a summary. The generated summary is then input into a third-party translation API for translation into the desired language (e.g., Japanese). Based on the output of the emotion engine, the server adjusts the summary and translation results to optimally respond to the user.

[0243] Specific examples

[0244] Here's a concrete example: Suppose the user types the following long sentence:

[0245] “Artificial intelligence has seen rapid advancements in recent years. These advancements are evident in various fields such as healthcare, finance, and education. The implementation of AI systems has led to more efficient processes, better decision-making, and significant cost savings. However, ethical concerns remain, particularly in terms of data privacy and algorithmic bias.”

[0246] When the device sends this text to the server, the server first uses its emotion engine to recognize emotions such as:

[0247] "Positive emotions: progress, efficiency, cost savings

[0248] Negative Emotions: Ethical Concerns, Data Privacy, and Algorithmic Bias.”

[0249] The server then uses this information to generate a summary:

[0250] “Artificial intelligence has seen rapid advancements in recent years, improving efficiency and decision-making in various fields such as healthcare, finance, and education. However, ethical concerns like data privacy and bias remain.”

[0251] The server then translates this summary into a different language, say Japanese, producing the final result:

[0252] "Artificial intelligence has advanced rapidly in recent years, improving efficiency and decision-making in sectors such as healthcare, finance and education. However, ethical concerns such as data privacy and bias remain."

[0253] The translation takes into account the output of the emotion engine and makes appropriate adjustments, such as emphasizing key points and appropriately reflecting positive and negative sentiment.

[0254] This enables a series of processes to efficiently summarize text entered by a user, translate that summary into different languages, and simultaneously recognize and appropriately reflect the sentiment of the text. This system, combined with an emotion engine, can enrich multilingual communication and more accurately meet user needs.

[0255] The processing flow will be explained below.

[0256] Step 1:

[0257] A user uses a terminal to input a long text in a natural language into a text input form.

[0258] Example: "Artificial intelligence has seen rapid advancements in recent years. ..."

[0259] Step 2:

[0260] The device sends the entered text to the server as a POST request, with the request body containing the text in JSON format.

[0261] Example: {"text": "Artificial intelligence has seen rapid advancements in recent years. ..."}

[0262] Step 3:

[0263] The server receives the POST request, extracts the JSON data from the request body, and uses request.json to get the text field.

[0264] Example: text = data['text']

[0265] Step 4:

[0266] The server uses an emotion engine to recognize the emotion of the received text, a process that classifies emotions such as positive, negative, or neutral based on the content of the text.

[0267] Example: emotion = emotion_engine.analyze(text)

[0268] Step 5:

[0269] The server tokenizes the received text using a pre-trained natural language processing model (BART model). The tokenizer.encode method is used to convert the text into tokens for summarization.

[0270] Example: inputs = tokenizer.encode("summarize: " + text, return_tensors='pt')

[0271] Step 6:

[0272] The server passes the tokenized input to a natural language processing model to generate a summary of the text. The summary result is obtained using the model.generate method.

[0273] Example: summary_ids = model.generate(inputs, max_length=150, min_length=40, ...)

[0274] Step 7:

[0275] The server decodes the generated summary and converts it into a natural language form that the user can understand. Use the tokenizer.decode method to obtain the summary text.

[0276] Example: summary = tokenizer.decode(summary_ids[0], skip_special_tokens=True)

[0277] Step 8:

[0278] The server inputs the summarized text into a third-party translation API (Google Translate API) to translate it into a different language, and retrieves the translation result using the translator.translate method.

[0279] Example: translation = translator.translate(summary, src='en', dest='ja').text

[0280] Step 9:

[0281] The server adjusts the summarization and translation results based on the output of the emotion engine, for example by changing the wording to emphasize positive emotions.

[0282] Example: adjusted_translation = adjust_based_on_emotion(translation, emotion)

[0283] Step 10:

[0284] The server generates a response containing the summary and adjusted translation results in JSON format. The response is created using the jsonify function.

[0285] Example: return jsonify({'summary': summary, 'translation': adjusted_translation})

[0286] Step 11:

[0287] The server sends a JSON response back to the device, which receives the response and displays the summary and translation results to the user.

[0288] Example: {"summary": "Artificial intelligence has seen rapid advancements...", "translation": "Artificial intelligence has seen rapid advancements..."}

[0289] This allows for a system that efficiently summarizes user-entered text, translates the summary into different languages, and recognizes the sentiment of the text and reflects it in the results.

[0290] Example 2

[0291] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0292] Many conventional natural language processing systems are unable to recognize and reflect the sentiment of text when summarizing or translating it. This can result in summaries and translations that do not adequately reflect the user's intentions and sentiments, potentially reducing the quality of communication. Furthermore, the lack of automatic adjustments to translation results increases the user's workload.

[0293] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0294] In this invention, the server includes means for receiving natural language text from a user, means for recognizing emotions from the text, means for using a pre-trained model to summarize the text, means for translating the summarized text into a different language, means for adjusting the translation result based on the recognized emotions, and means for returning the translation result to the user, thereby making it possible to provide a summary and translation result that appropriately reflects the user's intentions and emotions.

[0295] A "user" is a person or entity that utilizes the system to input natural language text and receive summaries and translation results.

[0296] "Natural language text" is a sentence or text written in a normal human language.

[0297] "Means for recognizing emotions" refers to means that use engines or algorithms, such as sentiment analysis models, to analyze and identify emotions within text.

[0298] A "pre-trained model" is an artificial intelligence model that has been trained in advance based on a large amount of text data to perform a specific task (in this case, summarization) with high accuracy.

[0299] A "means of translation" is a means of using technology or services, such as a translation API, to convert text written in one language into another language.

[0300] "Means for responding to the user" refers to the technology or service that sends the processing results to the user's terminal so that the user can view them.

[0301] A "server" is a computer system that receives requests from users, performs various processes internally, and returns the final results.

[0302] The present invention combines a system that summarizes text written by a user in a natural language and then translates the summary into a different language with an emotion engine that recognizes emotions from the user's text. The system receives text entered by a user, summarizes it, translates the summary into another language, and recognizes emotions from the text and reflects them in the results.

[0303] Program processing

[0304] The processing of the system program is explained below.

[0305] The server uses a pre-trained natural language processing model (e.g., the BART model) to summarize the text, a third-party translation API (e.g., Google Translate) for translation, and an emotion engine for emotion recognition.

[0306] First, a user accesses the server using a terminal and inputs a long piece of text. This text is sent to the server as a POST request. The server receives the text and uses an emotion engine to recognize emotions. The recognized emotions influence the subsequent summarization and translation processes.

[0307] The server then inputs the text into a pre-trained model to automatically generate a summary. The generated summary is then input into a third-party translation API for translation into the desired language (e.g., Japanese). Based on the output of the emotion engine, the server adjusts the summary and translation results to optimally respond to the user.

[0308] Specific examples

[0309] Here's a concrete example: Suppose the user types the following long sentence:

[0310] “Artificial intelligence has seen rapid advancements in recent years. These advancements are evident in various fields such as healthcare, finance, and education. The implementation of AI systems has led to more efficient processes, better decision-making, and significant cost savings. However, ethical concerns remain, particularly in terms of data privacy and algorithmic bias.”

[0311] When the device sends this text to the server, the server first uses its emotion engine to recognize emotions such as:

[0312] "Positive emotions: progress, efficiency, cost savings

[0313] Negative Emotions: Ethical Concerns, Data Privacy, and Algorithmic Bias.”

[0314] The server then uses this information to generate a summary:

[0315] “Artificial intelligence has seen rapid advancements in recent years, improving efficiency and decision-making in various fields such as healthcare, finance, and education. However, ethical concerns like data privacy and bias remain.”

[0316] The server then translates this summary into a different language, say Japanese, producing the final result:

[0317] "Artificial intelligence has advanced rapidly in recent years, improving efficiency and decision-making in sectors such as healthcare, finance and education. However, ethical concerns such as data privacy and bias remain."

[0318] The translation takes into account the output of the emotion engine and makes appropriate adjustments, such as emphasizing key points and appropriately reflecting positive and negative sentiment.

[0319] This enables a series of processes to efficiently summarize text entered by a user, translate that summary into different languages, and simultaneously recognize and appropriately reflect the sentiment of the text. This system, combined with an emotion engine, can enrich multilingual communication and more accurately meet user needs.

[0320] Prompt Sentence Examples

[0321] An example of a prompt to use when a user has entered a long text and wants a summary of that text is:

[0322] "Summarize the following long English sentence and translate it into Japanese. Also, please specify the emotion that is recognized."

[0323] Then, here's an example of the specific text the user would enter:

[0324] “Artificial intelligence has seen rapid advancements in recent years. These advancements are evident in various fields such as healthcare, finance, and education. The implementation of AI systems has led to more efficient processes, better decision-making, and significant cost savings. However, ethical concerns remain, particularly in terms of data privacy and algorithmic bias.”

[0325] This provides clear instructions for the system to perform the expected functions accurately.

[0326] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0327] Step 1:

[0328] Entering and sending text

[0329] First, a user accesses the system using a terminal and inputs a long piece of natural language text. When the user inputs the text and clicks the "Submit" button, the terminal sends the input text to the server as a POST request. Specifically, the text is input as a browser form, and JavaScript processes it and sends it to the server.

[0330] Input: Natural language text entered by the user

[0331] Output: POST request from the terminal to the server

[0332] Step 2:

[0333] Text Receiving and Emotion Recognition

[0334] The server receives the POST request sent from the device and passes the received text to an emotion engine (e.g., an emotion analysis model) to analyze the type and intensity of emotion. This analysis process identifies positive and negative emotions contained in the text.

[0335] Input: The user's natural language text included in the POST request

[0336] Output: Sentiment analysis results (positive and negative) of the text

[0337] Step 3:

[0338] Summary Generation

[0339] The server inputs the emotion-recognized text into a pre-trained natural language processing model (e.g., the BART model). The model analyzes the text, extracts key information, and generates a summary. The summary is generated quickly and reduces the content while preserving the essence of the original text.

[0340] Input: Natural language text with sentiment information

[0341] Output: Generated summary text

[0342] Step 4:

[0343] Executing the translation

[0344] The server sends the generated summary text to a translation API provided by a third party (e.g., a translation service) for translation into the desired language. The translation API connects to the external service via an HTTP request and returns the translated text.

[0345] Input: Summarized text

[0346] Output: The translated text

[0347] Step 5:

[0348] Adjusting the translation results

[0349] The server adjusts the translation results based on the output of the emotion engine, specifically by emphasizing key points and modifying the translation to appropriately reflect positive and negative emotions, so that the information is conveyed in the most optimal way for the user.

[0350] Input: Sentiment analysis results and translated text

[0351] Output: Sentiment-informed translated text

[0352] Step 6:

[0353] Returning the results

[0354] Finally, the server sends the adjusted translation back to the user's device. The server generates an HTTP response and sends the text in HTML or JSON format to the user's device, where the user can view the translation results.

[0355] Input: Adjusted translation text

[0356] Output: Translation results displayed on the user's device

[0357] Through these steps, users can efficiently summarize natural language text, translate the summary into different languages, and even recognize emotions to achieve optimally reflected results.

[0358] (Application example 2)

[0359] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0360] The present invention relates to a system that efficiently performs summarization, translation, and emotion recognition for long natural language texts entered by users, and aims to provide appropriate summarization and translation results, particularly in multilingual content distribution services. Furthermore, because conventional technologies have been insufficient in recognizing emotions and reflecting the results, there is a need for a system that can provide more natural and appropriate output by recognizing emotions in text.

[0361] The identification process by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for receiving natural language text from a user, means for using a pre-trained model to summarize the text, means for translating the summarized text into a different language, means for analyzing emotions extracted from the text, means for adjusting the summary and translation result based on the emotions, and means for returning the adjusted translation result to the user. This allows the user to obtain a translation result in which the input text is efficiently summarized and appropriate emotions are reflected.

[0362] "User" means any person or entity that utilizes the System to enter text and receive results.

[0363] A "natural language" is a language that humans use on a daily basis (e.g., Japanese, English, etc.).

[0364] "Text" refers to natural language sentences entered by a user.

[0365] A "pre-trained model" is a machine learning model for natural language processing that has been trained in advance using a large amount of data.

[0366] "Summarization" is the process of extracting important information from a longer text and summarizing it in a short form.

[0367] "Translation" is the process of converting text written in one language into a different language.

[0368] "Emotion" refers to a user's feelings or emotional state, such as positive or negative, extracted from text.

[0369] "Analyzing sentiment" is the process of recognizing sentiment in text and determining whether it is positive, negative, or neutral.

[0370] "Adjustment" refers to optimizing the summary and translation results by reflecting the results of sentiment analysis.

[0371] "Returning results" means providing the processed summary and translation, as well as sentiment analysis results, to the user.

[0372] The present invention is a system that summarizes text written by a user in a natural language and then translates the summary into a different language, incorporating a process for recognizing emotions from the user's text. This system is particularly effective in multilingual content distribution services.

[0373] First, the server receives natural language text from the user. The received text is then sent to a sentiment analysis unit to identify the sentiment within the text. A pre-trained natural language processing model, such as the BERT model, is used for sentiment analysis. The analyzed sentiment is reflected in the subsequent summarization and translation process.

[0374] The server then uses a pre-trained natural language processing model, such as the BART model, to summarize the text entered by the user. This summarization process extracts important information from long pieces of text and summarizes it in a concise form.

[0375] The summarized text is then translated into different languages ​​using a third-party translation API, such as Google Translate.

[0376] The translation result is adjusted as needed, taking into account the results of sentiment analysis, so that the translation provided to the user appropriately reflects the sentiment of the input text.

[0377] Finally, the server returns the adjusted summary and translation results to the user. Through this process, the user can efficiently summarize long texts and obtain translation results that appropriately reflect the sentiment of the text.

[0378] Hardware and software used

[0379] Hardware: Server (AWS (registered trademark), GCP, etc.)

[0380] software:

[0381] Python

[0382] Transformers library (BART, BERT models)

[0383] Translation API (e.g., Google Translate API)

[0384] Specific examples

[0385] Suppose the user types a long sentence like this:

[0386] “Artificial intelligence has seen rapid advancements in recent years. These advancements are evident in various fields such as healthcare, finance, and education. The implementation of AI systems has led to more efficient processes, better decision-making, and significant cost savings. However, ethical concerns remain, particularly in terms of data privacy and algorithmic bias.”

[0387] The server receives this text and first uses an emotion engine to recognize emotions such as:

[0388] "Positive Emotions: Progress, Efficiency, and Cost Reduction"

[0389] "Negative Emotions: Ethical Concerns, Data Privacy, and Algorithmic Bias"

[0390] Then, generate a summary using the pre-trained model:

[0391] “Artificial intelligence has seen rapid advancements in recent years, improving efficiency and decision-making in various fields such as healthcare, finance, and education. However, ethical concerns like data privacy and bias remain.”

[0392] Translate this summary into a different language, for example Japanese:

[0393] "Artificial intelligence has advanced rapidly in recent years, improving efficiency and decision-making in sectors such as healthcare, finance and education. However, ethical concerns such as data privacy and bias remain."

[0394] Example prompt sentence:

[0395] "Summarize a text about recent advances in AI technology, translate it into Japanese, and perform sentiment analysis."

[0396] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0397] Step 1:

[0398] The user inputs text in a natural language and sends it from the device to the server. An example of input text could be a long sentence such as "Artificial intelligence has seen rapid advancements in recent years." The device then sends this long text to the server as a POST request.

[0399] Step 2:

[0400] The server inputs the received text into an emotion recognition engine, which uses a pre-trained BERT model to analyze the sentiment within this text. In this process, keywords and phrases are extracted from the text and classified as positive, negative, or neutral. For example, "progress" is recognized as positive, and "ethical concerns" as negative.

[0401] Step 3:

[0402] The server then sends the input text to a text summarization engine, which uses a pre-trained BART model to convert long text into concise summaries. Specifically, it preserves the meaning of the text while extracting only the important information and summarizing it into short sentences. The input to this process is long text, and the output is a summary.

[0403] Step 4:

[0404] The server sends the summarized text to a translation engine, which uses a third-party translation API (e.g., Google Translate) to translate the summarized text into the specified language. The input to this process is the summarized text, and the output is the translated text. For example, the summary sentence "Artificial intelligence has made rapid advancements." is translated to "Artificial intelligence has made rapid advancements."

[0405] Step 5:

[0406] The server adjusts the translation results based on the results of sentiment analysis. Based on the results of sentiment analysis, the summary and translation results are adjusted to emphasize specific emotions. The input to this process is the sentiment analysis results and the translated text, and the output is the adjusted translation result. For example, the sentence may be restructured to emphasize parts with positive sentiment.

[0407] Step 6:

[0408] The server returns the final result to the user. By returning the adjusted translation result, the user can get a proper summary and translation result based on the input text, which makes it easier for users to understand long sentences and grasp the nuances of emotions in different languages.

[0409] The specific processing unit 290 transmits the result of the specific processing to the smart device 14. In the smart device 14, the control unit 46A causes the output device 40 to output the result of the specific processing. The microphone 38B acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[0410] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (registered trademark) (Internet search engine).<URL: https: / / openai.com / blog / chatgpt> ), Gemini (registered trademark) (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0411] In the above embodiment, an example in which the specific process is performed by the data processing device 12 has been given, but the technology of the present disclosure is not limited to this, and the specific process may be performed by the smart device 14.

[0412] [Second embodiment]

[0413] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0414] 3, the data processing system 210 includes the data processing device 12 and smart glasses 214. An example of the data processing device 12 is a server.

[0415] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0416] The smart glasses 214 include a computer 36, a microphone 238, a speaker 240, a camera 42, and a communication I / F 44. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, and the camera 42 are also connected to the bus 52.

[0417] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[0418] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[0419] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[0420] Fig. 4 shows an example of the main functions of the data processing device 12 and the smart glasses 214. As shown in Fig. 4, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[0421] The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0422] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0423] In the smart glasses 214, the processor 46 performs the reception output process. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0424] Next, a description will be given of the identification process performed by the identification processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as the "server" and the smart glasses 214 will be referred to as the "terminal."

[0425] The present invention provides a system that summarizes text written in a natural language by a user and then translates the summary into a different language. This system receives text entered by a user, summarizes it briefly, and automatically translates it into other languages, such as Japanese, thereby supporting efficient understanding of information and communication between multiple languages.

[0426] Program processing

[0427] The program for this system is as follows: The server summarizes text using a pre-trained natural language processing model (e.g., the BART model), and uses a translation API provided by a third party (e.g., Google Translate) for translation.

[0428] First, a user accesses the server using a terminal and inputs a long piece of text. This text is sent to the server as a POST request. The server receives the text and inputs it into a pre-trained model to automatically generate a summary.

[0429] The server then inputs the generated summary into a third-party translation API, which translates the summarized text into the desired language (e.g., Japanese). Once the translation is complete, the server returns the summary and translation results to the device as a JSON response. The user can view the summary and translation results on the device screen.

[0430] Specific examples

[0431] Here's a concrete example: Suppose the user types the following long sentence:

[0432] “Artificial intelligence has seen rapid advancements in recent years. These advancements are evident in various fields such as healthcare, finance, and education. The implementation of AI systems has led to more efficient processes, better decision-making, and significant cost savings. However, ethical concerns remain, particularly in terms of data privacy and algorithmic bias.”

[0433] When the terminal sends this text to the server, the server generates a summary like this:

[0434] “Artificial intelligence has seen rapid advancements in recent years, improving efficiency and decision-making in various fields such as healthcare, finance, and education. However, ethical concerns like data privacy and bias remain.”

[0435] The server then translates this summary into a different language, for example Japanese, to produce the final result:

[0436] "Artificial intelligence has advanced rapidly in recent years, improving efficiency and decision-making in sectors such as healthcare, finance and education. However, ethical concerns such as data privacy and bias remain."

[0437] In this way, the present invention realizes a series of processes for summarizing text entered by a user and then translating the summary into a different language. The embodiment of the invention embodies a series of processes including multiple steps, such as data exchange between a server and a terminal, summarizing the text, and translating it.

[0438] The processing flow will be explained below.

[0439] Step 1:

[0440] A user uses a terminal to input a long text in a natural language into a text input form.

[0441] Example: "Artificial intelligence has seen rapid advancements in recent years. ..."

[0442] Step 2:

[0443] The device sends the entered text to the server as a POST request, with the request body containing the text in JSON format.

[0444] Example: {"text": "Artificial intelligence has seen rapid advancements in recent years. ..."}

[0445] Step 3:

[0446] The server receives the POST request, extracts the JSON data from the request body, and uses request.json to get the text field.

[0447] Example: text = data['text']

[0448] Step 4:

[0449] The server tokenizes the received text using a pre-trained natural language processing model (BART model). The tokenizer.encode method is used to convert the text into tokens for summarization.

[0450] Example: inputs = tokenizer.encode("summarize: " + text, return_tensors='pt')

[0451] Step 5:

[0452] The server passes the tokenized input to a natural language processing model to generate a summary of the text. The summary result is obtained using the model.generate method.

[0453] Example: summary_ids = model.generate(inputs, max_length=150, min_length=40, ...)

[0454] Step 6:

[0455] The server decodes the generated summary and converts it into a natural language form that the user can understand. Use the tokenizer.decode method to obtain the summary text.

[0456] Example: summary = tokenizer.decode(summary_ids[0], skip_special_tokens=True)

[0457] Step 7:

[0458] The server inputs the summarized text into a third-party translation API (Google Translate API) to translate it into a different language, and retrieves the translation result using the translator.translate method.

[0459] Example: translation = translator.translate(summary, src='en', dest='ja').text

[0460] Step 8:

[0461] The server compiles the summary and translation results into JSON format and generates a response. The jsonify function is used to create the response.

[0462] Example: return jsonify({'summary': summary, 'translation': translation})

[0463] Step 9:

[0464] The server sends a JSON response back to the device, which receives the response and displays the summary and translation results to the user.

[0465] Example: {"summary": "Artificial intelligence has seen rapid advancements...", "translation": "Artificial intelligence has seen rapid advancements..."}

[0466] This completes the process of efficiently summarizing user-entered text and translating the summary into different languages.

[0467] Example 1

[0468] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0469] In today's world, the exchange and understanding of information across multiple languages ​​is extremely important. However, it takes a great deal of time and effort for users to understand long texts written in languages ​​other than their native language. Furthermore, there are limited technologies that can automatically summarize texts and appropriately translate them without specialized knowledge. This makes it difficult for users to easily exchange information across multiple languages, hindering efficient communication.

[0470] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[0471] In this invention, the server includes means for receiving text written in a natural language from a user, means for using a pre-trained generative model to automatically summarize the text, means for using a translation service to translate the summarized text into a different language, and means for returning the translation result to the user's terminal, thereby enabling the user to easily summarize long texts and further translate them into different languages, thereby enabling efficient information comprehension and smooth communication between multiple languages.

[0472] "Text written in a natural language" refers to text written in a language that humans use on a daily basis.

[0473] "User" refers to a person or organization that uses the system.

[0474] A "server" refers to a computer that provides services to other computers over a network.

[0475] "Terminal" refers to the computer or device through which a user interacts with a server.

[0476] A "summary" is a concise description of the main information extracted from a longer original text.

[0477] A "pre-trained generative model" refers to a model that has been trained in advance on a large amount of data and can automatically perform a specific task (in this case, summarization).

[0478] "Translation" refers to the conversion of text written in one language into a different language.

[0479] "Translation Service" means an online application or API that provides text translation.

[0480] "Translation Result" means text that has been converted into a different language by a Translation Service.

[0481] A "pre-trained natural language processing model" refers to a model that has been trained on a large amount of text data in advance to perform natural language processing tasks.

[0482] "Translation API" refers to an application programming interface provided to enable translation functionality to be used from any program.

[0483] "JSON format" refers to the data exchange format called JavaScript Object Notation.

[0484] This invention is a system that automatically summarizes natural language text entered by a user and translates the summary into a different language. The system has the following main functions:

[0485] 1. User enters text

[0486] A user inputs long pieces of text through a web browser or application.

[0487] 2. Sending text from the device to the server

[0488] The device sends the text entered by the user to the server using an HTTP POST request.

[0489] 3. Server-generated text summarization

[0490] The server uses a pre-trained generative model (e.g., the BART model) for text summarization, which summarizes long texts into short, concise summaries.

[0491] 4. Server-Based Abstract Translation

[0492] The summarized text is sent via the server to a translation service (e.g., Google Translate API) and translated into the desired language.

[0493] 5. The server returns the translation results

[0494] The server returns the translated text in JSON format to the device, where the user can view a summary and translation results.

[0495] As a concrete example, consider the case where a user enters the following long sentence:

[0496] Artificial intelligence has seen rapid advancements in recent years. These advancements are evident in various fields such as healthcare, finance, and education. The implementation of AI systems has led to more efficient processes, better decision-making, and significant cost savings. However, ethical concerns remain, particularly in terms of data privacy and algorithmic bias.

[0497] When the device sends this text to the server, the server generates a summary like this:

[0498] Artificial intelligence has seen rapid advancements in recent years, improving efficiency and decision-making in various fields such as healthcare, finance, and education. However, ethical concerns like data privacy and bias remain.

[0499] The server then translates this summary into a different language, for example Japanese:

[0500] Artificial intelligence has advanced rapidly in recent years, improving efficiency and decision-making in fields such as medicine, finance, and education, but ethical concerns such as data privacy and bias remain.

[0501] In this way, users can easily summarize long texts they input and then translate the summaries into different languages. This system allows users to efficiently exchange information between multiple languages.

[0502] Prompt Sentence Examples

[0503] Summarize the following long passage and translate it into Japanese:

[0504] "Artificial intelligence has seen rapid advancements in recent years. These advancements are evident in various fields such as healthcare, finance, and education. The implementation of AI systems has led to more efficient processes, better decision-making, and significant cost savings. However, ethical concerns remain, particularly in terms of data privacy and algorithmic bias."

[0505] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0506] System program processing flow

[0507] Step 1: User Enters Text

[0508] Specific description:

[0509] A user inputs a long piece of natural language text into an input field in a web browser or application on his or her device.

[0510] Specific behavior:

[0511] A user enters the following text into a browser input field:

[0512] Artificial intelligence has seen rapid advancements in recent years. These advancements are evident in various fields such as healthcare, finance, and education. The implementation of AI systems has led to more efficient processes, better decision-making, and significant cost savings. However, ethical concerns remain, particularly in terms of data privacy and algorithmic bias.

[0513] Step 2: Send text to the server using the device

[0514] Specific description:

[0515] The terminal receives the text entered by the user and sends it to the server as a POST request.

[0516] Input and Output:

[0517] Input: Long text entered by the user

[0518] Output: HTTP POST request sent to the server

[0519] Specific behavior:

[0520] The device sends the following HTTP POST request to the server:

[0521] POST / summarize_translate

[0522] Content-Type: application / json

[0523] {

[0524] "text": "Artificial intelligence has seen rapid advancements in recent years. These advancements are evident in various fields such as healthcare, finance, and education. The implementation of AI systems has led to more efficient processes, better decision-making, and significant cost savings. However, ethical concerns remain, particularly in terms of data privacy and algorithmic bias."

[0525] }

[0526] Step 3: Server-generated text summarization

[0527] Specific description:

[0528] The server inputs the received text into a pre-trained generative model (the BART model) and automatically generates a summary.

[0529] Input and Output:

[0530] Input: Long text sent from the terminal

[0531] Output: A short summary of the text

[0532] Specific behavior:

[0533] The server invokes the BART model and generates a summary like this:

[0534] Artificial intelligence has seen rapid advancements in recent years, improving efficiency and decision-making in various fields such as healthcare, finance, and education. However, ethical concerns like data privacy and bias remain.

[0535] Step 4: Server translation of the summary

[0536] Specific description:

[0537] The server sends the generated summary to a translation service (Google Translate API) for translation into the desired language.

[0538] Input and Output:

[0539] Input: A short text to be summarized

[0540] Output: The translated text

[0541] Specific behavior:

[0542] The server calls the Google Translate API and obtains the translation result as follows:

[0543] Artificial intelligence has advanced rapidly in recent years, improving efficiency and decision-making in fields such as medicine, finance, and education, but ethical concerns such as data privacy and bias remain.

[0544] Step 5: The server returns the translation results to the device

[0545] Specific description:

[0546] The server returns the translated text and a summary to the device in JSON format.

[0547] Input and Output:

[0548] Input: translated text and summary

[0549] Output: JSON response sent back to the device

[0550] Specific behavior:

[0551] The server sends the following JSON response to the device:

[0552] {

[0553] "summary": "Artificial intelligence has seen rapid advancements in recent years, improving efficiency and decision-making in various fields such as healthcare, finance, and education. However, ethical concerns like data privacy and bias remain.",

[0554] "translation": "Artificial intelligence has advanced rapidly in recent years, improving efficiency and decision-making in sectors like healthcare, finance, and education. However, ethical concerns such as data privacy and bias remain."

[0555] }

[0556] Step 6: User summarization and translation confirmation

[0557] Specific description:

[0558] The user checks the returned summary and translation results on the screen of the terminal.

[0559] Input and Output:

[0560] Input: JSON response returned from the server

[0561] Output: Summary and translation results displayed on the terminal

[0562] Specific behavior:

[0563] The user sees the summary and translation results in the browser or application screen, as shown below.

[0564] summary:

[0565] Artificial intelligence has seen rapid advancements in recent years, improving efficiency and decision-making in various fields such as healthcare, finance, and education. However, ethical concerns like data privacy and bias remain.

[0566] translation:

[0567] Artificial intelligence has advanced rapidly in recent years, improving efficiency and decision-making in fields such as medicine, finance, and education, but ethical concerns such as data privacy and bias remain.

[0568] (Application example 1)

[0569] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0570] On modern e-commerce sites, reviews posted by users are written in a wide variety of languages, making it difficult for multilingual users to understand the content of the reviews. Furthermore, reading the entire lengthy review is inefficient, and summaries are required. The present invention addresses this issue.

[0571] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[0572] In this invention, the server includes means for receiving natural language text from a user, means for using a pre-trained model to summarize the text, means for translating the summarized text into a different language, means for applying the summarized text and the translation result to user reviews on an e-commerce site, and means for returning the translation result to the user, thereby enabling multilingual users to efficiently understand the review content.

[0573] A "user" is a user of the system, and is an individual or organization that inputs text in a natural language.

[0574] "Natural language" refers to a language that humans use on a daily basis and that does not have a specific structure or form.

[0575] "Text" is character string information written in a natural language.

[0576] "Pre-trained model" refers to a machine learning model that has been pre-trained with a large amount of reference data for the purpose of natural language processing.

[0577] A "summary" is a short sentence that succinctly expresses the main points of a long text.

[0578] "Translation" is the act of converting text written in one language into another language.

[0579] "Different language" refers to a language that is different from the natural language input by the user.

[0580] "Translation result" refers to the output of the translated text.

[0581] An "e-commerce site" is a website that sells goods and services over the Internet.

[0582] "User reviews" refer to texts in which users express their opinions or evaluations of products or services they have purchased.

[0583] "Replying" refers to the act of providing an appropriate output in response to an input from a user.

[0584] "Means" refers to elements necessary to achieve a specific function.

[0585] The present invention is a system that automatically summarizes natural language text entered by a user, translates the summary into different languages, and applies it to user reviews on e-commerce sites.

[0586] The system includes the following means:

[0587] 1. A means for users to input natural language text

[0588] 2. Using a pre-trained model to summarize input text

[0589] 3. A means of translating the summarized text into different languages

[0590] 4. A method for applying summarized text and translation results to user reviews on e-commerce sites

[0591] 5. Means of returning translation results to the user

[0592] Hardware and software used

[0593] Hardware: Server, user devices (smartphones, PCs)

[0594] Software: BART model (summarization), Google Translate API (translation), web service (RESTful API), front-end (React, Angular, etc.)

[0595] Processing procedures and examples

[0596] 1. Processing when posting a review

[0597] A user submits a review on an e-commerce site, and the text is sent to the server as a POST request.

[0598] 2. Summary Generation

[0599] The server inputs the received reviews into a pre-trained BART model, a machine learning model pre-trained on large amounts of data for the purpose of natural language processing, to generate summaries.

[0600] 3. Translation

[0601] The server inputs the summarized text into the Google Translate API, which is a third-party web API that provides translation services, and translates it into a different specified language (e.g., Japanese).

[0602] 4. Application

[0603] The server applies the summarized text and translation results to user reviews on an e-commerce site, where the translated summaries are displayed in the review section.

[0604] 5. Response Generation

[0605] The server stores the summary and translation results in a cloud database and returns them to the user's device in JSON format, where the user can view the summary and translation results in real time.

[0606] Specific examples

[0607] For example, if a user posts a review like this:

[0608] “Artificial intelligence has seen rapid advancements in recent years. These advancements are evident in various fields such as healthcare, finance, and education. The implementation of AI systems has led to more efficient processes, better decision-making, and significant cost savings. However, ethical concerns remain, particularly in terms of data privacy and algorithmic bias.”

[0609] The system generates a summary like this, which is then translated into Japanese:

[0610] Abstract: "Artificial intelligence has seen rapid advancements in recent years, improving efficiency and decision-making in various fields such as healthcare, finance, and education. However, concerns like data privacy and ethical bias remain."

[0611] Translation result: "Artificial intelligence has advanced rapidly in recent years, improving efficiency and decision-making in fields such as medicine, finance, and education. However, ethical concerns such as data privacy and bias remain."

[0612] In this way, multilingual users can efficiently understand the review content.

[0613] Prompt Sentence Examples

[0614] "Please summarize the long review below to a maximum of 150 characters and then translate it into Japanese. Review: {Review content}"

[0615] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0616] Step 1:

[0617] A user submits a review by entering natural language text into an e-commerce site. This text is sent from the user's device to the server as a POST request. The input text is a string of natural language characters entered by the user as a review. When the user presses the submit button for the text they entered, the text data is sent to the server in JSON format.

[0618] Step 2:

[0619] The server detects the review text received from the user and inputs it into a pre-trained generative AI model (e.g., the BART model) to summarize it. The input text data is the natural language review text received from the user. The server inputs this text data into the BART model, which summarizes the text based on important information. This results in a shorter summary being output.

[0620] Step 3:

[0621] The server inputs the generated summary into a translation API (for example, Google Translate API) to translate it into a different specified language. The input data is the summary and the language information for the translation. The server sends the summary to the translation API and receives the translation result. As a result, the summary translated into the specified language is output.

[0622] Step 4:

[0623] The server applies the translated summary to user reviews on the e-commerce site. The input data is the translated summary. The server inserts the summary into the review section, making it visible to other users. This step allows the summarized translation to be displayed in the review section.

[0624] Step 5:

[0625] The server stores the summary and translation results in a cloud database and returns them to the user's device in JSON format. The input data is the summary and translation. The server stores this data in a cloud database and sends it to the user's device as a response. The user can check the summary and translation results on their device.

[0626] The above processing steps enable users who speak different languages ​​to efficiently understand the review content.

[0627] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[0628] The present invention combines a system that summarizes text written by a user in a natural language and then translates the summary into a different language with an emotion engine that recognizes emotions from the user's text. The system receives text entered by a user, summarizes it, translates the summary into another language, and recognizes emotions from the text and reflects them in the results.

[0629] Program processing

[0630] The program for this system is as follows. The server summarizes text using a pre-trained natural language processing model (e.g., the BART model). It also uses a translation API provided by a third party (e.g., Google Translate) for translation and an emotion engine for emotion recognition.

[0631] First, a user accesses the server using a terminal and inputs a long piece of text. This text is sent to the server as a POST request. The server receives the text and recognizes emotions using an emotion engine. The recognized emotions influence the subsequent summarization and translation processes.

[0632] The server then inputs the text into a pre-trained model to automatically generate a summary. The generated summary is then input into a third-party translation API for translation into the desired language (e.g., Japanese). Based on the output of the emotion engine, the server adjusts the summary and translation results to optimally respond to the user.

[0633] Specific examples

[0634] Here's a concrete example: Suppose the user types the following long sentence:

[0635] “Artificial intelligence has seen rapid advancements in recent years. These advancements are evident in various fields such as healthcare, finance, and education. The implementation of AI systems has led to more efficient processes, better decision-making, and significant cost savings. However, ethical concerns remain, particularly in terms of data privacy and algorithmic bias.”

[0636] When the device sends this text to the server, the server first uses its emotion engine to recognize emotions such as:

[0637] "Positive emotions: progress, efficiency, cost savings

[0638] Negative Emotions: Ethical Concerns, Data Privacy, and Algorithmic Bias.”

[0639] The server then uses this information to generate a summary:

[0640] “Artificial intelligence has seen rapid advancements in recent years, improving efficiency and decision-making in various fields such as healthcare, finance, and education. However, ethical concerns like data privacy and bias remain.”

[0641] The server then translates this summary into a different language, say Japanese, producing the final result:

[0642] "Artificial intelligence has advanced rapidly in recent years, improving efficiency and decision-making in sectors such as healthcare, finance and education. However, ethical concerns such as data privacy and bias remain."

[0643] The translation takes into account the output of the emotion engine and makes appropriate adjustments, such as emphasizing key points and appropriately reflecting positive and negative sentiment.

[0644] This enables a series of processes to efficiently summarize text entered by a user, translate that summary into different languages, and simultaneously recognize and appropriately reflect the sentiment of the text. This system, combined with an emotion engine, can enrich multilingual communication and more accurately meet user needs.

[0645] The processing flow will be explained below.

[0646] Step 1:

[0647] A user uses a terminal to input a long text in a natural language into a text input form.

[0648] Example: "Artificial intelligence has seen rapid advancements in recent years. ..."

[0649] Step 2:

[0650] The device sends the entered text to the server as a POST request, with the request body containing the text in JSON format.

[0651] Example: {"text": "Artificial intelligence has seen rapid advancements in recent years. ..."}

[0652] Step 3:

[0653] The server receives the POST request, extracts the JSON data from the request body, and uses request.json to get the text field.

[0654] Example: text = data['text']

[0655] Step 4:

[0656] The server uses an emotion engine to recognize the emotion of the received text, a process that classifies emotions such as positive, negative, or neutral based on the content of the text.

[0657] Example: emotion = emotion_engine.analyze(text)

[0658] Step 5:

[0659] The server tokenizes the received text using a pre-trained natural language processing model (BART model). The tokenizer.encode method is used to convert the text into tokens for summarization.

[0660] Example: inputs = tokenizer.encode("summarize: " + text, return_tensors='pt')

[0661] Step 6:

[0662] The server passes the tokenized input to a natural language processing model to generate a summary of the text. The summary result is obtained using the model.generate method.

[0663] Example: summary_ids = model.generate(inputs, max_length=150, min_length=40, ...)

[0664] Step 7:

[0665] The server decodes the generated summary and converts it into a natural language form that the user can understand. Use the tokenizer.decode method to obtain the summary text.

[0666] Example: summary = tokenizer.decode(summary_ids[0], skip_special_tokens=True)

[0667] Step 8:

[0668] The server inputs the summarized text into a third-party translation API (Google Translate API) to translate it into a different language, and retrieves the translation result using the translator.translate method.

[0669] Example: translation = translator.translate(summary, src='en', dest='ja').text

[0670] Step 9:

[0671] The server adjusts the summarization and translation results based on the output of the emotion engine, for example by changing the wording to emphasize positive emotions.

[0672] Example: adjusted_translation = adjust_based_on_emotion(translation, emotion)

[0673] Step 10:

[0674] The server generates a response containing the summary and adjusted translation results in JSON format. The response is created using the jsonify function.

[0675] Example: return jsonify({'summary': summary, 'translation': adjusted_translation})

[0676] Step 11:

[0677] The server sends a JSON response back to the device, which receives the response and displays the summary and translation results to the user.

[0678] Example: {"summary": "Artificial intelligence has seen rapid advancements...", "translation": "Artificial intelligence has seen rapid advancements..."}

[0679] This allows for a system that efficiently summarizes user-entered text, translates the summary into different languages, and recognizes the sentiment of the text and reflects it in the results.

[0680] Example 2

[0681] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0682] Many conventional natural language processing systems are unable to recognize and reflect the sentiment of text when summarizing or translating it. This can result in summaries and translations that do not adequately reflect the user's intentions and sentiments, potentially reducing the quality of communication. Furthermore, the lack of automatic adjustments to translation results increases the user's workload.

[0683] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0684] In this invention, the server includes means for receiving natural language text from a user, means for recognizing emotions from the text, means for using a pre-trained model to summarize the text, means for translating the summarized text into a different language, means for adjusting the translation result based on the recognized emotions, and means for returning the translation result to the user, thereby making it possible to provide a summary and translation result that appropriately reflects the user's intentions and emotions.

[0685] A "user" is a person or entity that utilizes the system to input natural language text and receive summaries and translation results.

[0686] "Natural language text" is a sentence or text written in a normal human language.

[0687] "Means for recognizing emotions" refers to means that use engines or algorithms, such as sentiment analysis models, to analyze and identify emotions within text.

[0688] A "pre-trained model" is an artificial intelligence model that has been trained in advance based on a large amount of text data to perform a specific task (in this case, summarization) with high accuracy.

[0689] A "means of translation" is a means of using technology or services, such as a translation API, to convert text written in one language into another language.

[0690] "Means for responding to the user" refers to the technology or service that sends the processing results to the user's terminal so that the user can view them.

[0691] A "server" is a computer system that receives requests from users, performs various processes internally, and returns the final results.

[0692] The present invention combines a system that summarizes text written by a user in a natural language and then translates the summary into a different language with an emotion engine that recognizes emotions from the user's text. The system receives text entered by a user, summarizes it, translates the summary into another language, and recognizes emotions from the text and reflects them in the results.

[0693] Program processing

[0694] The processing of the system program is explained below.

[0695] The server uses a pre-trained natural language processing model (e.g., the BART model) to summarize the text, a third-party translation API (e.g., Google Translate) for translation, and an emotion engine for emotion recognition.

[0696] First, a user accesses the server using a terminal and inputs a long piece of text. This text is sent to the server as a POST request. The server receives the text and uses an emotion engine to recognize emotions. The recognized emotions influence the subsequent summarization and translation processes.

[0697] The server then inputs the text into a pre-trained model to automatically generate a summary. The generated summary is then input into a third-party translation API for translation into the desired language (e.g., Japanese). Based on the output of the emotion engine, the server adjusts the summary and translation results to optimally respond to the user.

[0698] Specific examples

[0699] Here's a concrete example: Suppose the user types the following long sentence:

[0700] “Artificial intelligence has seen rapid advancements in recent years. These advancements are evident in various fields such as healthcare, finance, and education. The implementation of AI systems has led to more efficient processes, better decision-making, and significant cost savings. However, ethical concerns remain, particularly in terms of data privacy and algorithmic bias.”

[0701] When the device sends this text to the server, the server first uses its emotion engine to recognize emotions such as:

[0702] "Positive emotions: progress, efficiency, cost savings

[0703] Negative Emotions: Ethical Concerns, Data Privacy, and Algorithmic Bias.”

[0704] The server then uses this information to generate a summary:

[0705] “Artificial intelligence has seen rapid advancements in recent years, improving efficiency and decision-making in various fields such as healthcare, finance, and education. However, ethical concerns like data privacy and bias remain.”

[0706] The server then translates this summary into a different language, say Japanese, producing the final result:

[0707] "Artificial intelligence has advanced rapidly in recent years, improving efficiency and decision-making in sectors such as healthcare, finance and education. However, ethical concerns such as data privacy and bias remain."

[0708] The translation takes into account the output of the emotion engine and makes appropriate adjustments, such as emphasizing key points and appropriately reflecting positive and negative sentiment.

[0709] This enables a series of processes to efficiently summarize text entered by a user, translate that summary into different languages, and simultaneously recognize and appropriately reflect the sentiment of the text. This system, combined with an emotion engine, can enrich multilingual communication and more accurately meet user needs.

[0710] Prompt Sentence Examples

[0711] An example of a prompt to use when a user has entered a long text and wants a summary of that text is:

[0712] "Summarize the following long English sentence and translate it into Japanese. Also, please specify the emotion that is recognized."

[0713] Then, here's an example of the specific text the user would enter:

[0714] “Artificial intelligence has seen rapid advancements in recent years. These advancements are evident in various fields such as healthcare, finance, and education. The implementation of AI systems has led to more efficient processes, better decision-making, and significant cost savings. However, ethical concerns remain, particularly in terms of data privacy and algorithmic bias.”

[0715] This provides clear instructions for the system to perform the expected functions accurately.

[0716] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0717] Step 1:

[0718] Entering and sending text

[0719] First, a user accesses the system using a terminal and inputs a long piece of natural language text. When the user inputs the text and clicks the "Submit" button, the terminal sends the input text to the server as a POST request. Specifically, the text is input as a browser form, and JavaScript processes it and sends it to the server.

[0720] Input: Natural language text entered by the user

[0721] Output: POST request from the terminal to the server

[0722] Step 2:

[0723] Text Receiving and Emotion Recognition

[0724] The server receives the POST request sent from the device and passes the received text to an emotion engine (e.g., an emotion analysis model) to analyze the type and intensity of emotion. This analysis process identifies positive and negative emotions contained in the text.

[0725] Input: The user's natural language text included in the POST request

[0726] Output: Sentiment analysis results (positive and negative) of the text

[0727] Step 3:

[0728] Summary Generation

[0729] The server inputs the emotion-recognized text into a pre-trained natural language processing model (e.g., the BART model). The model analyzes the text, extracts key information, and generates a summary. The summary is generated quickly and reduces the content while preserving the essence of the original text.

[0730] Input: Natural language text with sentiment information

[0731] Output: Generated summary text

[0732] Step 4:

[0733] Executing the translation

[0734] The server sends the generated summary text to a translation API provided by a third party (e.g., a translation service) for translation into the desired language. The translation API connects to the external service via an HTTP request and returns the translated text.

[0735] Input: Summarized text

[0736] Output: The translated text

[0737] Step 5:

[0738] Adjusting the translation results

[0739] The server adjusts the translation results based on the output of the emotion engine, specifically by emphasizing key points and modifying the translation to appropriately reflect positive and negative emotions, so that the information is conveyed in the most optimal way for the user.

[0740] Input: Sentiment analysis results and translated text

[0741] Output: Sentiment-informed translated text

[0742] Step 6:

[0743] Returning the results

[0744] Finally, the server sends the adjusted translation back to the user's device. The server generates an HTTP response and sends the text in HTML or JSON format to the user's device, where the user can view the translation results.

[0745] Input: Adjusted translation text

[0746] Output: Translation results displayed on the user's device

[0747] Through these steps, users can efficiently summarize natural language text, translate the summary into different languages, and even recognize emotions to achieve optimally reflected results.

[0748] (Application example 2)

[0749] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0750] The present invention relates to a system that efficiently performs summarization, translation, and emotion recognition for long natural language texts entered by users, and aims to provide appropriate summarization and translation results, particularly in multilingual content distribution services. Furthermore, because conventional technologies have been insufficient in recognizing emotions and reflecting the results, there is a need for a system that can provide more natural and appropriate output by recognizing emotions in text.

[0751] The identification process by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for receiving natural language text from a user, means for using a pre-trained model to summarize the text, means for translating the summarized text into a different language, means for analyzing emotions extracted from the text, means for adjusting the summary and translation result based on the emotions, and means for returning the adjusted translation result to the user. This allows the user to obtain a translation result in which the input text is efficiently summarized and appropriate emotions are reflected.

[0752] "User" means any person or entity that utilizes the System to enter text and receive results.

[0753] A "natural language" is a language that humans use on a daily basis (e.g., Japanese, English, etc.).

[0754] "Text" refers to natural language sentences entered by a user.

[0755] A "pre-trained model" is a machine learning model for natural language processing that has been trained in advance using a large amount of data.

[0756] "Summarization" is the process of extracting important information from a longer text and summarizing it in a short form.

[0757] "Translation" is the process of converting text written in one language into a different language.

[0758] "Emotion" refers to a user's feelings or emotional state, such as positive or negative, extracted from text.

[0759] "Analyzing sentiment" is the process of recognizing sentiment in text and determining whether it is positive, negative, or neutral.

[0760] "Adjustment" refers to optimizing the summary and translation results by reflecting the results of sentiment analysis.

[0761] "Returning results" means providing the processed summary and translation, as well as sentiment analysis results, to the user.

[0762] The present invention is a system that summarizes text written by a user in a natural language and then translates the summary into a different language, incorporating a process for recognizing emotions from the user's text. This system is particularly effective in multilingual content distribution services.

[0763] First, the server receives natural language text from the user. The received text is then sent to a sentiment analysis unit to identify the sentiment within the text. A pre-trained natural language processing model, such as the BERT model, is used for sentiment analysis. The analyzed sentiment is reflected in the subsequent summarization and translation process.

[0764] The server then uses a pre-trained natural language processing model, such as the BART model, to summarize the text entered by the user. This summarization process extracts important information from long pieces of text and summarizes it in a concise form.

[0765] The summarized text is then translated into different languages ​​using a third-party translation API, such as Google Translate.

[0766] The translation result is adjusted as needed, taking into account the results of sentiment analysis, so that the translation provided to the user appropriately reflects the sentiment of the input text.

[0767] Finally, the server returns the adjusted summary and translation results to the user. Through this process, the user can efficiently summarize long texts and obtain translation results that appropriately reflect the sentiment of the text.

[0768] Hardware and software used

[0769] Hardware: Server (AWS, GCP, etc.)

[0770] software:

[0771] Python

[0772] Transformers library (BART, BERT models)

[0773] Translation API (e.g., Google Translate API)

[0774] Specific examples

[0775] Suppose the user types a long sentence like this:

[0776] “Artificial intelligence has seen rapid advancements in recent years. These advancements are evident in various fields such as healthcare, finance, and education. The implementation of AI systems has led to more efficient processes, better decision-making, and significant cost savings. However, ethical concerns remain, particularly in terms of data privacy and algorithmic bias.”

[0777] The server receives this text and first uses an emotion engine to recognize emotions such as:

[0778] "Positive Emotions: Progress, Efficiency, and Cost Reduction"

[0779] "Negative Emotions: Ethical Concerns, Data Privacy, and Algorithmic Bias"

[0780] Then, generate a summary using the pre-trained model:

[0781] “Artificial intelligence has seen rapid advancements in recent years, improving efficiency and decision-making in various fields such as healthcare, finance, and education. However, ethical concerns like data privacy and bias remain.”

[0782] Translate this summary into a different language, for example Japanese:

[0783] "Artificial intelligence has advanced rapidly in recent years, improving efficiency and decision-making in sectors such as healthcare, finance and education. However, ethical concerns such as data privacy and bias remain."

[0784] Example prompt sentence:

[0785] "Summarize a text about recent advances in AI technology, translate it into Japanese, and perform sentiment analysis."

[0786] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0787] Step 1:

[0788] The user inputs text in a natural language and sends it from the device to the server. An example of input text could be a long sentence such as "Artificial intelligence has seen rapid advancements in recent years." The device then sends this long text to the server as a POST request.

[0789] Step 2:

[0790] The server inputs the received text into an emotion recognition engine, which uses a pre-trained BERT model to analyze the sentiment within this text. In this process, keywords and phrases are extracted from the text and classified as positive, negative, or neutral. For example, "progress" is recognized as positive, and "ethical concerns" as negative.

[0791] Step 3:

[0792] The server then sends the input text to a text summarization engine, which uses a pre-trained BART model to convert long text into concise summaries. Specifically, it preserves the meaning of the text while extracting only the important information and summarizing it into short sentences. The input to this process is long text, and the output is a summary.

[0793] Step 4:

[0794] The server sends the summarized text to a translation engine, which uses a third-party translation API (e.g., Google Translate) to translate the summarized text into the specified language. The input to this process is the summarized text, and the output is the translated text. For example, the summary sentence "Artificial intelligence has made rapid advancements." is translated to "Artificial intelligence has made rapid advancements."

[0795] Step 5:

[0796] The server adjusts the translation results based on the results of sentiment analysis. Based on the results of sentiment analysis, the summary and translation results are adjusted to emphasize specific emotions. The input to this process is the sentiment analysis results and the translated text, and the output is the adjusted translation result. For example, the sentence may be restructured to emphasize parts with positive sentiment.

[0797] Step 6:

[0798] The server returns the final result to the user. By returning the adjusted translation result, the user can get a proper summary and translation result based on the input text, which makes it easier for users to understand long sentences and grasp the nuances of emotions in different languages.

[0799] The specific processing unit 290 transmits the result of the specific processing to the smart glasses 214. In the smart glasses 214, the control unit 46A causes the speaker 240 to output the result of the specific processing. The microphone 238 acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[0800] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0801] In the above embodiment, an example in which the specific processing is performed by the data processing device 12 has been given, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the smart glasses 214.

[0802] [Third embodiment]

[0803] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[0804] 5, the data processing system 310 includes the data processing device 12 and a headset type terminal 314. An example of the data processing device 12 is a server.

[0805] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0806] The headset type terminal 314 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication I / F 44, and a display 343. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, the camera 42, and the display 343 are also connected to the bus 52.

[0807] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[0808] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[0809] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[0810] Fig. 6 shows an example of the main functions of the data processing device 12 and the headset type terminal 314. As shown in Fig. 6, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[0811] The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0812] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0813] In the headset type terminal 314, a reception output process is performed by the processor 46. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0814] Next, a description will be given of the identification process performed by the identification processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as the "server" and the headset type terminal 314 will be referred to as the "terminal."

[0815] The present invention provides a system that summarizes text written in a natural language by a user and then translates the summary into a different language. This system receives text entered by a user, summarizes it briefly, and automatically translates it into other languages, such as Japanese, thereby supporting efficient understanding of information and communication between multiple languages.

[0816] Program processing

[0817] The program for this system is as follows: The server summarizes text using a pre-trained natural language processing model (e.g., the BART model), and uses a translation API provided by a third party (e.g., Google Translate) for translation.

[0818] First, a user accesses the server using a terminal and inputs a long piece of text. This text is sent to the server as a POST request. The server receives the text and inputs it into a pre-trained model to automatically generate a summary.

[0819] The server then inputs the generated summary into a third-party translation API, which translates the summarized text into the desired language (e.g., Japanese). Once the translation is complete, the server returns the summary and translation results to the device as a JSON response. The user can view the summary and translation results on the device screen.

[0820] Specific examples

[0821] Here's a concrete example: Suppose the user types the following long sentence:

[0822] “Artificial intelligence has seen rapid advancements in recent years. These advancements are evident in various fields such as healthcare, finance, and education. The implementation of AI systems has led to more efficient processes, better decision-making, and significant cost savings. However, ethical concerns remain, particularly in terms of data privacy and algorithmic bias.”

[0823] When the terminal sends this text to the server, the server generates a summary like this:

[0824] “Artificial intelligence has seen rapid advancements in recent years, improving efficiency and decision-making in various fields such as healthcare, finance, and education. However, ethical concerns like data privacy and bias remain.”

[0825] The server then translates this summary into a different language, for example Japanese, to produce the final result:

[0826] "Artificial intelligence has advanced rapidly in recent years, improving efficiency and decision-making in sectors such as healthcare, finance and education. However, ethical concerns such as data privacy and bias remain."

[0827] In this way, the present invention realizes a series of processes for summarizing text entered by a user and then translating the summary into a different language. The embodiment of the invention embodies a series of processes including multiple steps, such as data exchange between a server and a terminal, summarizing the text, and translating it.

[0828] The processing flow will be explained below.

[0829] Step 1:

[0830] A user uses a terminal to input a long text in a natural language into a text input form.

[0831] Example: "Artificial intelligence has seen rapid advancements in recent years. ..."

[0832] Step 2:

[0833] The device sends the entered text to the server as a POST request, with the request body containing the text in JSON format.

[0834] Example: {"text": "Artificial intelligence has seen rapid advancements in recent years. ..."}

[0835] Step 3:

[0836] The server receives the POST request, extracts the JSON data from the request body, and uses request.json to get the text field.

[0837] Example: text = data['text']

[0838] Step 4:

[0839] The server tokenizes the received text using a pre-trained natural language processing model (BART model). The tokenizer.encode method is used to convert the text into tokens for summarization.

[0840] Example: inputs = tokenizer.encode("summarize: " + text, return_tensors='pt')

[0841] Step 5:

[0842] The server passes the tokenized input to a natural language processing model to generate a summary of the text. The summary result is obtained using the model.generate method.

[0843] Example: summary_ids = model.generate(inputs, max_length=150, min_length=40, ...)

[0844] Step 6:

[0845] The server decodes the generated summary and converts it into a natural language form that the user can understand. Use the tokenizer.decode method to obtain the summary text.

[0846] Example: summary = tokenizer.decode(summary_ids[0], skip_special_tokens=True)

[0847] Step 7:

[0848] The server inputs the summarized text into a third-party translation API (Google Translate API) to translate it into a different language, and retrieves the translation result using the translator.translate method.

[0849] Example: translation = translator.translate(summary, src='en', dest='ja').text

[0850] Step 8:

[0851] The server compiles the summary and translation results into JSON format and generates a response. The jsonify function is used to create the response.

[0852] Example: return jsonify({'summary': summary, 'translation': translation})

[0853] Step 9:

[0854] The server sends a JSON response back to the device, which receives the response and displays the summary and translation results to the user.

[0855] Example: {"summary": "Artificial intelligence has seen rapid advancements...", "translation": "Artificial intelligence has seen rapid advancements..."}

[0856] This completes the process of efficiently summarizing user-entered text and translating the summary into different languages.

[0857] Example 1

[0858] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[0859] In today's world, the exchange and understanding of information across multiple languages ​​is extremely important. However, it takes a great deal of time and effort for users to understand long texts written in languages ​​other than their native language. Furthermore, there are limited technologies that can automatically summarize texts and appropriately translate them without specialized knowledge. This makes it difficult for users to easily exchange information across multiple languages, hindering efficient communication.

[0860] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[0861] In this invention, the server includes means for receiving text written in a natural language from a user, means for using a pre-trained generative model to automatically summarize the text, means for using a translation service to translate the summarized text into a different language, and means for returning the translation result to the user's terminal, thereby enabling the user to easily summarize long texts and further translate them into different languages, thereby enabling efficient information comprehension and smooth communication between multiple languages.

[0862] "Text written in a natural language" refers to text written in a language that humans use on a daily basis.

[0863] "User" refers to a person or organization that uses the system.

[0864] A "server" refers to a computer that provides services to other computers over a network.

[0865] "Terminal" refers to the computer or device through which a user interacts with a server.

[0866] A "summary" is a concise description of the main information extracted from a longer original text.

[0867] A "pre-trained generative model" refers to a model that has been trained in advance on a large amount of data and can automatically perform a specific task (in this case, summarization).

[0868] "Translation" refers to the conversion of text written in one language into a different language.

[0869] "Translation Service" means an online application or API that provides text translation.

[0870] "Translation Result" means text that has been converted into a different language by a Translation Service.

[0871] A "pre-trained natural language processing model" refers to a model that has been trained on a large amount of text data in advance to perform natural language processing tasks.

[0872] "Translation API" refers to an application programming interface provided to enable translation functionality to be used from any program.

[0873] "JSON format" refers to the data exchange format called JavaScript Object Notation.

[0874] This invention is a system that automatically summarizes natural language text entered by a user and translates the summary into a different language. The system has the following main functions:

[0875] 1. User enters text

[0876] A user inputs long pieces of text through a web browser or application.

[0877] 2. Sending text from the device to the server

[0878] The device sends the text entered by the user to the server using an HTTP POST request.

[0879] 3. Server-generated text summarization

[0880] The server uses a pre-trained generative model (e.g., the BART model) for text summarization, which summarizes long texts into short, concise summaries.

[0881] 4. Server-Based Abstract Translation

[0882] The summarized text is sent via the server to a translation service (e.g., Google Translate API) and translated into the desired language.

[0883] 5. The server returns the translation results

[0884] The server returns the translated text in JSON format to the device, where the user can view a summary and translation results.

[0885] As a concrete example, consider the case where a user enters the following long sentence:

[0886] Artificial intelligence has seen rapid advancements in recent years. These advancements are evident in various fields such as healthcare, finance, and education. The implementation of AI systems has led to more efficient processes, better decision-making, and significant cost savings. However, ethical concerns remain, particularly in terms of data privacy and algorithmic bias.

[0887] When the device sends this text to the server, the server generates a summary like this:

[0888] Artificial intelligence has seen rapid advancements in recent years, improving efficiency and decision-making in various fields such as healthcare, finance, and education. However, ethical concerns like data privacy and bias remain.

[0889] The server then translates this summary into a different language, for example Japanese:

[0890] Artificial intelligence has advanced rapidly in recent years, improving efficiency and decision-making in fields such as medicine, finance, and education, but ethical concerns such as data privacy and bias remain.

[0891] In this way, users can easily summarize long texts they input and then translate the summaries into different languages. This system allows users to efficiently exchange information between multiple languages.

[0892] Prompt Sentence Examples

[0893] Summarize the following long passage and translate it into Japanese:

[0894] "Artificial intelligence has seen rapid advancements in recent years. These advancements are evident in various fields such as healthcare, finance, and education. The implementation of AI systems has led to more efficient processes, better decision-making, and significant cost savings. However, ethical concerns remain, particularly in terms of data privacy and algorithmic bias."

[0895] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0896] System program processing flow

[0897] Step 1: User Enters Text

[0898] Specific description:

[0899] A user inputs a long piece of natural language text into an input field in a web browser or application on his or her device.

[0900] Specific behavior:

[0901] A user enters the following text into a browser input field:

[0902] Artificial intelligence has seen rapid advancements in recent years. These advancements are evident in various fields such as healthcare, finance, and education. The implementation of AI systems has led to more efficient processes, better decision-making, and significant cost savings. However, ethical concerns remain, particularly in terms of data privacy and algorithmic bias.

[0903] Step 2: Send text to the server using the device

[0904] Specific description:

[0905] The terminal receives the text entered by the user and sends it to the server as a POST request.

[0906] Input and Output:

[0907] Input: Long text entered by the user

[0908] Output: HTTP POST request sent to the server

[0909] Specific behavior:

[0910] The device sends the following HTTP POST request to the server:

[0911] POST / summarize_translate

[0912] Content-Type: application / json

[0913] {

[0914] "text": "Artificial intelligence has seen rapid advancements in recent years. These advancements are evident in various fields such as healthcare, finance, and education. The implementation of AI systems has led to more efficient processes, better decision-making, and significant cost savings. However, ethical concerns remain, particularly in terms of data privacy and algorithmic bias."

[0915] }

[0916] Step 3: Server-generated text summarization

[0917] Specific description:

[0918] The server inputs the received text into a pre-trained generative model (the BART model) and automatically generates a summary.

[0919] Input and Output:

[0920] Input: Long text sent from the terminal

[0921] Output: A short summary of the text

[0922] Specific behavior:

[0923] The server invokes the BART model and generates a summary like this:

[0924] Artificial intelligence has seen rapid advancements in recent years, improving efficiency and decision-making in various fields such as healthcare, finance, and education. However, ethical concerns like data privacy and bias remain.

[0925] Step 4: Server translation of the summary

[0926] Specific description:

[0927] The server sends the generated summary to a translation service (Google Translate API) for translation into the desired language.

[0928] Input and Output:

[0929] Input: A short text to be summarized

[0930] Output: The translated text

[0931] Specific behavior:

[0932] The server calls the Google Translate API and obtains the translation result as follows:

[0933] Artificial intelligence has advanced rapidly in recent years, improving efficiency and decision-making in fields such as medicine, finance, and education, but ethical concerns such as data privacy and bias remain.

[0934] Step 5: The server returns the translation results to the device

[0935] Specific description:

[0936] The server returns the translated text and a summary to the device in JSON format.

[0937] Input and Output:

[0938] Input: translated text and summary

[0939] Output: JSON response sent back to the device

[0940] Specific behavior:

[0941] The server sends the following JSON response to the device:

[0942] {

[0943] "summary": "Artificial intelligence has seen rapid advancements in recent years, improving efficiency and decision-making in various fields such as healthcare, finance, and education. However, ethical concerns like data privacy and bias remain.",

[0944] "translation": "Artificial intelligence has advanced rapidly in recent years, improving efficiency and decision-making in sectors like healthcare, finance, and education. However, ethical concerns such as data privacy and bias remain."

[0945] }

[0946] Step 6: User summarization and translation confirmation

[0947] Specific description:

[0948] The user checks the returned summary and translation results on the screen of the terminal.

[0949] Input and Output:

[0950] Input: JSON response returned from the server

[0951] Output: Summary and translation results displayed on the terminal

[0952] Specific behavior:

[0953] The user sees the summary and translation results in the browser or application screen, as shown below.

[0954] summary:

[0955] Artificial intelligence has seen rapid advancements in recent years, improving efficiency and decision-making in various fields such as healthcare, finance, and education. However, ethical concerns like data privacy and bias remain.

[0956] translation:

[0957] Artificial intelligence has advanced rapidly in recent years, improving efficiency and decision-making in fields such as medicine, finance, and education, but ethical concerns such as data privacy and bias remain.

[0958] (Application example 1)

[0959] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[0960] On modern e-commerce sites, reviews posted by users are written in a wide variety of languages, making it difficult for multilingual users to understand the content of the reviews. Furthermore, reading the entire lengthy review is inefficient, and summaries are required. The present invention addresses this issue.

[0961] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[0962] In this invention, the server includes means for receiving natural language text from a user, means for using a pre-trained model to summarize the text, means for translating the summarized text into a different language, means for applying the summarized text and the translation result to user reviews on an e-commerce site, and means for returning the translation result to the user, thereby enabling multilingual users to efficiently understand the review content.

[0963] A "user" is a user of the system, and is an individual or organization that inputs text in a natural language.

[0964] "Natural language" refers to a language that humans use on a daily basis and that does not have a specific structure or form.

[0965] "Text" is character string information written in a natural language.

[0966] "Pre-trained model" refers to a machine learning model that has been pre-trained with a large amount of reference data for the purpose of natural language processing.

[0967] A "summary" is a short sentence that succinctly expresses the main points of a long text.

[0968] "Translation" is the act of converting text written in one language into another language.

[0969] "Different language" refers to a language that is different from the natural language input by the user.

[0970] "Translation result" refers to the output of the translated text.

[0971] An "e-commerce site" is a website that sells goods and services over the Internet.

[0972] "User reviews" refer to texts in which users express their opinions or evaluations of products or services they have purchased.

[0973] "Replying" refers to the act of providing an appropriate output in response to an input from a user.

[0974] "Means" refers to elements necessary to achieve a specific function.

[0975] The present invention is a system that automatically summarizes natural language text entered by a user, translates the summary into different languages, and applies it to user reviews on e-commerce sites.

[0976] The system includes the following means:

[0977] 1. A means for users to input natural language text

[0978] 2. Using a pre-trained model to summarize input text

[0979] 3. A means of translating the summarized text into different languages

[0980] 4. A method for applying summarized text and translation results to user reviews on e-commerce sites

[0981] 5. Means of returning translation results to the user

[0982] Hardware and software used

[0983] Hardware: Server, user devices (smartphones, PCs)

[0984] Software: BART model (summarization), Google Translate API (translation), web service (RESTful API), front-end (React, Angular, etc.)

[0985] Processing procedures and examples

[0986] 1. Processing when posting a review

[0987] A user submits a review on an e-commerce site, and the text is sent to the server as a POST request.

[0988] 2. Summary Generation

[0989] The server inputs the received reviews into a pre-trained BART model, a machine learning model pre-trained on large amounts of data for the purpose of natural language processing, to generate summaries.

[0990] 3. Translation

[0991] The server inputs the summarized text into the Google Translate API, which is a third-party web API that provides translation services, and translates it into a different specified language (e.g., Japanese).

[0992] 4. Application

[0993] The server applies the summarized text and translation results to user reviews on an e-commerce site, where the translated summaries are displayed in the review section.

[0994] 5. Response Generation

[0995] The server stores the summary and translation results in a cloud database and returns them to the user's device in JSON format, where the user can view the summary and translation results in real time.

[0996] Specific examples

[0997] For example, if a user posts a review like this:

[0998] “Artificial intelligence has seen rapid advancements in recent years. These advancements are evident in various fields such as healthcare, finance, and education. The implementation of AI systems has led to more efficient processes, better decision-making, and significant cost savings. However, ethical concerns remain, particularly in terms of data privacy and algorithmic bias.”

[0999] The system generates a summary like this, which is then translated into Japanese:

[1000] Abstract: "Artificial intelligence has seen rapid advancements in recent years, improving efficiency and decision-making in various fields such as healthcare, finance, and education. However, concerns like data privacy and ethical bias remain."

[1001] Translation result: "Artificial intelligence has advanced rapidly in recent years, improving efficiency and decision-making in fields such as medicine, finance, and education. However, ethical concerns such as data privacy and bias remain."

[1002] In this way, multilingual users can efficiently understand the review content.

[1003] Prompt Sentence Examples

[1004] "Please summarize the long review below to a maximum of 150 characters and then translate it into Japanese. Review: {Review content}"

[1005] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1006] Step 1:

[1007] A user submits a review by entering natural language text into an e-commerce site. This text is sent from the user's device to the server as a POST request. The input text is a string of natural language characters entered by the user as a review. When the user presses the submit button for the text they entered, the text data is sent to the server in JSON format.

[1008] Step 2:

[1009] The server detects the review text received from the user and inputs it into a pre-trained generative AI model (e.g., the BART model) to summarize it. The input text data is the natural language review text received from the user. The server inputs this text data into the BART model, which summarizes the text based on important information. This results in a shorter summary being output.

[1010] Step 3:

[1011] The server inputs the generated summary into a translation API (for example, Google Translate API) to translate it into a different specified language. The input data is the summary and the language information for the translation. The server sends the summary to the translation API and receives the translation result. As a result, the summary translated into the specified language is output.

[1012] Step 4:

[1013] The server applies the translated summary to user reviews on the e-commerce site. The input data is the translated summary. The server inserts the summary into the review section, making it visible to other users. This step allows the summarized translation to be displayed in the review section.

[1014] Step 5:

[1015] The server stores the summary and translation results in a cloud database and returns them to the user's device in JSON format. The input data is the summary and translation. The server stores this data in a cloud database and sends it to the user's device as a response. The user can check the summary and translation results on their device.

[1016] The above processing steps enable users who speak different languages ​​to efficiently understand the review content.

[1017] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[1018] The present invention combines a system that summarizes text written by a user in a natural language and then translates the summary into a different language with an emotion engine that recognizes emotions from the user's text. The system receives text entered by a user, summarizes it, translates the summary into another language, and recognizes emotions from the text and reflects them in the results.

[1019] Program processing

[1020] The program for this system is as follows. The server summarizes text using a pre-trained natural language processing model (e.g., the BART model). It also uses a translation API provided by a third party (e.g., Google Translate) for translation and an emotion engine for emotion recognition.

[1021] First, a user accesses the server using a terminal and inputs a long piece of text. This text is sent to the server as a POST request. The server receives the text and recognizes emotions using an emotion engine. The recognized emotions influence the subsequent summarization and translation processes.

[1022] The server then inputs the text into a pre-trained model to automatically generate a summary. The generated summary is then input into a third-party translation API for translation into the desired language (e.g., Japanese). Based on the output of the emotion engine, the server adjusts the summary and translation results to optimally respond to the user.

[1023] Specific examples

[1024] Here's a concrete example: Suppose the user types the following long sentence:

[1025] “Artificial intelligence has seen rapid advancements in recent years. These advancements are evident in various fields such as healthcare, finance, and education. The implementation of AI systems has led to more efficient processes, better decision-making, and significant cost savings. However, ethical concerns remain, particularly in terms of data privacy and algorithmic bias.”

[1026] When the device sends this text to the server, the server first uses its emotion engine to recognize emotions such as:

[1027] "Positive emotions: progress, efficiency, cost savings

[1028] Negative Emotions: Ethical Concerns, Data Privacy, and Algorithmic Bias.”

[1029] The server then uses this information to generate a summary:

[1030] “Artificial intelligence has seen rapid advancements in recent years, improving efficiency and decision-making in various fields such as healthcare, finance, and education. However, ethical concerns like data privacy and bias remain.”

[1031] The server then translates this summary into a different language, say Japanese, producing the final result:

[1032] "Artificial intelligence has advanced rapidly in recent years, improving efficiency and decision-making in sectors such as healthcare, finance and education. However, ethical concerns such as data privacy and bias remain."

[1033] The translation takes into account the output of the emotion engine and makes appropriate adjustments, such as emphasizing key points and appropriately reflecting positive and negative sentiment.

[1034] This enables a series of processes to efficiently summarize text entered by a user, translate that summary into different languages, and simultaneously recognize and appropriately reflect the sentiment of the text. This system, combined with an emotion engine, can enrich multilingual communication and more accurately meet user needs.

[1035] The processing flow will be explained below.

[1036] Step 1:

[1037] A user uses a terminal to input a long text in a natural language into a text input form.

[1038] Example: "Artificial intelligence has seen rapid advancements in recent years. ..."

[1039] Step 2:

[1040] The device sends the entered text to the server as a POST request, with the request body containing the text in JSON format.

[1041] Example: {"text": "Artificial intelligence has seen rapid advancements in recent years. ..."}

[1042] Step 3:

[1043] The server receives the POST request, extracts the JSON data from the request body, and uses request.json to get the text field.

[1044] Example: text = data['text']

[1045] Step 4:

[1046] The server uses an emotion engine to recognize the emotion of the received text, a process that classifies emotions such as positive, negative, or neutral based on the content of the text.

[1047] Example: emotion = emotion_engine.analyze(text)

[1048] Step 5:

[1049] The server tokenizes the received text using a pre-trained natural language processing model (BART model). The tokenizer.encode method is used to convert the text into tokens for summarization.

[1050] Example: inputs = tokenizer.encode("summarize: " + text, return_tensors='pt')

[1051] Step 6:

[1052] The server passes the tokenized input to a natural language processing model to generate a summary of the text. The summary result is obtained using the model.generate method.

[1053] Example: summary_ids = model.generate(inputs, max_length=150, min_length=40, ...)

[1054] Step 7:

[1055] The server decodes the generated summary and converts it into a natural language form that the user can understand. Use the tokenizer.decode method to obtain the summary text.

[1056] Example: summary = tokenizer.decode(summary_ids[0], skip_special_tokens=True)

[1057] Step 8:

[1058] The server inputs the summarized text into a third-party translation API (Google Translate API) to translate it into a different language, and retrieves the translation result using the translator.translate method.

[1059] Example: translation = translator.translate(summary, src='en', dest='ja').text

[1060] Step 9:

[1061] The server adjusts the summarization and translation results based on the output of the emotion engine, for example by changing the wording to emphasize positive emotions.

[1062] Example: adjusted_translation = adjust_based_on_emotion(translation, emotion)

[1063] Step 10:

[1064] The server generates a response containing the summary and adjusted translation results in JSON format. The response is created using the jsonify function.

[1065] Example: return jsonify({'summary': summary, 'translation': adjusted_translation})

[1066] Step 11:

[1067] The server sends a JSON response back to the device, which receives the response and displays the summary and translation results to the user.

[1068] Example: {"summary": "Artificial intelligence has seen rapid advancements...", "translation": "Artificial intelligence has seen rapid advancements..."}

[1069] This allows for a system that efficiently summarizes user-entered text, translates the summary into different languages, and recognizes the sentiment of the text and reflects it in the results.

[1070] Example 2

[1071] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1072] Many conventional natural language processing systems are unable to recognize and reflect the sentiment of text when summarizing or translating it. This can result in summaries and translations that do not adequately reflect the user's intentions and sentiments, potentially reducing the quality of communication. Furthermore, the lack of automatic adjustments to translation results increases the user's workload.

[1073] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[1074] In this invention, the server includes means for receiving natural language text from a user, means for recognizing emotions from the text, means for using a pre-trained model to summarize the text, means for translating the summarized text into a different language, means for adjusting the translation result based on the recognized emotions, and means for returning the translation result to the user, thereby making it possible to provide a summary and translation result that appropriately reflects the user's intentions and emotions.

[1075] A "user" is a person or entity that utilizes the system to input natural language text and receive summaries and translation results.

[1076] "Natural language text" is a sentence or text written in a normal human language.

[1077] "Means for recognizing emotions" refers to means that use engines or algorithms, such as sentiment analysis models, to analyze and identify emotions within text.

[1078] A "pre-trained model" is an artificial intelligence model that has been trained in advance based on a large amount of text data to perform a specific task (in this case, summarization) with high accuracy.

[1079] A "means of translation" is a means of using technology or services, such as a translation API, to convert text written in one language into another language.

[1080] "Means for responding to the user" refers to the technology or service that sends the processing results to the user's terminal so that the user can view them.

[1081] A "server" is a computer system that receives requests from users, performs various processes internally, and returns the final results.

[1082] The present invention combines a system that summarizes text written by a user in a natural language and then translates the summary into a different language with an emotion engine that recognizes emotions from the user's text. The system receives text entered by a user, summarizes it, translates the summary into another language, and recognizes emotions from the text and reflects them in the results.

[1083] Program processing

[1084] The processing of the system program is explained below.

[1085] The server uses a pre-trained natural language processing model (e.g., the BART model) to summarize the text, a third-party translation API (e.g., Google Translate) for translation, and an emotion engine for emotion recognition.

[1086] First, a user accesses the server using a terminal and inputs a long piece of text. This text is sent to the server as a POST request. The server receives the text and uses an emotion engine to recognize emotions. The recognized emotions influence the subsequent summarization and translation processes.

[1087] The server then inputs the text into a pre-trained model to automatically generate a summary. The generated summary is then input into a third-party translation API for translation into the desired language (e.g., Japanese). Based on the output of the emotion engine, the server adjusts the summary and translation results to optimally respond to the user.

[1088] Specific examples

[1089] Here's a concrete example: Suppose the user types the following long sentence:

[1090] “Artificial intelligence has seen rapid advancements in recent years. These advancements are evident in various fields such as healthcare, finance, and education. The implementation of AI systems has led to more efficient processes, better decision-making, and significant cost savings. However, ethical concerns remain, particularly in terms of data privacy and algorithmic bias.”

[1091] When the device sends this text to the server, the server first uses its emotion engine to recognize emotions such as:

[1092] "Positive emotions: progress, efficiency, cost savings

[1093] Negative Emotions: Ethical Concerns, Data Privacy, and Algorithmic Bias.”

[1094] The server then uses this information to generate a summary:

[1095] “Artificial intelligence has seen rapid advancements in recent years, improving efficiency and decision-making in various fields such as healthcare, finance, and education. However, ethical concerns like data privacy and bias remain.”

[1096] The server then translates this summary into a different language, say Japanese, producing the final result:

[1097] "Artificial intelligence has advanced rapidly in recent years, improving efficiency and decision-making in sectors such as healthcare, finance and education. However, ethical concerns such as data privacy and bias remain."

[1098] The translation takes into account the output of the emotion engine and makes appropriate adjustments, such as emphasizing key points and appropriately reflecting positive and negative sentiment.

[1099] This enables a series of processes to efficiently summarize text entered by a user, translate that summary into different languages, and simultaneously recognize and appropriately reflect the sentiment of the text. This system, combined with an emotion engine, can enrich multilingual communication and more accurately meet user needs.

[1100] Prompt Sentence Examples

[1101] An example of a prompt to use when a user has entered a long text and wants a summary of that text is:

[1102] "Summarize the following long English sentence and translate it into Japanese. Also, please specify the emotion that is recognized."

[1103] Then, here's an example of the specific text the user would enter:

[1104] “Artificial intelligence has seen rapid advancements in recent years. These advancements are evident in various fields such as healthcare, finance, and education. The implementation of AI systems has led to more efficient processes, better decision-making, and significant cost savings. However, ethical concerns remain, particularly in terms of data privacy and algorithmic bias.”

[1105] This provides clear instructions for the system to perform the expected functions accurately.

[1106] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1107] Step 1:

[1108] Entering and sending text

[1109] First, a user accesses the system using a terminal and inputs a long piece of natural language text. When the user inputs the text and clicks the "Submit" button, the terminal sends the input text to the server as a POST request. Specifically, the text is input as a browser form, and JavaScript processes it and sends it to the server.

[1110] Input: Natural language text entered by the user

[1111] Output: POST request from the terminal to the server

[1112] Step 2:

[1113] Text Receiving and Emotion Recognition

[1114] The server receives the POST request sent from the device and passes the received text to an emotion engine (e.g., an emotion analysis model) to analyze the type and intensity of emotion. This analysis process identifies positive and negative emotions contained in the text.

[1115] Input: The user's natural language text included in the POST request

[1116] Output: Sentiment analysis results (positive and negative) of the text

[1117] Step 3:

[1118] Summary Generation

[1119] The server inputs the emotion-recognized text into a pre-trained natural language processing model (e.g., the BART model). The model analyzes the text, extracts key information, and generates a summary. The summary is generated quickly and reduces the content while preserving the essence of the original text.

[1120] Input: Natural language text with sentiment information

[1121] Output: Generated summary text

[1122] Step 4:

[1123] Executing the translation

[1124] The server sends the generated summary text to a translation API provided by a third party (e.g., a translation service) for translation into the desired language. The translation API connects to the external service via an HTTP request and returns the translated text.

[1125] Input: Summarized text

[1126] Output: The translated text

[1127] Step 5:

[1128] Adjusting the translation results

[1129] The server adjusts the translation results based on the output of the emotion engine, specifically by emphasizing key points and modifying the translation to appropriately reflect positive and negative emotions, so that the information is conveyed in the most optimal way for the user.

[1130] Input: Sentiment analysis results and translated text

[1131] Output: Sentiment-informed translated text

[1132] Step 6:

[1133] Returning the results

[1134] Finally, the server sends the adjusted translation back to the user's device. The server generates an HTTP response and sends the text in HTML or JSON format to the user's device, where the user can view the translation results.

[1135] Input: Adjusted translation text

[1136] Output: Translation results displayed on the user's device

[1137] Through these steps, users can efficiently summarize natural language text, translate the summary into different languages, and even recognize emotions to achieve optimally reflected results.

[1138] (Application example 2)

[1139] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1140] The present invention relates to a system that efficiently performs summarization, translation, and emotion recognition for long natural language texts entered by users, and aims to provide appropriate summarization and translation results, particularly in multilingual content distribution services. Furthermore, because conventional technologies have been insufficient in recognizing emotions and reflecting the results, there is a need for a system that can provide more natural and appropriate output by recognizing emotions in text.

[1141] The identification process by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for receiving natural language text from a user, means for using a pre-trained model to summarize the text, means for translating the summarized text into a different language, means for analyzing emotions extracted from the text, means for adjusting the summary and translation result based on the emotions, and means for returning the adjusted translation result to the user. This allows the user to obtain a translation result in which the input text is efficiently summarized and appropriate emotions are reflected.

[1142] "User" means any person or entity that utilizes the System to enter text and receive results.

[1143] A "natural language" is a language that humans use on a daily basis (e.g., Japanese, English, etc.).

[1144] "Text" refers to natural language sentences entered by a user.

[1145] A "pre-trained model" is a machine learning model for natural language processing that has been trained in advance using a large amount of data.

[1146] "Summarization" is the process of extracting important information from a longer text and summarizing it in a short form.

[1147] "Translation" is the process of converting text written in one language into a different language.

[1148] "Emotion" refers to a user's feelings or emotional state, such as positive or negative, extracted from text.

[1149] "Analyzing sentiment" is the process of recognizing sentiment in text and determining whether it is positive, negative, or neutral.

[1150] "Adjustment" refers to optimizing the summary and translation results by reflecting the results of sentiment analysis.

[1151] "Returning results" means providing the processed summary and translation, as well as sentiment analysis results, to the user.

[1152] The present invention is a system that summarizes text written by a user in a natural language and then translates the summary into a different language, incorporating a process for recognizing emotions from the user's text. This system is particularly effective in multilingual content distribution services.

[1153] First, the server receives natural language text from the user. The received text is then sent to a sentiment analysis unit to identify the sentiment within the text. A pre-trained natural language processing model, such as the BERT model, is used for sentiment analysis. The analyzed sentiment is reflected in the subsequent summarization and translation process.

[1154] The server then uses a pre-trained natural language processing model, such as the BART model, to summarize the text entered by the user. This summarization process extracts important information from long pieces of text and summarizes it in a concise form.

[1155] The summarized text is then translated into different languages ​​using a third-party translation API, such as Google Translate.

[1156] The translation result is adjusted as needed, taking into account the results of sentiment analysis, so that the translation provided to the user appropriately reflects the sentiment of the input text.

[1157] Finally, the server returns the adjusted summary and translation results to the user. Through this process, the user can efficiently summarize long texts and obtain translation results that appropriately reflect the sentiment of the text.

[1158] Hardware and software used

[1159] Hardware: Server (AWS, GCP, etc.)

[1160] software:

[1161] Python

[1162] Transformers library (BART, BERT models)

[1163] Translation API (e.g., Google Translate API)

[1164] Specific examples

[1165] Suppose the user types a long sentence like this:

[1166] “Artificial intelligence has seen rapid advancements in recent years. These advancements are evident in various fields such as healthcare, finance, and education. The implementation of AI systems has led to more efficient processes, better decision-making, and significant cost savings. However, ethical concerns remain, particularly in terms of data privacy and algorithmic bias.”

[1167] The server receives this text and first uses an emotion engine to recognize emotions such as:

[1168] "Positive Emotions: Progress, Efficiency, and Cost Reduction"

[1169] "Negative Emotions: Ethical Concerns, Data Privacy, and Algorithmic Bias"

[1170] Then, generate a summary using the pre-trained model:

[1171] “Artificial intelligence has seen rapid advancements in recent years, improving efficiency and decision-making in various fields such as healthcare, finance, and education. However, ethical concerns like data privacy and bias remain.”

[1172] Translate this summary into a different language, for example Japanese:

[1173] "Artificial intelligence has advanced rapidly in recent years, improving efficiency and decision-making in sectors such as healthcare, finance and education. However, ethical concerns such as data privacy and bias remain."

[1174] Example prompt sentence:

[1175] "Summarize a text about recent advances in AI technology, translate it into Japanese, and perform sentiment analysis."

[1176] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1177] Step 1:

[1178] The user inputs text in a natural language and sends it from the device to the server. An example of input text could be a long sentence such as "Artificial intelligence has seen rapid advancements in recent years." The device then sends this long text to the server as a POST request.

[1179] Step 2:

[1180] The server inputs the received text into an emotion recognition engine, which uses a pre-trained BERT model to analyze the sentiment within this text. In this process, keywords and phrases are extracted from the text and classified as positive, negative, or neutral. For example, "progress" is recognized as positive, and "ethical concerns" as negative.

[1181] Step 3:

[1182] The server then sends the input text to a text summarization engine, which uses a pre-trained BART model to convert long text into concise summaries. Specifically, it preserves the meaning of the text while extracting only the important information and summarizing it into short sentences. The input to this process is long text, and the output is a summary.

[1183] Step 4:

[1184] The server sends the summarized text to a translation engine, which uses a third-party translation API (e.g., Google Translate) to translate the summarized text into the specified language. The input to this process is the summarized text, and the output is the translated text. For example, the summary sentence "Artificial intelligence has made rapid advancements." is translated to "Artificial intelligence has made rapid advancements."

[1185] Step 5:

[1186] The server adjusts the translation results based on the results of sentiment analysis. Based on the results of sentiment analysis, the summary and translation results are adjusted to emphasize specific emotions. The input to this process is the sentiment analysis results and the translated text, and the output is the adjusted translation result. For example, the sentence may be restructured to emphasize parts with positive sentiment.

[1187] Step 6:

[1188] The server returns the final result to the user. By returning the adjusted translation result, the user can get a proper summary and translation result based on the input text, which makes it easier for users to understand long sentences and grasp the nuances of emotions in different languages.

[1189] The specific processing unit 290 transmits the result of the specific processing to the headset type terminal 314. In the headset type terminal 314, the control unit 46A causes the speaker 240 and the display 343 to output the result of the specific processing. The microphone 238 acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[1190] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1191] In the above embodiment, an example was given in which the specific processing is performed by the data processing device 12, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the headset type terminal 314.

[1192] [Fourth embodiment]

[1193] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[1194] 7, a data processing system 410 includes a data processing device 12 and a robot 414. An example of the data processing device 12 is a server.

[1195] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[1196] The robot 414 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication I / F 44, and a control target 443. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, the camera 42, and the control target 443 are also connected to the bus 52.

[1197] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[1198] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[1199] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[1200] The control object 443 includes a display device, LEDs in the eyes, and motors for driving the arms, hands, and feet. The posture and gestures of the robot 414 are controlled by controlling the motors of the arms, hands, and feet. Some of the emotions of the robot 414 can be expressed by controlling these motors. In addition, the facial expressions of the robot 414 can also be expressed by controlling the light emission state of the LEDs in the eyes of the robot 414.

[1201] Fig. 8 shows an example of the main functions of the data processing device 12 and the robot 414. As shown in Fig. 8, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[1202] The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[1203] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[1204] In the robot 414, the processor 46 performs the reception output process. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[1205] Next, a description will be given of the specific processing performed by the specific processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1206] The present invention provides a system that summarizes text written in a natural language by a user and then translates the summary into a different language. This system receives text entered by a user, summarizes it briefly, and automatically translates it into other languages, such as Japanese, thereby supporting efficient understanding of information and communication between multiple languages.

[1207] Program processing

[1208] The program for this system is as follows: The server summarizes text using a pre-trained natural language processing model (e.g., the BART model), and uses a translation API provided by a third party (e.g., Google Translate) for translation.

[1209] First, a user accesses the server using a terminal and inputs a long piece of text. This text is sent to the server as a POST request. The server receives the text and inputs it into a pre-trained model to automatically generate a summary.

[1210] The server then inputs the generated summary into a third-party translation API, which translates the summarized text into the desired language (e.g., Japanese). Once the translation is complete, the server returns the summary and translation results to the device as a JSON response. The user can view the summary and translation results on the device screen.

[1211] Specific examples

[1212] Here's a concrete example: Suppose the user types the following long sentence:

[1213] “Artificial intelligence has seen rapid advancements in recent years. These advancements are evident in various fields such as healthcare, finance, and education. The implementation of AI systems has led to more efficient processes, better decision-making, and significant cost savings. However, ethical concerns remain, particularly in terms of data privacy and algorithmic bias.”

[1214] When the terminal sends this text to the server, the server generates a summary like this:

[1215] “Artificial intelligence has seen rapid advancements in recent years, improving efficiency and decision-making in various fields such as healthcare, finance, and education. However, ethical concerns like data privacy and bias remain.”

[1216] The server then translates this summary into a different language, for example Japanese, to produce the final result:

[1217] "Artificial intelligence has advanced rapidly in recent years, improving efficiency and decision-making in sectors such as healthcare, finance and education. However, ethical concerns such as data privacy and bias remain."

[1218] In this way, the present invention realizes a series of processes for summarizing text entered by a user and then translating the summary into a different language. The embodiment of the invention embodies a series of processes including multiple steps, such as data exchange between a server and a terminal, summarizing the text, and translating it.

[1219] The processing flow will be explained below.

[1220] Step 1:

[1221] A user uses a terminal to input a long text in a natural language into a text input form.

[1222] Example: "Artificial intelligence has seen rapid advancements in recent years. ..."

[1223] Step 2:

[1224] The device sends the entered text to the server as a POST request, with the request body containing the text in JSON format.

[1225] Example: {"text": "Artificial intelligence has seen rapid advancements in recent years. ..."}

[1226] Step 3:

[1227] The server receives the POST request, extracts the JSON data from the request body, and uses request.json to get the text field.

[1228] Example: text = data['text']

[1229] Step 4:

[1230] The server tokenizes the received text using a pre-trained natural language processing model (BART model). The tokenizer.encode method is used to convert the text into tokens for summarization.

[1231] Example: inputs = tokenizer.encode("summarize: " + text, return_tensors='pt')

[1232] Step 5:

[1233] The server passes the tokenized input to a natural language processing model to generate a summary of the text. The summary result is obtained using the model.generate method.

[1234] Example: summary_ids = model.generate(inputs, max_length=150, min_length=40, ...)

[1235] Step 6:

[1236] The server decodes the generated summary and converts it into a natural language form that the user can understand. Use the tokenizer.decode method to obtain the summary text.

[1237] Example: summary = tokenizer.decode(summary_ids[0], skip_special_tokens=True)

[1238] Step 7:

[1239] The server inputs the summarized text into a third-party translation API (Google Translate API) to translate it into a different language, and retrieves the translation result using the translator.translate method.

[1240] Example: translation = translator.translate(summary, src='en', dest='ja').text

[1241] Step 8:

[1242] The server compiles the summary and translation results into JSON format and generates a response. The jsonify function is used to create the response.

[1243] Example: return jsonify({'summary': summary, 'translation': translation})

[1244] Step 9:

[1245] The server sends a JSON response back to the device, which receives the response and displays the summary and translation results to the user.

[1246] Example: {"summary": "Artificial intelligence has seen rapid advancements...", "translation": "Artificial intelligence has seen rapid advancements..."}

[1247] This completes the process of efficiently summarizing user-entered text and translating the summary into different languages.

[1248] Example 1

[1249] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1250] In today's world, the exchange and understanding of information across multiple languages ​​is extremely important. However, it takes a great deal of time and effort for users to understand long texts written in languages ​​other than their native language. Furthermore, there are limited technologies that can automatically summarize texts and appropriately translate them without specialized knowledge. This makes it difficult for users to easily exchange information across multiple languages, hindering efficient communication.

[1251] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[1252] In this invention, the server includes means for receiving text written in a natural language from a user, means for using a pre-trained generative model to automatically summarize the text, means for using a translation service to translate the summarized text into a different language, and means for returning the translation result to the user's terminal, thereby enabling the user to easily summarize long texts and further translate them into different languages, thereby enabling efficient information comprehension and smooth communication between multiple languages.

[1253] "Text written in a natural language" refers to text written in a language that humans use on a daily basis.

[1254] "User" refers to a person or organization that uses the system.

[1255] A "server" refers to a computer that provides services to other computers over a network.

[1256] "Terminal" refers to the computer or device through which a user interacts with a server.

[1257] A "summary" is a concise description of the main information extracted from a longer original text.

[1258] A "pre-trained generative model" refers to a model that has been trained in advance on a large amount of data and can automatically perform a specific task (in this case, summarization).

[1259] "Translation" refers to the conversion of text written in one language into a different language.

[1260] "Translation Service" means an online application or API that provides text translation.

[1261] "Translation Result" means text that has been converted into a different language by a Translation Service.

[1262] A "pre-trained natural language processing model" refers to a model that has been trained on a large amount of text data in advance to perform natural language processing tasks.

[1263] "Translation API" refers to an application programming interface provided to enable translation functionality to be used from any program.

[1264] "JSON format" refers to the data exchange format called JavaScript Object Notation.

[1265] This invention is a system that automatically summarizes natural language text entered by a user and translates the summary into a different language. The system has the following main functions:

[1266] 1. User enters text

[1267] A user inputs long pieces of text through a web browser or application.

[1268] 2. Sending text from the device to the server

[1269] The device sends the text entered by the user to the server using an HTTP POST request.

[1270] 3. Server-generated text summarization

[1271] The server uses a pre-trained generative model (e.g., the BART model) for text summarization, which summarizes long texts into short, concise summaries.

[1272] 4. Server-Based Abstract Translation

[1273] The summarized text is sent via the server to a translation service (e.g., Google Translate API) and translated into the desired language.

[1274] 5. The server returns the translation results

[1275] The server returns the translated text in JSON format to the device, where the user can view a summary and translation results.

[1276] As a concrete example, consider the case where a user enters the following long sentence:

[1277] Artificial intelligence has seen rapid advancements in recent years. These advancements are evident in various fields such as healthcare, finance, and education. The implementation of AI systems has led to more efficient processes, better decision-making, and significant cost savings. However, ethical concerns remain, particularly in terms of data privacy and algorithmic bias.

[1278] When the device sends this text to the server, the server generates a summary like this:

[1279] Artificial intelligence has seen rapid advancements in recent years, improving efficiency and decision-making in various fields such as healthcare, finance, and education. However, ethical concerns like data privacy and bias remain.

[1280] The server then translates this summary into a different language, for example Japanese:

[1281] Artificial intelligence has advanced rapidly in recent years, improving efficiency and decision-making in fields such as medicine, finance, and education, but ethical concerns such as data privacy and bias remain.

[1282] In this way, users can easily summarize long texts they input and then translate the summaries into different languages. This system allows users to efficiently exchange information between multiple languages.

[1283] Prompt Sentence Examples

[1284] Summarize the following long passage and translate it into Japanese:

[1285] "Artificial intelligence has seen rapid advancements in recent years. These advancements are evident in various fields such as healthcare, finance, and education. The implementation of AI systems has led to more efficient processes, better decision-making, and significant cost savings. However, ethical concerns remain, particularly in terms of data privacy and algorithmic bias."

[1286] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1287] System program processing flow

[1288] Step 1: User Enters Text

[1289] Specific description:

[1290] A user inputs a long piece of natural language text into an input field in a web browser or application on his or her device.

[1291] Specific behavior:

[1292] A user enters the following text into a browser input field:

[1293] Artificial intelligence has seen rapid advancements in recent years. These advancements are evident in various fields such as healthcare, finance, and education. The implementation of AI systems has led to more efficient processes, better decision-making, and significant cost savings. However, ethical concerns remain, particularly in terms of data privacy and algorithmic bias.

[1294] Step 2: Send text to the server using the device

[1295] Specific description:

[1296] The terminal receives the text entered by the user and sends it to the server as a POST request.

[1297] Input and Output:

[1298] Input: Long text entered by the user

[1299] Output: HTTP POST request sent to the server

[1300] Specific behavior:

[1301] The device sends the following HTTP POST request to the server:

[1302] POST / summarize_translate

[1303] Content-Type: application / json

[1304] {

[1305] "text": "Artificial intelligence has seen rapid advancements in recent years. These advancements are evident in various fields such as healthcare, finance, and education. The implementation of AI systems has led to more efficient processes, better decision-making, and significant cost savings. However, ethical concerns remain, particularly in terms of data privacy and algorithmic bias."

[1306] }

[1307] Step 3: Server-generated text summarization

[1308] Specific description:

[1309] The server inputs the received text into a pre-trained generative model (the BART model) and automatically generates a summary.

[1310] Input and Output:

[1311] Input: Long text sent from the terminal

[1312] Output: A short summary of the text

[1313] Specific behavior:

[1314] The server invokes the BART model and generates a summary like this:

[1315] Artificial intelligence has seen rapid advancements in recent years, improving efficiency and decision-making in various fields such as healthcare, finance, and education. However, ethical concerns like data privacy and bias remain.

[1316] Step 4: Server translation of the summary

[1317] Specific description:

[1318] The server sends the generated summary to a translation service (Google Translate API) for translation into the desired language.

[1319] Input and Output:

[1320] Input: A short text to be summarized

[1321] Output: The translated text

[1322] Specific behavior:

[1323] The server calls the Google Translate API and obtains the translation result as follows:

[1324] Artificial intelligence has advanced rapidly in recent years, improving efficiency and decision-making in fields such as medicine, finance, and education, but ethical concerns such as data privacy and bias remain.

[1325] Step 5: The server returns the translation results to the device

[1326] Specific description:

[1327] The server returns the translated text and a summary to the device in JSON format.

[1328] Input and Output:

[1329] Input: translated text and summary

[1330] Output: JSON response sent back to the device

[1331] Specific behavior:

[1332] The server sends the following JSON response to the device:

[1333] {

[1334] "summary": "Artificial intelligence has seen rapid advancements in recent years, improving efficiency and decision-making in various fields such as healthcare, finance, and education. However, ethical concerns like data privacy and bias remain.",

[1335] "translation": "Artificial intelligence has advanced rapidly in recent years, improving efficiency and decision-making in sectors like healthcare, finance, and education. However, ethical concerns such as data privacy and bias remain."

[1336] }

[1337] Step 6: User summarization and translation confirmation

[1338] Specific description:

[1339] The user checks the returned summary and translation results on the screen of the terminal.

[1340] Input and Output:

[1341] Input: JSON response returned from the server

[1342] Output: Summary and translation results displayed on the terminal

[1343] Specific behavior:

[1344] The user sees the summary and translation results in the browser or application screen, as shown below.

[1345] summary:

[1346] Artificial intelligence has seen rapid advancements in recent years, improving efficiency and decision-making in various fields such as healthcare, finance, and education. However, ethical concerns like data privacy and bias remain.

[1347] translation:

[1348] Artificial intelligence has advanced rapidly in recent years, improving efficiency and decision-making in fields such as medicine, finance, and education, but ethical concerns such as data privacy and bias remain.

[1349] (Application example 1)

[1350] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1351] On modern e-commerce sites, reviews posted by users are written in a wide variety of languages, making it difficult for multilingual users to understand the content of the reviews. Furthermore, reading the entire lengthy review is inefficient, and summaries are required. The present invention addresses this issue.

[1352] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[1353] In this invention, the server includes means for receiving natural language text from a user, means for using a pre-trained model to summarize the text, means for translating the summarized text into a different language, means for applying the summarized text and the translation result to user reviews on an e-commerce site, and means for returning the translation result to the user, thereby enabling multilingual users to efficiently understand the review content.

[1354] A "user" is a user of the system, and is an individual or organization that inputs text in a natural language.

[1355] "Natural language" refers to a language that humans use on a daily basis and that does not have a specific structure or form.

[1356] "Text" is character string information written in a natural language.

[1357] "Pre-trained model" refers to a machine learning model that has been pre-trained with a large amount of reference data for the purpose of natural language processing.

[1358] A "summary" is a short sentence that succinctly expresses the main points of a long text.

[1359] "Translation" is the act of converting text written in one language into another language.

[1360] "Different language" refers to a language that is different from the natural language input by the user.

[1361] "Translation result" refers to the output of the translated text.

[1362] An "e-commerce site" is a website that sells goods and services over the Internet.

[1363] "User reviews" refer to texts in which users express their opinions or evaluations of products or services they have purchased.

[1364] "Replying" refers to the act of providing an appropriate output in response to an input from a user.

[1365] "Means" refers to elements necessary to achieve a specific function.

[1366] The present invention is a system that automatically summarizes natural language text entered by a user, translates the summary into different languages, and applies it to user reviews on e-commerce sites.

[1367] The system includes the following means:

[1368] 1. A means for users to input natural language text

[1369] 2. Using a pre-trained model to summarize input text

[1370] 3. A means of translating the summarized text into different languages

[1371] 4. A method for applying summarized text and translation results to user reviews on e-commerce sites

[1372] 5. Means of returning translation results to the user

[1373] Hardware and software used

[1374] Hardware: Server, user devices (smartphones, PCs)

[1375] Software: BART model (summarization), Google Translate API (translation), web service (RESTful API), front-end (React, Angular, etc.)

[1376] Processing procedures and examples

[1377] 1. Processing when posting a review

[1378] A user submits a review on an e-commerce site, and the text is sent to the server as a POST request.

[1379] 2. Summary Generation

[1380] The server inputs the received reviews into a pre-trained BART model, a machine learning model pre-trained on large amounts of data for the purpose of natural language processing, to generate summaries.

[1381] 3. Translation

[1382] The server inputs the summarized text into the Google Translate API, which is a third-party web API that provides translation services, and translates it into a different specified language (e.g., Japanese).

[1383] 4. Application

[1384] The server applies the summarized text and translation results to user reviews on an e-commerce site, where the translated summaries are displayed in the review section.

[1385] 5. Response Generation

[1386] The server stores the summary and translation results in a cloud database and returns them to the user's device in JSON format, where the user can view the summary and translation results in real time.

[1387] Specific examples

[1388] For example, if a user posts a review like this:

[1389] “Artificial intelligence has seen rapid advancements in recent years. These advancements are evident in various fields such as healthcare, finance, and education. The implementation of AI systems has led to more efficient processes, better decision-making, and significant cost savings. However, ethical concerns remain, particularly in terms of data privacy and algorithmic bias.”

[1390] The system generates a summary like this, which is then translated into Japanese:

[1391] Abstract: "Artificial intelligence has seen rapid advancements in recent years, improving efficiency and decision-making in various fields such as healthcare, finance, and education. However, concerns like data privacy and ethical bias remain."

[1392] Translation result: "Artificial intelligence has advanced rapidly in recent years, improving efficiency and decision-making in fields such as medicine, finance, and education. However, ethical concerns such as data privacy and bias remain."

[1393] In this way, multilingual users can efficiently understand the review content.

[1394] Prompt Sentence Examples

[1395] "Please summarize the long review below to a maximum of 150 characters and then translate it into Japanese. Review: {Review content}"

[1396] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1397] Step 1:

[1398] A user submits a review by entering natural language text into an e-commerce site. This text is sent from the user's device to the server as a POST request. The input text is a string of natural language characters entered by the user as a review. When the user presses the submit button for the text they entered, the text data is sent to the server in JSON format.

[1399] Step 2:

[1400] The server detects the review text received from the user and inputs it into a pre-trained generative AI model (e.g., the BART model) to summarize it. The input text data is the natural language review text received from the user. The server inputs this text data into the BART model, which summarizes the text based on important information. This results in a shorter summary being output.

[1401] Step 3:

[1402] The server inputs the generated summary into a translation API (for example, Google Translate API) to translate it into a different specified language. The input data is the summary and the language information for the translation. The server sends the summary to the translation API and receives the translation result. As a result, the summary translated into the specified language is output.

[1403] Step 4:

[1404] The server applies the translated summary to user reviews on the e-commerce site. The input data is the translated summary. The server inserts the summary into the review section, making it visible to other users. This step allows the summarized translation to be displayed in the review section.

[1405] Step 5:

[1406] The server stores the summary and translation results in a cloud database and returns them to the user's device in JSON format. The input data is the summary and translation. The server stores this data in a cloud database and sends it to the user's device as a response. The user can check the summary and translation results on their device.

[1407] The above processing steps enable users who speak different languages ​​to efficiently understand the review content.

[1408] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[1409] The present invention combines a system that summarizes text written by a user in a natural language and then translates the summary into a different language with an emotion engine that recognizes emotions from the user's text. The system receives text entered by a user, summarizes it, translates the summary into another language, and recognizes emotions from the text and reflects them in the results.

[1410] Program processing

[1411] The program for this system is as follows. The server summarizes text using a pre-trained natural language processing model (e.g., the BART model). It also uses a translation API provided by a third party (e.g., Google Translate) for translation and an emotion engine for emotion recognition.

[1412] First, a user accesses the server using a terminal and inputs a long piece of text. This text is sent to the server as a POST request. The server receives the text and recognizes emotions using an emotion engine. The recognized emotions influence the subsequent summarization and translation processes.

[1413] The server then inputs the text into a pre-trained model to automatically generate a summary. The generated summary is then input into a third-party translation API for translation into the desired language (e.g., Japanese). Based on the output of the emotion engine, the server adjusts the summary and translation results to optimally respond to the user.

[1414] Specific examples

[1415] Here's a concrete example: Suppose the user types the following long sentence:

[1416] “Artificial intelligence has seen rapid advancements in recent years. These advancements are evident in various fields such as healthcare, finance, and education. The implementation of AI systems has led to more efficient processes, better decision-making, and significant cost savings. However, ethical concerns remain, particularly in terms of data privacy and algorithmic bias.”

[1417] When the device sends this text to the server, the server first uses its emotion engine to recognize emotions such as:

[1418] "Positive emotions: progress, efficiency, cost savings

[1419] Negative Emotions: Ethical Concerns, Data Privacy, and Algorithmic Bias.”

[1420] The server then uses this information to generate a summary:

[1421] “Artificial intelligence has seen rapid advancements in recent years, improving efficiency and decision-making in various fields such as healthcare, finance, and education. However, ethical concerns like data privacy and bias remain.”

[1422] The server then translates this summary into a different language, say Japanese, producing the final result:

[1423] "Artificial intelligence has advanced rapidly in recent years, improving efficiency and decision-making in sectors such as healthcare, finance and education. However, ethical concerns such as data privacy and bias remain."

[1424] The translation takes into account the output of the emotion engine and makes appropriate adjustments, such as emphasizing key points and appropriately reflecting positive and negative sentiment.

[1425] This enables a series of processes to efficiently summarize text entered by a user, translate that summary into different languages, and simultaneously recognize and appropriately reflect the sentiment of the text. This system, combined with an emotion engine, can enrich multilingual communication and more accurately meet user needs.

[1426] The processing flow will be explained below.

[1427] Step 1:

[1428] A user uses a terminal to input a long text in a natural language into a text input form.

[1429] Example: "Artificial intelligence has seen rapid advancements in recent years. ..."

[1430] Step 2:

[1431] The device sends the entered text to the server as a POST request, with the request body containing the text in JSON format.

[1432] Example: {"text": "Artificial intelligence has seen rapid advancements in recent years. ..."}

[1433] Step 3:

[1434] The server receives the POST request, extracts the JSON data from the request body, and uses request.json to get the text field.

[1435] Example: text = data['text']

[1436] Step 4:

[1437] The server uses an emotion engine to recognize the emotion of the received text, a process that classifies emotions such as positive, negative, or neutral based on the content of the text.

[1438] Example: emotion = emotion_engine.analyze(text)

[1439] Step 5:

[1440] The server tokenizes the received text using a pre-trained natural language processing model (BART model). The tokenizer.encode method is used to convert the text into tokens for summarization.

[1441] Example: inputs = tokenizer.encode("summarize: " + text, return_tensors='pt')

[1442] Step 6:

[1443] The server passes the tokenized input to a natural language processing model to generate a summary of the text. The summary result is obtained using the model.generate method.

[1444] Example: summary_ids = model.generate(inputs, max_length=150, min_length=40, ...)

[1445] Step 7:

[1446] The server decodes the generated summary and converts it into a natural language form that the user can understand. Use the tokenizer.decode method to obtain the summary text.

[1447] Example: summary = tokenizer.decode(summary_ids[0], skip_special_tokens=True)

[1448] Step 8:

[1449] The server inputs the summarized text into a third-party translation API (Google Translate API) to translate it into a different language, and retrieves the translation result using the translator.translate method.

[1450] Example: translation = translator.translate(summary, src='en', dest='ja').text

[1451] Step 9:

[1452] The server adjusts the summarization and translation results based on the output of the emotion engine, for example by changing the wording to emphasize positive emotions.

[1453] Example: adjusted_translation = adjust_based_on_emotion(translation, emotion)

[1454] Step 10:

[1455] The server generates a response containing the summary and adjusted translation results in JSON format. The response is created using the jsonify function.

[1456] Example: return jsonify({'summary': summary, 'translation': adjusted_translation})

[1457] Step 11:

[1458] The server sends a JSON response back to the device, which receives the response and displays the summary and translation results to the user.

[1459] Example: {"summary": "Artificial intelligence has seen rapid advancements...", "translation": "Artificial intelligence has seen rapid advancements..."}

[1460] This allows for a system that efficiently summarizes user-entered text, translates the summary into different languages, and recognizes the sentiment of the text and reflects it in the results.

[1461] Example 2

[1462] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1463] Many conventional natural language processing systems are unable to recognize and reflect the sentiment of text when summarizing or translating it. This can result in summaries and translations that do not adequately reflect the user's intentions and sentiments, potentially reducing the quality of communication. Furthermore, the lack of automatic adjustments to translation results increases the user's workload.

[1464] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[1465] In this invention, the server includes means for receiving natural language text from a user, means for recognizing emotions from the text, means for using a pre-trained model to summarize the text, means for translating the summarized text into a different language, means for adjusting the translation result based on the recognized emotions, and means for returning the translation result to the user, thereby making it possible to provide a summary and translation result that appropriately reflects the user's intentions and emotions.

[1466] A "user" is a person or entity that utilizes the system to input natural language text and receive summaries and translation results.

[1467] "Natural language text" is a sentence or text written in a normal human language.

[1468] "Means for recognizing emotions" refers to means that use engines or algorithms, such as sentiment analysis models, to analyze and identify emotions within text.

[1469] A "pre-trained model" is an artificial intelligence model that has been trained in advance based on a large amount of text data to perform a specific task (in this case, summarization) with high accuracy.

[1470] A "means of translation" is a means of using technology or services, such as a translation API, to convert text written in one language into another language.

[1471] "Means for responding to the user" refers to the technology or service that sends the processing results to the user's terminal so that the user can view them.

[1472] A "server" is a computer system that receives requests from users, performs various processes internally, and returns the final results.

[1473] The present invention combines a system that summarizes text written by a user in a natural language and then translates the summary into a different language with an emotion engine that recognizes emotions from the user's text. The system receives text entered by a user, summarizes it, translates the summary into another language, and recognizes emotions from the text and reflects them in the results.

[1474] Program processing

[1475] The processing of the system program is explained below.

[1476] The server uses a pre-trained natural language processing model (e.g., the BART model) to summarize the text, a third-party translation API (e.g., Google Translate) for translation, and an emotion engine for emotion recognition.

[1477] First, a user accesses the server using a terminal and inputs a long piece of text. This text is sent to the server as a POST request. The server receives the text and uses an emotion engine to recognize emotions. The recognized emotions influence the subsequent summarization and translation processes.

[1478] The server then inputs the text into a pre-trained model to automatically generate a summary. The generated summary is then input into a third-party translation API for translation into the desired language (e.g., Japanese). Based on the output of the emotion engine, the server adjusts the summary and translation results to optimally respond to the user.

[1479] Specific examples

[1480] Here's a concrete example: Suppose the user types the following long sentence:

[1481] “Artificial intelligence has seen rapid advancements in recent years. These advancements are evident in various fields such as healthcare, finance, and education. The implementation of AI systems has led to more efficient processes, better decision-making, and significant cost savings. However, ethical concerns remain, particularly in terms of data privacy and algorithmic bias.”

[1482] When the device sends this text to the server, the server first uses its emotion engine to recognize emotions such as:

[1483] "Positive emotions: progress, efficiency, cost savings

[1484] Negative Emotions: Ethical Concerns, Data Privacy, and Algorithmic Bias.”

[1485] The server then uses this information to generate a summary:

[1486] “Artificial intelligence has seen rapid advancements in recent years, improving efficiency and decision-making in various fields such as healthcare, finance, and education. However, ethical concerns like data privacy and bias remain.”

[1487] The server then translates this summary into a different language, say Japanese, producing the final result:

[1488] "Artificial intelligence has advanced rapidly in recent years, improving efficiency and decision-making in sectors such as healthcare, finance and education. However, ethical concerns such as data privacy and bias remain."

[1489] The translation takes into account the output of the emotion engine and makes appropriate adjustments, such as emphasizing key points and appropriately reflecting positive and negative sentiment.

[1490] This enables a series of processes to efficiently summarize text entered by a user, translate that summary into different languages, and simultaneously recognize and appropriately reflect the sentiment of the text. This system, combined with an emotion engine, can enrich multilingual communication and more accurately meet user needs.

[1491] Prompt Sentence Examples

[1492] An example of a prompt to use when a user has entered a long text and wants a summary of that text is:

[1493] "Summarize the following long English sentence and translate it into Japanese. Also, please specify the emotion that is recognized."

[1494] Then, here's an example of the specific text the user would enter:

[1495] “Artificial intelligence has seen rapid advancements in recent years. These advancements are evident in various fields such as healthcare, finance, and education. The implementation of AI systems has led to more efficient processes, better decision-making, and significant cost savings. However, ethical concerns remain, particularly in terms of data privacy and algorithmic bias.”

[1496] This provides clear instructions for the system to perform the expected functions accurately.

[1497] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1498] Step 1:

[1499] Entering and sending text

[1500] First, a user accesses the system using a terminal and inputs a long piece of natural language text. When the user inputs the text and clicks the "Submit" button, the terminal sends the input text to the server as a POST request. Specifically, the text is input as a browser form, and JavaScript processes it and sends it to the server.

[1501] Input: Natural language text entered by the user

[1502] Output: POST request from the terminal to the server

[1503] Step 2:

[1504] Text Receiving and Emotion Recognition

[1505] The server receives the POST request sent from the device and passes the received text to an emotion engine (e.g., an emotion analysis model) to analyze the type and intensity of emotion. This analysis process identifies positive and negative emotions contained in the text.

[1506] Input: The user's natural language text included in the POST request

[1507] Output: Sentiment analysis results (positive and negative) of the text

[1508] Step 3:

[1509] Summary Generation

[1510] The server inputs the emotion-recognized text into a pre-trained natural language processing model (e.g., the BART model). The model analyzes the text, extracts key information, and generates a summary. The summary is generated quickly and reduces the content while preserving the essence of the original text.

[1511] Input: Natural language text with sentiment information

[1512] Output: Generated summary text

[1513] Step 4:

[1514] Executing the translation

[1515] The server sends the generated summary text to a translation API provided by a third party (e.g., a translation service) for translation into the desired language. The translation API connects to the external service via an HTTP request and returns the translated text.

[1516] Input: Summarized text

[1517] Output: The translated text

[1518] Step 5:

[1519] Adjusting the translation results

[1520] The server adjusts the translation results based on the output of the emotion engine, specifically by emphasizing key points and modifying the translation to appropriately reflect positive and negative emotions, so that the information is conveyed in the most optimal way for the user.

[1521] Input: Sentiment analysis results and translated text

[1522] Output: Sentiment-informed translated text

[1523] Step 6:

[1524] Returning the results

[1525] Finally, the server sends the adjusted translation back to the user's device. The server generates an HTTP response and sends the text in HTML or JSON format to the user's device, where the user can view the translation results.

[1526] Input: Adjusted translation text

[1527] Output: Translation results displayed on the user's device

[1528] Through these steps, users can efficiently summarize natural language text, translate the summary into different languages, and even recognize emotions to achieve optimally reflected results.

[1529] (Application example 2)

[1530] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1531] The present invention relates to a system that efficiently performs summarization, translation, and emotion recognition for long natural language texts entered by users, and aims to provide appropriate summarization and translation results, particularly in multilingual content distribution services. Furthermore, because conventional technologies have been insufficient in recognizing emotions and reflecting the results, there is a need for a system that can provide more natural and appropriate output by recognizing emotions in text.

[1532] The identification process by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for receiving natural language text from a user, means for using a pre-trained model to summarize the text, means for translating the summarized text into a different language, means for analyzing emotions extracted from the text, means for adjusting the summary and translation result based on the emotions, and means for returning the adjusted translation result to the user. This allows the user to obtain a translation result in which the input text is efficiently summarized and appropriate emotions are reflected.

[1533] "User" means any person or entity that utilizes the System to enter text and receive results.

[1534] A "natural language" is a language that humans use on a daily basis (e.g., Japanese, English, etc.).

[1535] "Text" refers to natural language sentences entered by a user.

[1536] A "pre-trained model" is a machine learning model for natural language processing that has been trained in advance using a large amount of data.

[1537] "Summarization" is the process of extracting important information from a longer text and summarizing it in a short form.

[1538] "Translation" is the process of converting text written in one language into a different language.

[1539] "Emotion" refers to a user's feelings or emotional state, such as positive or negative, extracted from text.

[1540] "Analyzing sentiment" is the process of recognizing sentiment in text and determining whether it is positive, negative, or neutral.

[1541] "Adjustment" refers to optimizing the summary and translation results by reflecting the results of sentiment analysis.

[1542] "Returning results" means providing the processed summary and translation, as well as sentiment analysis results, to the user.

[1543] The present invention is a system that summarizes text written by a user in a natural language and then translates the summary into a different language, incorporating a process for recognizing emotions from the user's text. This system is particularly effective in multilingual content distribution services.

[1544] First, the server receives natural language text from the user. The received text is then sent to a sentiment analysis unit to identify the sentiment within the text. A pre-trained natural language processing model, such as the BERT model, is used for sentiment analysis. The analyzed sentiment is reflected in the subsequent summarization and translation process.

[1545] The server then uses a pre-trained natural language processing model, such as the BART model, to summarize the text entered by the user. This summarization process extracts important information from long pieces of text and summarizes it in a concise form.

[1546] The summarized text is then translated into different languages ​​using a third-party translation API, such as Google Translate.

[1547] The translation result is adjusted as needed, taking into account the results of sentiment analysis, so that the translation provided to the user appropriately reflects the sentiment of the input text.

[1548] Finally, the server returns the adjusted summary and translation results to the user. Through this process, the user can efficiently summarize long texts and obtain translation results that appropriately reflect the sentiment of the text.

[1549] Hardware and software used

[1550] Hardware: Server (AWS, GCP, etc.)

[1551] software:

[1552] Python

[1553] Transformers library (BART, BERT models)

[1554] Translation API (e.g., Google Translate API)

[1555] Specific examples

[1556] Suppose the user types a long sentence like this:

[1557] “Artificial intelligence has seen rapid advancements in recent years. These advancements are evident in various fields such as healthcare, finance, and education. The implementation of AI systems has led to more efficient processes, better decision-making, and significant cost savings. However, ethical concerns remain, particularly in terms of data privacy and algorithmic bias.”

[1558] The server receives this text and first uses an emotion engine to recognize emotions such as:

[1559] "Positive Emotions: Progress, Efficiency, and Cost Reduction"

[1560] "Negative Emotions: Ethical Concerns, Data Privacy, and Algorithmic Bias"

[1561] Then, generate a summary using the pre-trained model:

[1562] “Artificial intelligence has seen rapid advancements in recent years, improving efficiency and decision-making in various fields such as healthcare, finance, and education. However, ethical concerns like data privacy and bias remain.”

[1563] Translate this summary into a different language, for example Japanese:

[1564] "Artificial intelligence has advanced rapidly in recent years, improving efficiency and decision-making in sectors such as healthcare, finance and education. However, ethical concerns such as data privacy and bias remain."

[1565] Example prompt sentence:

[1566] "Summarize a text about recent advances in AI technology, translate it into Japanese, and perform sentiment analysis."

[1567] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1568] Step 1:

[1569] The user inputs text in a natural language and sends it from the device to the server. An example of input text could be a long sentence such as "Artificial intelligence has seen rapid advancements in recent years." The device then sends this long text to the server as a POST request.

[1570] Step 2:

[1571] The server inputs the received text into an emotion recognition engine, which uses a pre-trained BERT model to analyze the sentiment within this text. In this process, keywords and phrases are extracted from the text and classified as positive, negative, or neutral. For example, "progress" is recognized as positive, and "ethical concerns" as negative.

[1572] Step 3:

[1573] The server then sends the input text to a text summarization engine, which uses a pre-trained BART model to convert long text into concise summaries. Specifically, it preserves the meaning of the text while extracting only the important information and summarizing it into short sentences. The input to this process is long text, and the output is a summary.

[1574] Step 4:

[1575] The server sends the summarized text to a translation engine, which uses a third-party translation API (e.g., Google Translate) to translate the summarized text into the specified language. The input to this process is the summarized text, and the output is the translated text. For example, the summary sentence "Artificial intelligence has made rapid advancements." is translated to "Artificial intelligence has made rapid advancements."

[1576] Step 5:

[1577] The server adjusts the translation results based on the results of sentiment analysis. Based on the results of sentiment analysis, the summary and translation results are adjusted to emphasize specific emotions. The input to this process is the sentiment analysis results and the translated text, and the output is the adjusted translation result. For example, the sentence may be restructured to emphasize parts with positive sentiment.

[1578] Step 6:

[1579] The server returns the final result to the user. By returning the adjusted translation result, the user can get a proper summary and translation result based on the input text, which makes it easier for users to understand long sentences and grasp the nuances of emotions in different languages.

[1580] The specific processing unit 290 transmits the result of the specific processing to the robot 414. In the robot 414, the control unit 46A causes the speaker 240 and the control target 443 to output the result of the specific processing. The microphone 238 acquires voice indicating a user input regarding the result of the specific processing. The control unit 46A transmits voice data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the voice data.

[1581] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1582] In the above embodiment, an example in which the specific processing is performed by the data processing device 12 has been given, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the robot 414.

[1583] The emotion identification model 59 as an emotion engine may determine the user's emotion according to a specific mapping. Specifically, the emotion identification model 59 may determine the user's emotion according to an emotion map (see FIG. 9), which is a specific mapping. Similarly, the emotion identification model 59 may determine the robot's emotion, and the identification processing unit 290 may perform identification processing using the robot's emotion.

[1584] FIG. 9 illustrates an emotion map 400 on which multiple emotions are mapped. In the emotion map 400, emotions are arranged in concentric circles radiating from the center. Emotions closer to the center of the concentric circles are more primitive. Emotions representing states and behaviors arising from a state of mind are arranged on the outer edges of the concentric circles. The concept of emotion includes both affect and mental states. Emotions generally generated from reactions occurring in the brain are arranged on the left side of the concentric circles. Emotions generally induced by situational judgment are arranged on the right side of the concentric circles. Emotions generally generated from reactions occurring in the brain and induced by situational judgment are arranged on the upper and lower sides of the concentric circles. Furthermore, the emotion of "pleasure" is arranged on the upper side of the concentric circles, and the emotion of "discomfort" is arranged on the lower side. In this way, in the emotion map 400, multiple emotions are mapped based on the structure by which emotions are generated, and emotions that tend to occur simultaneously are mapped close to each other.

[1585] These emotions are distributed in the 3 o'clock direction on emotion map 400, and typically fluctuate between relief and anxiety. In the right half of emotion map 400, situational awareness dominates over internal sensations, resulting in a sense of calm.

[1586] The inside of emotion map 400 represents what is going on in the mind, and the outside of emotion map 400 represents behavior, so the further you go outside emotion map 400, the more visible the emotions become (the more they are expressed in behavior).

[1587] Human emotions are based on various balances, such as posture and blood sugar levels. When these balances deviate from the ideal, a state of discomfort is indicated, and when they approach the ideal, a state of pleasure is indicated. Emotions can also be created for robots, automobiles, and motorcycles, based on various balances, such as posture and remaining battery life. When these balances deviate from the ideal, a state of discomfort is indicated, and when they approach the ideal, a state of pleasure is indicated. An emotion map can be generated, for example, based on Dr. Mitsuyoshi's emotion map (Research on Voice Emotion Recognition and Emotional Brain Physiological Signal Analysis Systems, Tokushima University, Doctoral Dissertation: https: / / ci.nii.ac.jp / naid / 500000375379). The left half of the emotion map lists emotions belonging to the "reaction" domain, where sensation is dominant. The right half of the emotion map lists emotions belonging to the "situation" domain, where situational awareness is dominant.

[1588] The emotion map defines two emotions that promote learning. One is a negative emotion on the situation side, around the middle of "repentance" or "reflection." In other words, this occurs when the robot experiences negative emotions such as "I never want to feel this way again" or "I don't want to be scolded again." The other is a positive emotion on the response side, around "desire." In other words, this occurs when the robot experiences positive feelings such as "I want more" or "I want to know more."

[1589] The emotion identification model 59 inputs user input into a pre-trained neural network, obtains emotion values ​​indicating each emotion shown in the emotion map 400, and determines the user's emotion. This neural network is pre-trained based on multiple pieces of training data that are combinations of user input and emotion values ​​indicating each emotion shown in the emotion map 400. Furthermore, this neural network is trained so that emotions that are located close to each other have similar values, as in the emotion map 900 shown in FIG. 10. FIG. 10 shows an example in which multiple emotions, "relieved," "calm," and "reassuring," have similar emotion values.

[1590] The system according to the present disclosure has been described above mainly with respect to the functions of the data processing device 12, but the system according to the present disclosure is not necessarily implemented on a server. The system according to the present disclosure may be implemented as a general information processing system. The present disclosure may be implemented, for example, as a software program running on a personal computer or an application running on a smartphone, etc. The method according to the present disclosure may be provided to users in the form of SaaS (Software as a Service).

[1591] In the above embodiment, an example was given in which the specific processing is performed by one computer 22, but the technology of the present disclosure is not limited to this, and the specific processing may be distributed and performed by a plurality of computers including the computer 22. For example, the data generation model 58 may be provided in an external device of the data processing device 12, and data may be generated in the external device in accordance with input data.

[1592] In the above embodiment, an example in which the specific processing program 56 is stored in the storage 32 has been described, but the technology of the present disclosure is not limited to this. For example, the specific processing program 56 may be stored in a portable, computer-readable, non-transitory storage medium such as a USB (Universal Serial Bus) memory. The specific processing program 56 stored in the non-transitory storage medium is installed in the computer 22 of the data processing device 12. The processor 28 executes the specific processing in accordance with the specific processing program 56.

[1593] Alternatively, the specific processing program 56 may be stored in a storage device such as a server connected to the data processing device 12 via the network 54, and the specific processing program 56 may be downloaded and installed on the computer 22 in response to a request from the data processing device 12.

[1594] It is not necessary to store all of the specific processing program 56 in a storage device such as a server connected to the data processing device 12 via the network 54, or to store all of the specific processing program 56 in the storage 32; only a portion of the specific processing program 56 may be stored.

[1595] The hardware resource for executing a specific process can be any of the following processors: An example of a processor is a CPU, which is a general-purpose processor that functions as a hardware resource for executing a specific process by executing software, i.e., a program. Another example of a processor is a dedicated electrical circuit, such as an FPGA (Field-Programmable Gate Array), a PLD (Programmable Logic Device), or an ASIC (Application Specific Integrated Circuit), which is a processor with a circuit configuration designed specifically for executing a specific process. Each processor has built-in or connected memory, and each processor uses the memory to execute the specific process.

[1596] The hardware resource that executes the specific processing may be configured with one of these various processors, or may be configured with a combination of two or more processors of the same or different types (for example, a combination of multiple FPGAs, or a combination of a CPU and an FPGA). Also, the hardware resource that executes the specific processing may be a single processor.

[1597] As an example of a system configured with a single processor, first, one processor is configured by combining one or more CPUs and software, and this processor functions as a hardware resource that executes a specific process. Second, there is a system that uses a processor that realizes the functions of an entire system including multiple hardware resources that execute a specific process on a single IC chip, as typified by SoC (System-on-a-chip). In this way, a specific process is realized using one or more of the above-mentioned various processors as hardware resources.

[1598] Furthermore, the hardware structure of these various processors can be, more specifically, an electric circuit that combines circuit elements such as semiconductor devices. The specific processing described above is merely an example. Therefore, it goes without saying that unnecessary steps may be deleted, new steps may be added, or the processing order may be rearranged, without departing from the spirit of the invention.

[1599] The above-described description and illustrations are a detailed explanation of the parts related to the technology of the present disclosure and are merely an example of the technology of the present disclosure. For example, the above description of the configuration, functions, actions, and effects is an explanation of an example of the configuration, functions, actions, and effects of the parts related to the technology of the present disclosure. Therefore, it goes without saying that unnecessary parts may be deleted, new elements may be added, or replacements may be made to the above-described description and illustrations within the scope of the gist of the technology of the present disclosure. Furthermore, to avoid confusion and facilitate understanding of the parts related to the technology of the present disclosure, the above-described description and illustrations omit explanations of common technical knowledge that do not require particular explanation to enable the implementation of the technology of the present disclosure.

[1600] All publications, patent applications, and technical standards mentioned in this specification are herein incorporated by reference to the same extent as if each individual publication, patent application, or technical standard was specifically and individually indicated to be incorporated by reference.

[1601] The following is further disclosed regarding the above embodiment.

[1602] (Claim 1)

[1603] means for receiving natural language text from a user;

[1604] means for using a pre-trained model to summarize the text;

[1605] means for translating the summarized text into different languages;

[1606] means for returning the translation result to the user;

[1607] A system including:

[1608] (Claim 2)

[1609] 10. The system of claim 1, wherein the text is summarized using a pre-trained natural language processing model.

[1610] (Claim 3)

[1611] 2. The system according to claim 1, wherein the translation uses a translation API provided by a third party.

[1612] "Example 1"

[1613] (Claim 1)

[1614] means for receiving text written in a natural language from a user;

[1615] means for using a pre-trained generative model to automatically summarize the text;

[1616] means for using a translation service to translate the summarized text into a different language;

[1617] means for returning the translation result to the user's terminal;

[1618] A system including:

[1619] (Claim 2)

[1620] 10. The system of claim 1, wherein the text is summarized using a pre-trained natural language processing model.

[1621] (Claim 3)

[1622] 2. The system according to claim 1, wherein the translation uses a translation API provided by a third party.

[1623] "Application Example 1"

[1624] (Claim 1)

[1625] means for receiving natural language text from a user;

[1626] means for using a pre-trained model to summarize the text;

[1627] means for translating the summarized text into different languages;

[1628] means for applying the summarized text and translation results to user reviews on an e-commerce site;

[1629] means for returning the translation result to the user;

[1630] A system including:

[1631] (Claim 2)

[1632] 10. The system of claim 1, wherein the text is summarized using a pre-trained natural language processing model.

[1633] (Claim 3)

[1634] 2. The system according to claim 1, wherein the translation uses a translation API provided by a third party.

[1635] "Example 2: Combining Emotion Engines"

[1636] (Claim 1)

[1637] means for receiving natural language text from a user;

[1638] means for recognizing emotions from the text;

[1639] means for using a pre-trained model to summarize the text;

[1640] means for translating the summarized text into different languages;

[1641] means for adjusting a translation result based on the recognized emotion;

[1642] means for returning the translation result to the user;

[1643] A system including:

[1644] (Claim 2)

[1645] 10. The system of claim 1, wherein the text is summarized using a pre-trained natural language processing model.

[1646] (Claim 3)

[1647] 2. The system according to claim 1, wherein the translation uses a translation API provided by a third party.

[1648] "Application example 2 when combining emotion engines"

[1649] (Claim 1)

[1650] means for receiving natural language text from a user;

[1651] means for using a pre-trained model to summarize the text;

[1652] means for translating the summarized text into different languages;

[1653] means for analyzing sentiment extracted from the text;

[1654] a means of adjusting summarization and translation results based on sentiment;

[1655] means for returning the adjusted translation result to the user;

[1656] A system including:

[1657] (Claim 2)

[1658] 10. The system of claim 1, wherein the text is summarized using a pre-trained natural language processing model.

[1659] (Claim 3)

[1660] 2. The system according to claim 1, wherein the translation uses a translation API provided by a third party. [Explanation of symbols]

[1661] 10, 210, 310, 410 Data Processing Systems 12 Data Processing Device 14 Smart Devices 214 Smart Glasses 314 Headset-type terminal 414 Robot< / url:> < / url:> < / url:> < / url:>

Claims

1. means for receiving natural language text from a user; means for using a pre-trained model to summarize the text; means for translating the summarized text into different languages; means for returning the translation result to the user; A system including:

2. The system of claim 1 , wherein the text summarization uses a pre-trained natural language processing model.

3. 2. The system according to claim 1, wherein the translation uses a translation API provided by a third party.

Images