Intelligent text generation method and device based on large model, equipment and medium

Abstract

The invention discloses an intelligent text generation method and device based on a large model, equipment and a medium, and relates to the technical field of computers. Comprising the steps of obtaining target input information, and performing natural language processing operation on the target input information to determine a key concept and a target theme corresponding to the target input information; generating original text content corresponding to the target input information by utilizing a preset large model based on the key concept and the target theme; obtaining feedback information for the original text content, and adjusting the original text content based on the feedback information to obtain target text content; and generating a file with a preset export format based on the target text content. Therefore, the text generation and analysis efficiency can be improved, and the professionality and consistency of the generated content are ensured.

Description

Technical Field

[0001] The present invention relates to the field of computer technology, and particularly to an intelligent text generation method, device, equipment and medium based on a large model. Background Art

[0002] In today's digital age, the demand for text data generation and analysis is increasing day by day, and it is widely used in many fields such as content creation, data analysis, intelligent customer service, etc. Traditional text processing methods mainly rely on manual editing and simple text replacement tools, and these methods have problems such as low efficiency, unstable quality, and difficulty in processing complex texts. In recent years, with the development of artificial intelligence and natural language processing technologies, text generation and analysis technologies based on large models have gradually emerged, but the existing technologies still have the following limitations:

[0003] Limited generation quality: The text generated by existing systems may have problems such as semantic incoherence and unclear logic, and it is difficult to meet the needs of high-quality content creation.

[0004] Insufficient analysis depth: The analysis of text mainly focuses on surface information, and it is difficult to deeply excavate deep semantic information such as implicit intentions and emotional tendencies in the text.

[0005] Poor adaptability: Existing systems have limited capabilities in text generation and analysis for different fields, and it is difficult to quickly adapt to the needs of new fields.

[0006] Lack of user interaction: Most existing systems are one-way outputs and lack interaction with users. Users cannot adjust and optimize the generation results in real time according to their needs.

[0007] As can be seen from the above, how to improve the efficiency of text generation and analysis and ensure the professionalism and consistency of the generated content is an urgent problem to be solved. Summary of the Invention

[0008] In view of this, the purpose of the present invention is to provide an intelligent text generation method, device, equipment and medium based on a large model, which can improve the efficiency of text generation and analysis and ensure the professionalism and consistency of the generated content. The specific solutions are as follows:

[0009] In a first aspect, the present application provides an intelligent text generation method based on a large model, including:

[0010] Obtain target input information, and perform natural language processing operations on the target input information to determine key concepts and a target theme corresponding to the target input information;

[0011] Generate original text content corresponding to the target input information based on the key concepts and the target theme by using a preset large model;

[0012] Obtain feedback information for the original text content, and adjust the original text content based on the feedback information to obtain the target text content;

[0013] Generate a file with a preset export format based on the target text content.

[0014] Optionally, the obtaining of the target input information includes:

[0015] Obtain the requirement information input by the target user through a preset user interaction interface, and generate corresponding target input information according to the requirement information;

[0016] Wherein, the requirement information includes text information and voice information.

[0017] Optionally, the natural language processing operation on the target input information to determine the key concepts and target theme corresponding to the target input information includes:

[0018] Perform a cleaning operation on the target input information to obtain the cleaned input information; the cleaning operation is an operation to remove irrelevant characters from the target input information and format the target input information;

[0019] Perform word segmentation and part-of-speech tagging operations on the cleaned input information to determine the grammatical framework corresponding to the target input information;

[0020] Based on the grammatical framework, use semantic understanding technology to determine the key concepts and target theme corresponding to the target input information.

[0021] Optionally, the generating of the original text content corresponding to the target input information by using a preset large model based on the key concepts and the target theme includes:

[0022] Determine the professional field corresponding to the target input information based on the key concepts and the target theme, and perform parameter adjustment and structure optimization on the preset large model based on the professional field;

[0023] Generate the original text content corresponding to the target input information by using the optimized preset large model and a preset knowledge base based on the key concepts and the target theme;

[0024] Wherein, the preset knowledge base includes professional field knowledge and preset templates, and updates the professional field knowledge and the preset templates according to a preset update time.

[0025] Optionally, after generating the original text content corresponding to the target input information by using a preset large model based on the key concepts and the target topic, the method further includes:

[0026] Evaluating and improving the semantic coherence and logical clarity of the original text content based on natural language processing techniques;

[0027] Removing redundant information from the original text content based on the target input information and adjusting the format of the original text content to complete the optimization process of the original text content.

[0028] Optionally, obtaining feedback information for the original text content and adjusting the original text content based on the feedback information to obtain the target text content includes:

[0029] Obtaining feedback information for the original text content and adjusting the original text content according to the priority of the feedback information to obtain the target text content;

[0030] Wherein, if the priority of the feedback information meets the preset priority condition, the original text content is preferentially adjusted according to the feedback information, and the corresponding adjusted content is highlighted.

[0031] Optionally, the intelligent text generation method based on a large model further includes:

[0032] Recording the feedback information and determining the feedback preference according to the feedback information;

[0033] If new target input information is obtained, performing a deep processing operation on the new target input information by using a preset large model based on the feedback preference to generate target text content that conforms to the feedback preference.

[0034] In a second aspect, the present application provides an intelligent text generation device based on a large model, including:

[0035] A target topic determination module, configured to obtain target input information and perform natural language processing operations on the target input information to determine key concepts and a target topic corresponding to the target input information;

[0036] An original text content generation module, configured to generate original text content corresponding to the target input information by using a preset large model based on the key concepts and the target topic;

[0037] An original text content adjustment module, configured to obtain feedback information for the original text content and adjust the original text content based on the feedback information to obtain the target text content;

[0038] A target text content export module, configured to generate a file with a preset export format based on the target text content.

[0039] In a third aspect, the present application provides an electronic device, including:

[0040] A memory, configured to store a computer program;

[0041] A processor, configured to execute the computer program to implement the foregoing intelligent text generation method based on a large model.

[0042] In a fourth aspect, the present application provides a computer-readable storage medium, configured to store a computer program, wherein when the computer program is executed by a processor, the foregoing intelligent text generation method based on a large model is implemented.

[0043] The present application provides an intelligent text generation method based on a large model. First, target input information is obtained, and natural language processing operations are performed on the target input information to determine key concepts and a target theme corresponding to the target input information; then, based on the key concepts and the target theme, a raw text content corresponding to the target input information is generated by using a preset large model; subsequently, feedback information for the raw text content is obtained, and the raw text content is adjusted based on the feedback information to obtain target text content; finally, a file with a preset export format is generated based on the target text content.

[0044] As can be seen from the above, through the deep learning ability of the large model, the present application deeply excavates deep semantic information such as implicit intentions and sentiment tendencies in the text, and provides more comprehensive text analysis results. And the generation result is adjusted and optimized in real time through the feedback information, improving the flexibility and practicality of the system. Thereby, the efficiency of text generation and analysis can be improved, and the professionalism and consistency of the generated content can be ensured. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings in the following description are only the embodiments of the present invention, and for those of ordinary skill in the art, other drawings can be obtained according to the provided drawings without creative efforts.

[0046] Figure 1 It is a flowchart of an intelligent text generation method based on a large model provided by the present application;

[0047] Figure 2 It is a system architecture diagram of a specific intelligent text generation method based on a large model provided by the present application;

[0048] Figure 3 A flowchart of a specific intelligent text generation method provided for this application based on a large model;

[0049] Figure 4 A schematic diagram of an intelligent text generation device provided for this application based on a large model;

[0050] Figure 5 A structural diagram of an electronic device provided for this application. Detailed implementation manners

[0051] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0052] In today's digital age, the demand for text data generation and analysis is increasing day by day, and it is widely used in many fields such as content creation, data analysis, intelligent customer service, etc. Traditional text processing methods mainly rely on manual editing and simple text replacement tools, and these methods have problems such as low efficiency, unstable quality, and difficulty in processing complex texts. In recent years, with the development of artificial intelligence and natural language processing technologies, text generation and analysis technologies based on large models have gradually emerged, but the existing technologies still have limitations including limited generation quality, insufficient analysis depth, poor adaptability, and lack of user interaction. For this reason, this application provides an intelligent text generation solution based on a large model, which can improve the efficiency of text generation and analysis and ensure the professionalism and consistency of the generated content.

[0053] See Figure 1 As shown, an intelligent text generation method based on a large model disclosed in an embodiment of this application includes:

[0054] Step S11, obtain target input information, and perform natural language processing operations on the target input information to determine key concepts and a target theme corresponding to the target input information.

[0055] In this embodiment, the target input information of the target user is obtained through a preset user interaction interface, and the target input information is preprocessed to facilitate natural language processing operations on the target input information. Among them, the specific type of the preset user interaction interface can be determined according to the actual application situation, including but not limited to mobile APPs, web pages, and mini-programs. For example, in a specific implementation, a mobile APP is selected as the preset user interaction interface. Specifically, the obtaining of the target input information may include: obtaining the demand information input by the target user through the mobile APP, and generating corresponding target input information according to the demand information. It can be understood that the specific type of the demand information can be determined according to the actual usage situation, including but not limited to text information and voice information.

[0056] Further, natural language processing operations are performed on the obtained target input information to determine the key concepts and target themes corresponding to the target input information. Specifically, the performing of natural language processing operations on the target input information to determine the key concepts and target themes corresponding to the target input information may include: performing a cleaning operation on the target input information to obtain the cleaned input information; the cleaning operation is an operation to remove irrelevant characters in the target input information and format the target input information; performing word segmentation and part-of-speech tagging operations on the cleaned input information to determine the grammatical framework corresponding to the target input information; based on the grammatical framework, using semantic understanding technology to determine the key concepts and target themes corresponding to the target input information. That is, first, the target input information is cleaned to remove irrelevant characters, extra spaces, and formatting errors in it to improve the neatness and consistency of the text. Then, the cleaned target input information is subjected to word segmentation processing to accurately split the target input information into independent lexical units. Subsequently, part-of-speech tagging is performed on each word to clarify its grammatical role in the sentence, such as noun, verb, adjective, etc., thereby constructing the grammatical framework of the text. Finally, through semantic understanding technology, the inherent meaning of the text is deeply explored, and important information such as key concepts, themes, and implicit sentiment tendencies is extracted, providing solid data support for the in-depth analysis and efficient generation of the text. By preprocessing the target input information, the quality of the text is optimized, laying a solid foundation for the subsequent intelligent processing process of the system and ensuring the accuracy and efficiency of text processing.

[0057] Step S12: Based on the key concepts and the target theme, use a preset large model to generate the original text content corresponding to the target input information.

[0058] In this embodiment, according to the input requirements of the user and the characteristics of a specific field, the pre-trained large model is fine-tuned. The fine-tuning process includes adjusting the model's parameters, optimizing the model's structure, adding domain-specific training data, etc., so that the model can better adapt to specific tasks and fields.

[0059] It is worth mentioning that in this embodiment, a comprehensive domain knowledge database is constructed and maintained. This database covers the professional knowledge and information of multiple industries. These fields include but are not limited to technology, medicine, law, finance, education, etc. By integrating this rich domain knowledge, the system can provide detailed reference information for the large model, thus ensuring the professionalism and accuracy of the output content when processing text generation and analysis tasks in specific fields. For example, when processing medical texts, the system can cite professional knowledge such as medical terms, disease information, treatment plans, etc., to generate high-quality medical texts. Specifically, the generating the original text content corresponding to the target input information by using the pre-trained large model based on the key concepts and the target theme may include: determining the professional field corresponding to the target input information based on the key concepts and the target theme, and performing parameter adjustment and structure optimization on the pre-trained large model based on the professional field; generating the original text content corresponding to the target input information by using the optimized pre-trained large model and the domain knowledge database based on the key concepts and the target theme. Further, a variety of predefined templates are also stored in the domain knowledge database. These templates cover common text types, such as news reports, research reports, business plans, academic papers, etc. Users can select appropriate templates according to specific needs, and the system will generate or analyze texts based on the selected templates. The template management function not only improves the speed of text generation, but also ensures that the structure and format of the generated texts conform to industry standards. For example, users can select an academic paper template, and the system will automatically generate a paper framework including abstract, introduction, methods, results, discussion, etc., and fill in relevant content.

[0060] It should be noted that in order to maintain the timeliness and accuracy of the knowledge base, the knowledge base has a real-time update function. Specifically, according to the user's input and the system operation situation, the knowledge base will automatically update its content. For example, when the user inputs new domain knowledge or updates existing information, the system will immediately integrate this information into the knowledge base. In addition, the system will regularly obtain the latest information from reliable external data sources to ensure that the data in the knowledge base is always up-to-date. This real-time update mechanism enables the system to quickly adapt to the ever-changing domain knowledge and user needs, and provide always accurate and up-to-date text generation and analysis services.

[0061] In this embodiment, after generating the text content, the text will be further optimized. This process includes evaluating and improving the semantic coherence and logical clarity of the text. Specifically, after generating the original text content corresponding to the target input information using a preset large model based on the key concepts and the target theme, it may further include: evaluating and improving the semantic coherence and logical clarity of the original text content based on natural language processing techniques; removing redundant information from the original text content based on the target input information and adjusting the format of the original text content to complete the optimization process of the original text content. That is, through natural language processing techniques, the structure and semantics of the text are analyzed to ensure that the generated text content is logically coherent and well-organized, avoiding problems such as semantic repetition or logical jumps.

[0062] Step S13, obtain feedback information for the original text content, and adjust the original text content based on the feedback information to obtain the target text content.

[0063] In this embodiment, to improve the user experience and work efficiency, a real-time feedback function is provided. After the user inputs the requirements, the text generation and analysis process will be immediately started, and the results will be displayed in real time. The user can view the results immediately during the generation process and make adjustments as needed. This real-time interaction method not only improves user participation but also ensures that the generated content fully meets the user's needs. Specifically, the obtaining of the feedback information for the original text content and the adjustment of the original text content based on the feedback information to obtain the target text content may include: obtaining the feedback information for the original text content and adjusting the original text content according to the priority of the feedback information to obtain the target text content; wherein, if the priority of the feedback information meets the preset priority condition, the original text content will be preferentially adjusted according to the feedback information, and the corresponding adjusted content will be highlighted. That is, by introducing a feedback priority mechanism, the feedback content can be intelligently sorted according to the urgency and importance of the user's feedback. When submitting feedback, the user can select the priority of the feedback (such as high, medium, low), and the system will classify and process the feedback according to the priority. For high-priority feedback, the system will preferentially adjust the generated result and highlight the adjusted part in the interface. This feedback priority mechanism can ensure that the system preferentially processes the problems that the user considers the most important, improving the system's response efficiency and user experience. At the same time, the intelligent sorting function can help the user quickly locate the parts that the system has adjusted, avoiding the user having to search through the long text one by one.

[0064] Furthermore, automatically record the user's feedback history and provide intelligent recommendations based on the history. For example, if the user mentions a preference for a specific theme or style in multiple feedbacks, the system will automatically adjust the generation strategy when generating subsequent content to meet the user's preference. At the same time, the system will also provide feedback templates in similar scenarios based on the feedback history to help the user express their needs more quickly. Specifically, the intelligent text generation method based on the large model may further include: recording the feedback information and determining the feedback preference according to the feedback information; if new target input information is obtained, then based on the feedback preference, use a preset large model to perform a deep processing operation on the new target input information to generate target text content that meets the feedback preference. That is, the feedback history record and the intelligent recommendation function can help the system better understand the user's long-term needs and provide more personalized services. The user does not need to elaborate on their needs each time, and the generated content can be automatically adjusted according to the history record, improving the user's usage efficiency and satisfaction.

[0065] Step S14, generate a file with a preset export format based on the target text content.

[0066] In this embodiment, a flexible result export function is provided, and the user can export the result into various common file formats, such as TXT, PDF, Word, etc. These formats are widely used in scenarios such as document editing, printing, and online sharing, facilitating the user to select a suitable format for subsequent use according to their needs. The design of the export function fully considers the diverse needs of the user and ensures the portability and operability of the generated content.

[0067] As can be seen above, in the embodiments of the present application, the target input information is first cleaned to remove irrelevant characters, extra spaces, and formatting errors therein, so as to improve the neatness and consistency of the text. Then, the cleaned target input information is tokenized to accurately split the target input information into independent lexical units. Subsequently, part-of-speech tagging is performed on each word to clarify its grammatical role in the sentence, such as noun, verb, adjective, etc., thereby constructing the grammatical framework of the text. Finally, through semantic understanding technology, the inherent meaning of the text is deeply explored, and important information such as key concepts, themes, and implicit sentiment tendencies is extracted, providing solid data support for the in-depth analysis and efficient generation of the text. By performing preprocessing operations on the target input information, the quality of the text is optimized, laying a solid foundation for the subsequent intelligent processing flow of the system and ensuring the accuracy and efficiency of text processing. A comprehensive domain knowledge database is constructed and maintained, which covers professional knowledge and information in multiple industries. These domains include but are not limited to technology, medicine, law, finance, education, etc. By integrating this rich domain knowledge, the system can provide detailed reference information for the large model, so as to ensure the professionalism and accuracy of the output content when processing text generation and analysis tasks in specific domains. Through the deep learning ability of the large model, the deep semantic information such as implicit intentions and sentiment tendencies in the text is deeply explored, providing more comprehensive text analysis results. And the generated results are adjusted and optimized in real time through the feedback information, improving the flexibility and practicality of the system. Thereby, the efficiency of text generation and analysis can be improved, and the professionalism and consistency of the generated content can be ensured.

[0068] Furthermore, the embodiments of the present application disclose a specific intelligent text generation method based on a large model, including:

[0069] See Figure 2 As shown, in this embodiment, the user interaction module, text preprocessing module, large model core module, result optimization module, and knowledge base module are divided according to the module functions.

[0070] The user interaction module designs an intuitive and easy-to-operate user interface through which users can input requirements in various ways. In addition to the traditional text input box, voice input is also supported, facilitating users to quickly express their requirements in different scenarios. The interface design is simple and the operation process is clear, enabling even first-time users to get started easily. Moreover, the system provides detailed usage guides and help documents to assist users in better understanding and using each function. To enhance the user experience and work efficiency, the system has a real-time feedback function. After the user inputs requirements, the system immediately initiates the text generation and analysis process and displays the results in real time. Users can view the results instantaneously during the generation process and make adjustments as needed. This real-time interaction method not only increases user engagement but also ensures that the generated content fully meets the user's requirements. The system also supports a real-time annotation function, allowing users to add comments or suggestions to any part of the generated results to further optimize the generated content. Considering that users may need to use the generation and analysis results in multiple scenarios, the system provides a flexible result export function. Users can export the results in various common file formats, such as TXT, PDF, Word, etc. These formats are widely used in scenarios such as document editing, printing, and online sharing, enabling users to select the appropriate format for subsequent use according to their needs. The design of the export function fully considers the diverse needs of users, ensuring the portability and operability of the generated content.

[0071] In the text preprocessing stage, the text preprocessing module first performs cleaning operations on the input text. This process includes removing irrelevant characters in the text, such as extra spaces, punctuation marks, special characters, etc., and formatting the text to ensure its neatness and consistency. Through these operations, the quality of the text can be effectively improved, laying a solid foundation for subsequent processing and analysis. For example, the system will automatically identify and remove HTML tags, extra line breaks and spaces in the text, making the text more standardized and easier to process. The cleaned text will be sent to the word segmentation module for processing. Word segmentation is a key step in natural language processing, which divides a continuous text string into individual lexical units. By using advanced word segmentation algorithms, the lexical boundaries of multiple languages such as Chinese and English can be accurately identified. At the same time of word segmentation, part-of-speech tagging is also performed on each word, that is, the part of speech of each word is marked, such as noun, verb, adjective, etc. Part-of-speech tagging provides an important basis for subsequent semantic analysis and text understanding. For example, the system can identify that "apple" is a noun and "eat" is a verb, so as to better understand the structure and meaning of the sentence. On the basis of word segmentation and part-of-speech tagging, semantic understanding is further carried out. The semantic information of the text is initially understood, and the key concepts and themes in the text are extracted. By analyzing the relationships between words, context information and semantic features, the system can identify the core content and main topics of the text. For example, in an article about artificial intelligence, the system can extract key concepts such as "artificial intelligence", "machine learning", "deep learning", and topics such as "technology development", "application cases", "future trends". This semantic understanding ability enables the system to analyze the text more deeply and provide more valuable information for users.

[0072] When the user inputs specific requirements through the interactive interface, the core large model module will utilize the powerful generation ability of the large model to quickly generate text that meets the user's needs. Whether it is writing an article, generating a report, creating a story, or writing code comments, the large model can generate smooth, coherent, and high-quality text content according to the requirements provided by the user, such as the theme, style, word count, etc. For example, the user can specify to generate a technology article on "The Application of Artificial Intelligence in the Medical Field", and the system will be able to generate a high-quality article with a complete structure, rich content, and accurate language. The core large model module can not only generate text but also conduct in-depth analysis of the text content. Through natural language processing technology, it can extract deep semantic content such as implicit intentions, sentiment tendencies, and theme information in the text. For example, when analyzing user feedback or customer reviews, the system can identify the positive or negative sentiment tendencies in the reviews, extract the key issues and suggestions that users are concerned about, thereby helping enterprises and organizations better understand customer needs and optimize products and services. In addition, the core large model module can also identify the logical structure in the text, analyze the coherence and consistency of the text, and provide an important basis for text quality assessment. To further improve the adaptability and accuracy of the core large model module, the core large model module has a model fine-tuning function. According to the user's input requirements and the characteristics of a specific domain, the core large model module will fine-tune the large model. The fine-tuning process includes adjusting the model's parameters, optimizing the model's structure, adding domain-specific training data, etc., to make the model better adapt to specific tasks and domains. For example, when processing legal texts, the system will load professional legal vocabulary and corpora and conduct targeted training on the model to improve the accuracy and professionalism of generating and analyzing legal texts. Through this flexible fine-tuning mechanism, the system can quickly adapt to changes in different domains and user needs and provide more accurate and efficient services.

[0073] The result optimization module will further optimize the text. This process includes evaluating and improving the semantic coherence and logical clarity of the text. The result optimization module analyzes the structure and semantics of the text through natural language processing techniques to ensure that the generated text content is logically coherent and well-organized, avoiding problems such as semantic repetition or logical leaps. In addition, the result optimization module will also optimize the keywords and phrases in the text to make them more in line with the user's expression habits and context requirements. To ensure the practicality and pertinence of the generated results, the result optimization module will screen out the most relevant and important information based on the user's needs and text analysis results. This process involves in-depth analysis of the text content, identifying the parts highly relevant to the user's needs, and removing redundant or irrelevant information. Through this screening mechanism, users can quickly obtain the most valuable information and improve work efficiency. The result optimization module also has a flexible format adjustment function, which can format the generated results according to the output format specified by the user. Whether it is common document formats such as TXT, PDF, Word, or specific format requirements, the result optimization module can automatically adjust the format of the text to ensure that the generated results meet the user's usage needs. For example, users can specify that the generated text needs to include format elements such as titles, subtitles, headers, and footers, and the system will automatically typeset according to these requirements. In addition, the system also supports users to customize format templates. Users can set format parameters such as font, font size, and line spacing of the text according to their preferences to generate personalized documents.

[0074] The knowledge base module covers professional knowledge and information in multiple industries. These fields include but are not limited to technology, healthcare, law, finance, education, etc. By integrating this rich domain knowledge, the knowledge base module can provide detailed reference information for the large model, thus ensuring the professionalism and accuracy of the output content when processing text generation and analysis tasks in specific fields. To improve the efficiency of text generation and analysis, the knowledge base module stores a variety of predefined templates. These templates cover common text types such as news reports, research reports, business plans, academic papers, etc. Users can select appropriate templates according to specific needs, and the system will generate or analyze text based on the selected templates. The template management function not only improves the speed of text generation but also ensures that the structure and format of the generated text meet industry standards. To maintain the timeliness and accuracy of the knowledge base, the knowledge base module has a real-time update function. According to the user's input and system operation status, the knowledge base will automatically update its content. In addition, the knowledge base module will regularly obtain the latest information from reliable external data sources to ensure that the data in the knowledge base is always up-to-date. This real-time update mechanism enables the system to quickly adapt to the changing domain knowledge and user needs, providing always accurate and up-to-date text generation and analysis services.

[0075] SeeFigure 3 As shown in the figure, users can input their specific requirements for text generation or analysis through a carefully designed interactive interface. This interface provides users with a concise and powerful platform, allowing users to easily express their ideas and requirements. Users can specify the topic of the text in detail, such as "the application of artificial intelligence in the medical field" or "the impact of climate change on the global economy", to ensure that the generated or analyzed text is closely centered on the core topic. At the same time, users can also choose the desired text style, such as formal, informal, academic, business or creative, to match different usage scenarios and target audiences. In addition, users can set a word limit for the text, whether it is a short summary or a long report, the system can accurately generate or analyze according to the user's needs. Through these meticulous demand input options, users can obtain highly customized text processing services to meet their diverse needs in many fields such as content creation, data analysis, and intelligent customer service.

[0076] As the basic link of the system, this preprocessing module undertakes the important task of preliminary sorting and analysis of the input text. It first cleans the text, removes irrelevant characters, extra spaces and formatting errors, so as to improve the neatness and consistency of the text. Then, the module will perform word segmentation on the cleaned text, accurately dividing the text into independent vocabulary units to prepare for subsequent analysis. On the basis of word segmentation, each word is further tagged with parts of speech to clarify its grammatical role in the sentence, such as nouns, verbs, adjectives, etc., so as to build the grammatical framework of the text. Finally, through semantic understanding technology, the intrinsic meaning of the text is deeply explored, and important information such as key concepts, themes and implicit emotional tendencies are extracted, providing solid data support for the in-depth analysis and efficient generation of the text. This series of preprocessing operations not only optimizes the quality of the text, but also lays a solid foundation for the subsequent intelligent processing flow of the system, ensuring the accuracy and efficiency of text processing.

[0077] The core module of the large model is the core component of this system. Based on the preliminary arrangement completed by the text preprocessing module, it further conducts in-depth processing according to the detailed requirements of users. This module utilizes advanced natural language processing technologies and the powerful capabilities of large-scale pre-trained models to accurately analyze and understand the preprocessed text. It can not only generate high-quality, semantically coherent, and logically clear text content to meet the diverse needs of users in different scenarios, such as writing articles, generating reports, or creating stories, but also deeply explore the implicit intentions, sentiment tendencies, theme information, and other deep semantic contents in the text. Through this in-depth analysis, the system can extract the core points and key information of the text, providing users with more valuable insights and analysis results. This comprehensive ability enables the core module of the large model to play an important role in multiple fields such as content creation, data analysis, and intelligent customer service, significantly improving the efficiency and quality of text processing.

[0078] The result optimization module plays a crucial role in the text generation and analysis process. It conducts refined processing on the preliminary results produced by the core module of the large model. This module first evaluates and improves the semantic coherence and logical clarity of the generated text content. Through natural language processing technologies, it adjusts the sentence structure and optimizes the word choice to ensure that the text is more fluent and natural in expression and more rigorous and orderly in logic. At the same time, the result optimization module will screen out the most relevant and important information according to the specific needs of users and the text analysis results, remove redundant content, and highlight the core points, thereby improving the relevance and practicality of the text. In addition, this module also has a flexible format adjustment function. It can accurately format the generated results according to the specified output format requirements of users, such as document types (TXT, PDF, Word, etc.), font styles, layout arrangements, etc., to ensure that the final output text fully meets the user's usage scenarios and preferences. Through this series of optimization operations, the result optimization module significantly improves the quality and usability of the text content, providing users with a high-quality and personalized text processing result.

[0079] The knowledge base module plays a crucial auxiliary role in this system. It provides the large model with rich domain knowledge and diverse template support, thus significantly enhancing the accuracy and adaptability of text generation and analysis. By constructing and maintaining a professional knowledge database covering multiple industries, the system can ensure that when the large model processes text in a specific field, it can reference accurate and relevant background information. For example, when processing medical text, the system can call on professional knowledge such as medical terms, disease information, and treatment plans to generate high-quality medical text. In addition, the knowledge base stores various templates for text generation and analysis. Users can select appropriate templates according to specific needs, and the system will generate or analyze text based on the selected templates to ensure that the structure and format of the output content meet industry standards. This template management function not only improves the speed of text generation but also ensures the professionalism and consistency of the generated text. At the same time, the knowledge base module has a real-time update function. It can update the content of the knowledge base in a timely manner according to user input and system operation conditions, maintaining the timeliness and accuracy of the knowledge base. This dynamic update mechanism enables the system to quickly adapt to changing domain knowledge and user needs, providing users with always accurate and up-to-date text processing services.

[0080] User interaction and adjustment is a highly innovative and practical function module in this system. Through a carefully designed interactive interface, it provides users with an intuitive and powerful operation platform. Users can clearly view the text results generated and analyzed by the system on this interface. These results are presented in an easy-to-understand manner, and whether it is text content, analysis charts, or key information summaries, they are all clearly visible. More importantly, users can make real-time adjustments and optimizations to these results according to their specific needs and preferences. For example, users can modify the style of the text, adjust the depth and breadth of the content, or customize the presentation method of the analysis results. The system will immediately respond to the user's operations and quickly update the results to ensure that users can obtain text content and analysis reports that best meet their needs. This high degree of interactivity and flexibility not only greatly enhances the user experience but also enables the system to better adapt to the diverse needs of different users in many fields such as content creation, data analysis, and intelligent customer service, truly realizing a user-centered intelligent text processing service.

[0081] At the final stage of the text generation and analysis process in this system, users can export the final results that have been carefully optimized and adjusted into files in a specified format. This process not only provides users with a convenient way to save and share, but also ensures the operability and portability of the results. The system supports a variety of common file formats, including but not limited to TXT, PDF, Word, etc. Users can choose the most suitable format for export according to their own needs and usage scenarios. For example, if users need to use the generated text for a printed report, they can choose to export it in PDF format to maintain format stability; if further editing is required, they can choose Word format. The design of the export function fully considers the diverse needs of users, enabling users to easily integrate the text and analysis results generated by the system into their work processes, and seamlessly connect for archiving, sharing, or further editing and processing. Through this efficient export mechanism, users can successfully complete the entire text generation and analysis process, transforming the powerful functions of the system into actual productivity and value.

[0082] As can be seen from the above, through the deep learning ability of the large model and the result optimization module in the embodiments of this application, high-quality, semantically coherent, and logically clear text content is generated. Deeply excavate the deep semantic information such as implicit intentions and sentiment tendencies in the text, and provide more comprehensive text analysis results. Through the knowledge base module and model fine-tuning technology, quickly adapt to the text generation and analysis needs in different fields. Provide an interactive user interface, where users can adjust and optimize the generation results in real time according to their needs, improving the flexibility and practicality of the system.

[0083] See Figure 4 As shown, the embodiments of this application disclose an intelligent text generation device based on a large model, including:

[0084] A target theme determination module 11, configured to obtain target input information and perform natural language processing operations on the target input information to determine key concepts and a target theme corresponding to the target input information;

[0085] An original text content generation module 12, configured to generate original text content corresponding to the target input information based on the key concepts and the target theme by using a preset large model;

[0086] An original text content adjustment module 13, configured to obtain feedback information for the original text content and adjust the original text content based on the feedback information to obtain target text content;

[0087] A target text content export module 14, configured to generate a file with a preset export format based on the target text content.

[0088] As can be seen from the above, through the deep learning ability of the large model, the embodiments of the present application deeply explore the implicit intentions, sentiment tendencies and other deep semantic information in the text, providing more comprehensive text analysis results. And through the feedback information, the generated results are adjusted and optimized in real time to improve the flexibility and practicality of the system. Thereby, the efficiency of text generation and analysis can be improved, ensuring the professionalism and consistency of the generated content.

[0089] In some specific embodiments, the target theme determination module 11 may specifically include:

[0090] A target input information acquisition unit, configured to acquire the requirement information input by the target user through a preset user interface, and generate corresponding target input information according to the requirement information; the requirement information includes text information and voice information;

[0091] A target input information cleaning unit, configured to perform a cleaning operation on the target input information to obtain the cleaned input information; the cleaning operation is an operation of removing irrelevant characters from the target input information and formatting the target input information;

[0092] A syntax framework determination unit, configured to perform word segmentation and part-of-speech tagging operations on the cleaned input information to determine the syntax framework corresponding to the target input information;

[0093] A target theme generation unit, configured to determine the key concepts and target theme corresponding to the target input information based on the syntax framework by using semantic understanding technology.

[0094] In some specific embodiments, the original text content generation module 12 may specifically include:

[0095] A preset large model optimization unit, configured to determine the professional field corresponding to the target input information based on the key concepts and the target theme, and perform parameter adjustment and structure optimization on the preset large model based on the professional field;

[0096] An original text content generation unit, configured to generate the original text content corresponding to the target input information based on the key concepts and the target theme by using the optimized preset large model and a preset knowledge base; the preset knowledge base includes professional field knowledge and preset templates;

[0097] A preset knowledge base update unit, configured to update the professional field knowledge and the preset templates according to a preset update time.

[0098] In some specific embodiments, the original text content adjustment module 13 may specifically include:

[0099] A feedback information acquisition unit, configured to acquire feedback information for the original text content, and adjust the original text content according to the priority of the feedback information to obtain target text content; if the priority of the feedback information meets a preset priority condition, adjust the original text content for the feedback information preferentially, and highlight the corresponding adjusted content.

[0100] A feedback information recording unit, configured to record the feedback information and determine a feedback preference according to the feedback information; if new target input information is acquired, perform a deep processing operation on the new target input information by using a preset large model based on the feedback preference to generate target text content that conforms to the feedback preference.

[0101] In some specific embodiments, the large model-based intelligent text generation device may further include:

[0102] An original text content optimization unit, configured to evaluate and improve the semantic coherence and logical clarity of the original text content based on natural language processing techniques.

[0103] An original text content adjustment unit, configured to eliminate redundant information in the original text content based on the target input information and adjust the format of the original text content to complete the optimization process of the original text content.

[0104] Furthermore, an embodiment of the present application also discloses an electronic device. Figure 5 It is a structural diagram of an electronic device 20 shown according to an exemplary embodiment, and the content in the figure should not be considered as any limitation to the scope of use of the present application. The electronic device 20 may specifically include: at least one processor 21, at least one memory 22, a power supply 23, a communication interface 24, an input / output interface 25, and a communication bus 26. Among them, the memory 22 is used to store a computer program, and the computer program is loaded and executed by the processor 21 to implement the relevant steps in the large model-based intelligent text generation method disclosed in any of the foregoing embodiments. In addition, the electronic device 20 in this embodiment may specifically be an electronic computer.

[0105] In this embodiment, the power supply 23 is used to provide operating voltages for each hardware device on the electronic device 20; the communication interface 24 can create a data transmission channel between the electronic device 20 and external devices, and the communication protocol it follows is any communication protocol applicable to the technical solution of the present application, and no specific limitation is imposed on it here; the input / output interface 25 is used to acquire external input data or output data to the outside, and its specific interface type can be selected according to specific application requirements, and no specific limitation is made here.

[0106] In addition, as a carrier for storing resources, the memory 22 can be a read-only memory, a random access memory, a magnetic disk, an optical disc, etc. The resources stored thereon can include an operating system 221, a computer program 222, etc. The storage method can be temporary storage or permanent storage.

[0107] Among them, the operating system 221 is used to manage and control each hardware device and the computer program 222 on the electronic device 20, and it can be Windows Server, Netware, Unix, Linux, etc. In addition to the computer program that can be used to complete the large model-based intelligent text generation method executed by the electronic device 20 disclosed in any of the foregoing embodiments, the computer program 222 can further include computer programs that can be used to complete other specific tasks.

[0108] Furthermore, the present application also discloses a computer-readable storage medium for storing a computer program; wherein, when the computer program is executed by a processor, it implements the foregoing disclosed large model-based intelligent text generation method. For the specific steps of this method, reference can be made to the corresponding content disclosed in the foregoing embodiments, and details will not be repeated here.

[0109] In this specification, each embodiment is described in a progressive manner. The key point of each embodiment is to illustrate the differences from other embodiments. The same or similar parts among the embodiments can be referred to each other. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the description of the method part.

[0110] Those skilled in the art can further realize that the units and algorithm steps of each example described in combination with the embodiments disclosed herein can be implemented by electronic hardware, computer software, or a combination of the two. To clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described according to functions in the above description. Whether these functions are executed in a hardware or software manner depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of the present application.

[0111] The steps of the method or algorithm described in combination with the embodiments disclosed herein can be directly implemented by hardware, a software module executed by a processor, or a combination of the two. The software module can be placed in a random access memory (RAM), internal memory, read-only memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, register, hard disk, removable disk, CD-ROM, or any other form of storage medium well-known in the technical field.

[0112] Finally, it should also be noted that in this text, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the term "comprising", "including" or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, article or device comprising a series of elements not only includes those elements, but also includes other elements not expressly listed, or further includes elements inherent to such process, method, article or device. Without further limitation, an element defined by the statement "comprising an..." does not exclude the existence of additional identical elements in the process, method, article or device comprising the said element.

[0113] The technical solutions provided in this application have been introduced in detail above. Specific examples are used in this text to elaborate on the principles and implementation manners of this application. The description of the above embodiments is only used to help understand the method and its core idea of this application; at the same time, for those of ordinary skill in the art, according to the idea of this application, there will be changes in the specific implementation manners and application scopes. In summary, the content of this specification should not be construed as a limitation to this application.

Claims

1. An intelligent text generation method based on a large model, characterized in that, Including: Obtain target input information, and perform natural language processing operations on the target input information to determine key concepts and a target theme corresponding to the target input information; Generate original text content corresponding to the target input information based on the key concepts and the target theme using a preset large model; Obtain feedback information for the original text content, and adjust the original text content based on the feedback information to obtain target text content; Generate a file with a preset export format based on the target text content.

2. The intelligent text generation method based on a large model according to claim 1, wherein The obtaining of the target input information includes: Obtain requirement information input by a target user through a preset user interface, and generate corresponding target input information according to the requirement information; Wherein, the requirement information includes text information and voice information.

3. The intelligent text generation method based on a large model according to claim 1, wherein The performing of natural language processing operations on the target input information to determine key concepts and a target theme corresponding to the target input information includes: Perform a cleaning operation on the target input information to obtain cleaned input information; the cleaning operation is an operation to remove irrelevant characters from the target input information and format the target input information; Perform word segmentation and part-of-speech tagging operations on the cleaned input information to determine a grammar framework corresponding to the target input information; Determine key concepts and a target theme corresponding to the target input information based on the grammar framework using semantic understanding technology.

4. The intelligent text generation method based on a large model according to claim 1, wherein The generating of original text content corresponding to the target input information based on the key concepts and the target theme using a preset large model includes: Determine the professional field corresponding to the target input information based on the key concepts and the target theme, and perform parameter adjustment and structure optimization on the preset large model based on the professional field; Generate original text content corresponding to the target input information based on the key concepts and the target theme using the optimized preset large model and a preset knowledge base; Wherein, the preset knowledge base includes professional field knowledge and preset templates, and updates the professional field knowledge and the preset templates according to a preset update time.

5. The intelligent text generation method based on a large model according to claim 1, characterized in that, After generating the original text content corresponding to the target input information based on the key concepts and the target theme, it further includes: Evaluate and improve the semantic coherence and logical clarity of the original text content based on natural language processing technology; Remove redundant information from the original text content based on the target input information, and adjust the format of the original text content to complete the optimization process of the original text content.

6. The intelligent text generation method based on a large model according to any one of claims 1 to 5, characterized in that, The obtaining of feedback information for the original text content and adjusting the original text content based on the feedback information to obtain target text content includes: Obtain feedback information for the original text content, and adjust the original text content according to the priority of the feedback information to obtain target text content; Among them, if the priority of the feedback information meets the preset priority condition, the original text content is preferentially adjusted according to the feedback information, and the corresponding adjusted content is highlighted.

7. The intelligent text generation method based on a large model according to claim 6, wherein It further includes: Recording the feedback information and determining the feedback preference according to the feedback information; If new target input information is obtained, the new target input information is deeply processed based on the feedback preference using a preset large model to generate target text content that conforms to the feedback preference.

8. An intelligent text generation device based on a large model, characterized in that, It includes: A target theme determination module, configured to obtain target input information and perform natural language processing operations on the target input information to determine the key concepts and target theme corresponding to the target input information; An original text content generation module, configured to generate original text content corresponding to the target input information based on the key concepts and the target theme using a preset large model; An original text content adjustment module, configured to obtain feedback information for the original text content and adjust the original text content based on the feedback information to obtain target text content; A target text content export module, configured to generate a file with a preset export format based on the target text content.

9. An electronic device, characterized in that, It includes: A memory for storing a computer program; A processor for executing the computer program to implement the large model-based intelligent text generation method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, For storing a computer program, wherein the computer program, when executed by the processor, implements the large model-based intelligent text generation method according to any one of claims 1 to 7.

Images