A multi-version text simplification method based on ChatGPT

By comparing the readability difference values ​​of text simplification results with the original text, and continuing to simplify when the difference value exceeds the threshold, combining Hugging Face's Transformer library and discriminator, the problem that the traditional ChatGPT model cannot provide multi-version text simplification is solved, and the implementation of High-quality text simplification for users with different reading abilities.

CN118734806BActive Publication Date: 2024-11-05YUNNAN NORMAL UNIV
View PDF 5 Cites 0 Cited by

Patent Information

Application Number
CN202411228090.X
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-09-03
Publication Date
2024-11-05
Estimated Expiration
2044-09-03

AI Technical Summary

Technical Problem

The traditional ChatGPT model can only generate a single version of simplified text, which cannot meet the needs of users with different reading abilities, and cannot provide multi-version text simplification.

Method used

By determining the readability difference between the text simplification result and the original text, compare with the preset difference threshold, and continue simplifying the text until the difference value is less than the threshold, use Hugging Face's Transformer library and discriminator to ensure simplification Quality of results.

Benefits of technology

Multi-version text simplification is achieved for users with different reading capabilities, ensuring the quality of text simplification and output reliability, and providing highly adaptable text simplification results.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118734806B_ABST
    Figure CN118734806B_ABST
Patent Text Reader

Abstract

The present application relates to the field of natural language processing technology, and in particular to a multi-version text simplification method based on ChatGPT. The method comprises: S1, determining the readability difference value between the current text simplification result and the original text, wherein the current text simplification result is generated by ChatGPT or a text simplifier model; S2, comparing the readability difference value with a preset difference threshold; S3, if it is less than or equal to the preset difference threshold, outputting the current text simplification result; S4, otherwise, simplifying the current text simplification result through the text simplifier, and returning the obtained new current text simplification result to step S1. It aims to solve the problem of how to provide multi-version text simplification for users with different reading abilities.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application relates to the technical field of natural language processing, and in particular to a multi-version text simplification method based on ChatGPT. Background Art

[0002] Text simplification is a language processing process that aims to transform complex, lengthy or professional text content into a more concise and understandable form. With the continuous development of natural language processing technology, text simplification can now be achieved through the ChatGPT model. When using ChatGPT for text simplification, users can enter clear prompts so that ChatGPT can understand and respond to user needs.

[0003] However, in the process of conceiving and implementing this application, the inventor discovered that different reader groups have different reading abilities, but the traditional ChatGPT model can only generate a single version of simplified text, which cannot meet the needs of different reader groups. Therefore, there is a need for a text simplification method that can provide multiple versions of text simplification for users with different reading abilities based on the ChatGPT model to improve the adaptability of text simplification. Summary of the invention

[0004] The main purpose of this application is to provide a multi-version text simplification method based on ChatGPT, aiming to solve the problem of how to provide multi-version text simplification for users with different reading abilities.

[0005] To achieve the above objectives, the present application provides a multi-version text simplification method based on ChatGPT, the method comprising:

[0006] S1, determining a readability difference value between a current text simplification result and an original text, wherein the current text simplification result is generated by ChatGPT or a text simplifier model;

[0007] S2, comparing the readability difference value with a preset difference threshold;

[0008] S3, if it is less than or equal to the preset difference threshold, output the current text simplification result;

[0009] S4: otherwise, simplify the current text simplification result by the text simplifier, and return the obtained new current text simplification result to step S1.

[0010] Optionally, the S1 specifically includes:

[0011] Determine a first readability value corresponding to the simplified result of the current text and a second readability value corresponding to the original text;

[0012] The difference between the first readability value and the second readability value is determined as the readability difference value.

[0013] Optionally, the readability value includes the first readability value and the second readability value, and the calculation formula of the readability value is:

[0014]

[0015] in, For input text The average sentence length of Represents input text The average number of syllables per word in .

[0016] Optionally, the outputting of the current text simplification result in S3 specifically includes:

[0017] The text simplification results generated in each round are annotated, wherein the annotated content includes the text simplification result, the number of simplifications, a label indicating whether the text can be further simplified, and the readability value corresponding to the text simplification result;

[0018] The labeled content is constructed as data samples for training and output.

[0019] Optionally, the simplification of the current text simplification result in S4 is performed by a text simplifier, wherein a pre-trained flan-t5-base model in the Transformer library of Hugging Face is used as the text simplifier.

[0020] Optionally, the pre-training goal of the text simplifier during the pre-training process is to maximize the likelihood estimate, satisfying the following expression:

[0021]

[0022] Among them, the above formula means to maximize the input text in each time step , and get the output probability.

[0023] Optionally, step S2 is performed by a trained discriminator, wherein the discriminator takes the text simplification result as input, and a flag indicating whether simplification can be continued as a training label, and the training goal is to minimize the binary cross entropy loss.

[0024] Optionally, the step of training the discriminator includes:

[0025] Use LSTM function to calculate the representation of text simplification results;

[0026] Using a multi-layer neural network to reduce the dimensionality of the representation;

[0027] use The function maps the result to Within the range.

[0028] In addition, to achieve the above-mentioned purpose, the present application also provides a text simplification system, which includes: a memory, a processor, and a multi-version text simplification program based on ChatGPT stored on the memory and run on the processor, and when the multi-version text simplification program based on ChatGPT is executed by the processor, the steps of the multi-version text simplification method based on ChatGPT as described above are implemented.

[0029] In addition, to achieve the above-mentioned purpose, the present application also provides a computer-readable storage medium, on which a multi-version text simplification program based on ChatGPT is stored. When the multi-version text simplification program based on ChatGPT is executed by a processor, the steps of the multi-version text simplification method based on ChatGPT as described above are implemented.

[0030] This application has at least the following technical effects:

[0031] By setting a discriminator to determine whether the text result after each simplification meets the user's current simplified text needs, the quality of text simplification is guaranteed and reliable multi-version simplified results are output. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] Figure 1 A schematic diagram of the architecture of the hardware operating environment of the text simplification system involved in the embodiment of the present application;

[0033] Figure 2 This is a flowchart of the first embodiment of the multi-version text simplification method based on ChatGPT of the present application;

[0034] Figure 3 This is a flowchart of the second embodiment of the ChatGPT-based multi-version text simplification method of the present application;

[0035] Figure 4 This is a flowchart of the fourth embodiment of the ChatGPT-based multi-version text simplification method of the present application;

[0036] The realization of the purpose, functional features and advantages of this application will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0037] In order to better understand the above technical solution, exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although exemplary embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments described herein. On the contrary, these embodiments are provided to enable a more thorough understanding of the present disclosure and to fully convey the scope of the present disclosure to those skilled in the art.

[0038] As an implementation plan, Figure 1 This is a schematic diagram of the architecture of the hardware operating environment of the text simplification system involved in the embodiment of the present application.

[0039] like Figure 1 As shown, the text simplification system may include: a processor 1001, such as a CPU, a memory 1005, a user interface 1003, a network interface 1004, and a communication bus 1002. Among them, the communication bus 1002 is used to realize the connection and communication between these components. The user interface 1003 may include a display screen (Display), an input unit such as a keyboard (Keyboard), and the user interface 1003 may optionally include a standard wired interface and a wireless interface. The network interface 1004 may optionally include a standard wired interface and a wireless interface (such as a WI-FI interface). The memory 1005 may be a high-speed RAM memory, or a stable memory (non-volatile memory), such as a disk memory. The memory 1005 may optionally be a storage device independent of the aforementioned processor 1001.

[0040] Those skilled in the art will understand that Figure 1 The text simplification system architecture shown in the text simplification system does not constitute a limitation on the text simplification system, and may include more or less components than those shown in the figure, or combine certain components, or arrange the components differently.

[0041] like Figure 1 As shown, the memory 1005 as a storage medium may include an operating system, a network communication module, a user interface module, and a multi-version text simplification program based on ChatGPT. Among them, the operating system is a program that manages and controls the hardware and software resources of the text simplification system, the multi-version text simplification program based on ChatGPT, and the operation of other software or programs.

[0042] exist Figure 1 In the text simplification system shown, the user interface 1003 is mainly used to connect to the terminal and communicate data with the terminal; the network interface 1004 is mainly used for the background server and communicates data with the background server; the processor 1001 can be used to call the ChatGPT-based multi-version text simplification program stored in the memory 1005.

[0043] In this embodiment, the text simplification system includes: a memory 1005, a processor 1001, and a ChatGPT-based multi-version text simplification program stored in the memory and executable on the processor, wherein:

[0044] When the processor 1001 calls the ChatGPT-based multi-version text simplification program stored in the memory 1005, the following operations are performed:

[0045] S1, determining a readability difference value between a current text simplification result and an original text, wherein the current text simplification result is generated by ChatGPT or a text simplifier model;

[0046] S2, comparing the readability difference value with a preset difference threshold;

[0047] S3, if it is less than or equal to the preset difference threshold, output the current text simplification result;

[0048] S4: otherwise, simplify the current text simplification result by the text simplifier, and return the obtained new current text simplification result to step S1.

[0049] When the processor 1001 calls the ChatGPT-based multi-version text simplification program stored in the memory 1005, the following operations are performed:

[0050] Determine a first readability value corresponding to the simplified result of the current text and a second readability value corresponding to the original text;

[0051] The difference between the first readability value and the second readability value is determined as the readability difference value.

[0052] When the processor 1001 calls the ChatGPT-based multi-version text simplification program stored in the memory 1005, the following operations are performed:

[0053] The text simplification results generated in each round are annotated, wherein the annotated content includes the text simplification result, the number of simplifications, a label indicating whether the text can be further simplified, and the readability value corresponding to the text simplification result;

[0054] Based on the hardware architecture of the text simplification system of the above-mentioned natural language processing technology, an embodiment of the multi-version text simplification method based on ChatGPT of the present application is proposed.

[0055] Reference Figure 2 In the first embodiment, the ChatGPT-based multi-version text simplification method includes the following steps:

[0056] S1, determining a readability difference value between a current text simplification result and an original text, wherein the current text simplification result is generated by ChatGPT or a text simplifier model;

[0057] In this embodiment, a text simplification system (hereinafter referred to as the system) first determines a readability difference value between a current text simplification result and an original text, wherein the current text simplification result may be generated by ChatGPT or a text simplifier model developed by the present embodiment.

[0058] At the beginning of the execution, that is, when the text is simplified for the first time, ChatGPT is used to simplify the text and then generate a text simplification result; and when the text is simplified again, the text simplifier model developed by this implementation is used to simplify and generate a text simplification result.

[0059] Optionally, how ChatGPT performs text simplification and generates a text simplification result: For example, first construct a template that implements the text simplification function:

[0060] 'Please simplify the following sentence: '

[0061] 'Please adjust the following sentences to make them easier to read: '

[0062] 'Rewrite this sentence to make them simpler: '

[0063] in, Indicates the sentence to be simplified that is to be filled in this position.

[0064] Next, input the original sentence Fill in any template to form input ,Right now:

[0065]

[0066] Then, initiate a ChatGPT conversation and use ChatGPT to conduct the conversation. The input of the conversation is , that is, ChatGPT is required to simplify the text and get the output ,Right now:

[0067]

[0068] in, Indicates Round dialogue results, when first input = .

[0069] In this embodiment, the readability difference value refers to a quantitative value of the difference between the current text simplification result and the original text before simplification, which is obtained by subtracting the readability value of the current text simplification result from the readability value of the current text simplification result.

[0070] Optionally, the readability value is calculated as follows:

[0071]

[0072] in, For input text The average sentence length of Represents input text The average number of syllables per word in .

[0073] It should be noted that the average sentence length is the number of words in the text divided by the total number of each word, and the average number of syllables is the number of syllables divided by the total number of each word.

[0074] S2, comparing the readability difference value with a preset difference threshold;

[0075] In this embodiment, after determining the readability difference value between the current text simplification result and the original text, the readability difference value is compared with a preset difference threshold.

[0076] Optionally, the preset difference threshold may be set to 2, 3, 4, or 1 / 10 of the readability of the complex sentence.

[0077] Example:

[0078] Complex sentences:

[0079] "Despite the fact that the professor had thoroughly explained the intricate concepts in the lecture, many students still found it challenging to grasp the underlying principles due to the abstract nature of the subjectmatter."

[0080] The ideal simplified sentence:

[0081] "Even though the professor explained the concepts well, many students struggled to understand the principles because the subject was abstract."

[0082] The calculated complex sentence score is 31.12

[0083] The calculated simple sentence score is 34.26

[0084] If the difference threshold is set to 2 or 3, the requirements can be met.

[0085] Optionally, the size of the preset difference threshold can be set based on the user's reading ability. Specifically, the system quantifies the user's reading ability by capturing the number of times the user's historical text simplification instructions are generated. The more times the user generates historical text simplification instructions, the weaker the user's reading ability is, and the smaller the corresponding preset difference threshold will be set, thereby increasing the number of simplifications in the process of simplifying the text by the system.

[0086] S3, if it is less than or equal to the preset difference threshold, output the current text simplification result;

[0087] S4: otherwise, simplify the current text simplification result by the text simplifier, and return the obtained new current text simplification result to step S1.

[0088] In this embodiment, if it is determined that the readability difference value is less than the preset difference threshold, it is determined that the current text simplification result is simple enough, and the current text simplification result is output.

[0089] If it is determined that the readability difference value is greater than the preset difference threshold, it means that the current text simplification result is still relatively complex. The current text simplification result is simplified by the text simplifier, and the new current text simplification result is returned to step S1. The readability difference value between the new current text simplification result after further simplification and the original text at the beginning is re-determined, and the new readability difference value is compared with the threshold again... This cycle is repeated until it is determined that the readability difference value is less than the preset difference threshold, and the current text simplification result is output.

[0090] In the technical solution provided in this embodiment, a discriminator is set to determine whether the text result after each simplification meets the user's current simplified text requirements, thereby ensuring the quality of text simplification and outputting reliable multi-version simplified results.

[0091] Second embodiment

[0092] As an optional embodiment, refer to Figure 3 Based on the first embodiment, in this embodiment, the output of the current text simplification result in S3 specifically includes:

[0093] S31, annotating the text simplification results generated in each round, wherein the annotation content includes the text simplification result, the number of simplifications, a label indicating whether the text can be further simplified, and a readability value corresponding to the text simplification result;

[0094] S32, constructing the annotated content into data samples for training and outputting them.

[0095] As an optional embodiment, in this embodiment, the system annotates the text simplification results generated in each round while outputting the current text simplification results each time. The annotated content includes the text simplification result, the number of simplifications, the label, and the readability value corresponding to the text simplification result, wherein the label represents whether the text can be further simplified. If the text can be further simplified, it is marked as a first-class label, and if the text cannot be further simplified, it is marked as a second-class label. The annotated content is then constructed as a data sample, and the data sample is used for training the discriminator and the text simplifier during the training process.

[0096] Third embodiment

[0097] As an optional embodiment, based on the first embodiment, in this embodiment, the current text simplification result is simplified in S4 by a text simplifier, wherein a pre-trained flan-t5-base model in the Transformer library of Hugging Face is used as the text simplifier.

[0098] It should be noted that the T5 model can convert complex sentences into simple and easy-to-understand sentences, making it easier for readers to understand the text content, which is beneficial for people with reading difficulties, such as children, non-native speakers, and people with low literacy levels. In addition, simplified texts are easier to be accurately translated by translation systems, improving the accuracy and fluency of cross-language communication.

[0099] Furthermore, in this embodiment, the pre-training goal of the text simplifier during the pre-training process is to maximize the likelihood estimate, satisfying the following expression:

[0100]

[0101] Among them, the above formula means to maximize the input text in each time step , and get the output The probability of A unit (token) representing each time step in the original input text, Represents the intermediate results produced by the autoregressive model at different time steps.

[0102] It should be noted that adjusting model parameters with maximizing likelihood estimation as the pre-training goal can make the text generated by the model more consistent with the target text in the training data, and help the model generate simplified text that is fluent, coherent, and grammatically correct. In addition, by maximizing likelihood estimation, the model can better capture the semantic and contextual information in the training data and reduce the risk of overfitting.

[0103] Fourth embodiment

[0104] As an optional embodiment, based on the first embodiment, in this embodiment, step S2 of comparing the readability difference value with a preset difference threshold is performed by a trained discriminator, wherein the discriminator takes the text simplification result as input, and a sign of whether simplification can continue as a training label, and the training goal is to minimize the binary cross entropy loss.

[0105] For example, the expression for minimizing the binary cross entropy loss is:

[0106]

[0107] In the formula, is the total sample size, is the true value of the sample, is the predicted value.

[0108] Further, see Figure 4 , the training steps of the discriminator include:

[0109] S100, uses LSTM function to calculate the representation of text simplification results;

[0110] S200, reducing the dimension of the representation using a multi-layer neural network;

[0111] S300, use The function maps the result to Within the range.

[0112] In this embodiment, the LSTM function is first used to calculate the text simplification result, and the input text simplification result is encoded into a vector representation of a fixed length.

[0113] For example, let the text simplification result be X. If Pytorch or other common neural network tools are used, LSTM can be applied to obtain the output of each time step and the hidden state h, c of the LSTM unit. This embodiment uses Pytorch as a tool, which is characterized as follows:

[0114]

[0115] in, is the output of each time step of LSTM, that is, the representation of the text simplification result, Represents the hidden state in the LSTM.

[0116] Next, a multi-layer neural network (MLP) is used to reduce the dimensionality of the LSTM representation. The multi-layer neural network contains multiple fully connected layers and uses non-linear activation functions to capture complex feature representations.

[0117] Optionally, the reduced dimension may be 32 dimensions.

[0118] Finally, the Sigmoid function is used to map the reduced representation to Within the interval, if the result is greater than 0.5, it indicates that the input text can still be simplified; if the result is less than 0.5, it indicates that the text cannot be simplified further.

[0119] In addition, it can be understood by a person skilled in the art that all or part of the processes in the method for implementing the above embodiment can be completed by instructing the relevant hardware through a computer program. The computer program includes program instructions, and the computer program can be stored in a storage medium, which is a computer-readable storage medium. The program instructions are executed by at least one processor in the text simplification system to implement the process steps of the embodiment of the above method.

[0120] Therefore, the present application also provides a computer-readable storage medium, which stores a multi-version text simplification program based on ChatGPT. When the multi-version text simplification program based on ChatGPT is executed by a processor, it implements the various steps of the multi-version text simplification method based on ChatGPT described in the above embodiment.

[0121] The computer-readable storage medium may be a USB flash drive, a mobile hard disk, a read-only memory (ROM), a magnetic disk, or an optical disk, etc., which are computer-readable storage media that can store program codes.

[0122] It should be noted that since the storage medium provided in the embodiment of the present application is the storage medium used to implement the method of the embodiment of the present application, based on the method introduced in the embodiment of the present application, the person skilled in the art can understand the specific structure and deformation of the storage medium, so it is not repeated here. All storage media used in the method of the embodiment of the present application belong to the scope of protection of this application.

[0123] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems or computer program products. Therefore, the present application may take the form of a complete hardware embodiment, a complete software embodiment or an embodiment combining software and hardware. Moreover, the present application may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program codes.

[0124] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems) and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0125] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0126] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0127] It should be noted that in the claims, any reference signs placed between brackets shall not be construed as limiting the claims. The word "comprising" does not exclude the presence of components or steps not listed in the claim. The word "a" or "an" preceding a component does not exclude the presence of a plurality of such components. The present application may be implemented by means of hardware comprising several different components and by means of a suitably programmed computer. In a unit claim enumerating several means, several of these means may be embodied by the same item of hardware. The use of the words first, second, and third etc. does not indicate any order. These words may be interpreted as names.

[0128] Although the preferred embodiments of the present application have been described, those skilled in the art may make additional changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications falling within the scope of the present application.

[0129] Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the spirit and scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the claims of the present application and their equivalents, the present application is also intended to include these modifications and variations.

Claims

1. A multi-version text simplification method based on ChatGPT, characterized in that: The method comprises the following steps: S1, determining a readability difference value between a current text simplification result and an original text, wherein the current text simplification result is generated by ChatGPT or a text simplifier; S2, comparing the readability difference value with a preset difference threshold; S3, if it is less than or equal to the preset difference threshold, output the current text simplification result; S4, otherwise, simplify the current text simplification result by the text simplifier, and return the obtained new current text simplification result to step S1; Simplifying the current text simplification result in S4 by using a text simplifier, wherein a pre-trained flan-t5-base model in the Transformer library of HuggingFace is used as the text simplifier; At the initial execution, that is, when text simplification is performed for the first time, ChatGPT is used to simplify the text and generate a text simplification result; and when text simplification is performed again, the text simplifier is used to simplify the text and generate a text simplification result.

2. The method according to claim 1, characterized in that The S1 specifically includes: Determine a first readability value corresponding to the simplified result of the current text and a second readability value corresponding to the original text; The difference between the first readability value and the second readability value is determined as the readability difference value.

3. The method according to claim 2, characterized in that The readability value includes the first readability value and the second readability value, and the calculation formula of the readability value is: ; in, For input text The average sentence length of Represents input text The average number of syllables per word in .

4. The method according to claim 1, characterized in that The current text simplification result is output in S3, which specifically includes: The text simplification results generated in each round are annotated, wherein the annotated content includes the text simplification result, the number of simplifications, a label indicating whether the text can be further simplified, and the readability value corresponding to the text simplification result; The labeled content is constructed as data samples for training and output.

5. The method according to claim 1, characterized in that The pre-training goal of the text simplifier during the pre-training process is to maximize the likelihood estimate, satisfying the following expression: ; Among them, the above formula means to maximize the input text in each time step , and get the output probability.

6. The method according to claim 1, characterized in that Step S2 is performed by the trained discriminator, wherein the discriminator takes the text simplification result as input, and a sign of whether to continue simplification is used as a training label, and the training goal is to minimize the binary cross entropy loss.

7. The method according to claim 6, characterized in that The training steps of the discriminator include: Use LSTM function to calculate the representation of text simplification results; Using a multi-layer neural network to reduce the dimensionality of the representation; use The function maps the result to Within the range.

8. A text simplification system, characterized in that: The text simplification system includes: a memory, a processor, and a ChatGPT-based multi-version text simplification program stored in the memory and executable on the processor. When the ChatGPT-based multi-version text simplification program is executed by the processor, the steps of the ChatGPT-based multi-version text simplification method as described in any one of claims 1 to 7 are implemented.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a ChatGPT-based multi-version text simplification program, which, when executed by a processor, implements the steps of the ChatGPT-based multi-version text simplification method as described in any one of claims 1 to 7.

Citation Information

Patent Citations

  • Model training method, text processing method and translation system

    CN116822604A

  • Text readability formula hierarchical calculation method and system and storage medium

    CN117131854A

  • Text simplification method based on Llama2

    CN117808124A

  • Systems and methods for real-time search based generative artificial intelligence

    US20240020538A1

  • Text simplification with minimal hallucination

    US20240119220A1