Data processing method and device

Abstract

The invention provides a data processing method and device, and relates to the technical field of computers. The data processing method comprises the steps of obtaining first input data and a response result for the first input data; performing identification processing on the response result to obtain first identification data; based on the first input data and the first identification data, performing corresponding processing on the response result by using a target processing strategy to output a target file; wherein the target file better conforms to the user intention of the target user generating the first input data compared with the response result.

Description

Technical Field

[0001] The present disclosure relates to the field of computer technology, and in particular to a data processing method and device. Background Art

[0002] At present, when performing generation tasks such as image generation, text generation or video generation, large language models (LLM) often generate images, texts or videos that do not match the user's input information, that is, the generation results are of low accuracy and the user experience is poor. Summary of the invention

[0003] The present disclosure provides a data processing method and device to at least solve the above technical problems existing in the prior art.

[0004] According to a first aspect of the present disclosure, a data processing method is provided, comprising: obtaining first input data and a response result to the first input data; performing identification processing on the response result to obtain first identification data; performing corresponding processing on the response result with a target processing strategy based on the first input data and the first identification data to output a target file; wherein, the target file is more in line with the user intention of the target user who generated the first input data than the response result.

[0005] In one possible implementation, obtaining first input data and a response result for the first input data includes at least one of the following: obtaining a first processing instruction for the first input data, and generating and processing the first input data using a first processing model that matches the first processing instruction to obtain the response result; generating and processing first target data in the first input data based on the type of the first input data and the user intent represented to obtain the response result; integrating and processing the first input data and the context data of its source application to obtain second input data, and generating and processing the second input data using a generative model to obtain the response result.

[0006] In one possible implementation, obtaining a first processing instruction for the first input data includes at least one of the following: generating the first processing instruction in response to obtaining a trigger operation for a first type of control acting on a first type of application, the first type of application being an application that can respond to the first input data; generating the first processing instruction in response to obtaining a trigger operation for first input data input to a second type of application, the second type of application being an application that can or cannot respond to the first input data; generating a corresponding first processing instruction based on application information of the third type of application in response to obtaining first input data input to a third type of application; generating a corresponding first processing instruction based on application information of the type of the first input data and / or source application.

[0007] In one possible implementation manner, the response result is identified and processed to obtain first identification data, including: determining a processing model or processing procedure corresponding to the response result based on the first input data and / or a first processing instruction for the first input data; identifying and processing the response result based on the determined processing model or processing procedure to obtain the first identification data; wherein a type of the first identification data is the same as or different from a type of the first input data.

[0008] In one possible implementation, determining a processing model or a processing program corresponding to the response result based on the first input data and / or a first processing instruction for the first input data includes at least one of the following: determining a second processing model or a first processing program for identifying the response result based on the type of the first input data, so as to process the response result into first identification data of the same type using the second processing model or the first processing program; determining a third processing model or a third processing program for identifying the response result based on the first processing model or the second processing program called by the first processing instruction to process the first input data into the response result, so as to process the response result into the first identification data using the third processing model or the third processing program; determining a fourth processing model or a fourth processing program for identifying the response result based on the instruction content of the first processing instruction, so as to process the response result into the first identification data using the fourth processing model or the fourth processing program; determining a fifth processing model or a fifth processing program for identifying the response result based on the content of the first input data or the user intention represented, so as to process the response result into the first identification data using the fifth processing model or the fifth processing program.

[0009] In one possible implementation, the response result is processed accordingly with a target processing strategy based on the first input data and the first identification data, including: obtaining a comparison result of the first input data and the first identification data, and processing the response result accordingly with a target processing strategy based on the comparison result, wherein different comparison results correspond to different processing strategies.

[0010] In one possible implementation, the response result is processed accordingly with a target processing strategy based on the comparison result, including at least one of the following: when the comparison result indicates that the similarity difference between the first input data and the first recognition data is less than a first threshold, the response result is directly output as the target file, or the response result is generated to obtain the target file; when the comparison result indicates that the similarity difference between the user intention represented by the first input data and the user intention represented by the first recognition data is less than a second threshold, the first input data is generated again and then the response result is replaced as the target file; when the comparison result indicates that the similarity difference between the first input data and the first recognition data is greater than a second threshold, the response result is optimized or compared with the generated result generated again for the first input data to output the target file.

[0011] In one possible implementation manner, optimizing the response result or comparing the response result with a generated result generated again from the first input data to output the target file includes at least one of the following: optimizing the response result or comparing it with the generated result based on the similarity difference and / or the type of the response result to output the optimized response result as the target file, or outputting the response result and / or the generated result after comparison as the target file; replacing or optimizing a target part of the response result to obtain the target file, the target part including a similarity difference part between the response result and the first input data; obtaining a generated result obtained by generating the first input data using a sixth processing model, comparing the generated result with the response result, and outputting the response result and / or the generated result as the target file based on the comparison result, wherein the sixth processing model is set in the electronic device or other device different from the electronic device.

[0012] In one possible implementation, the response result is processed accordingly with a target processing strategy based on the first input data and the first recognition data, and also includes at least one of the following: integrating the first input data and the first recognition data to obtain third input data, and replacing the response result based on the generated result corresponding to the third input data to obtain the target file; performing a similarity evaluation on the first input data and the first recognition data based on the user portrait information of the target user and / or the current environment parameters, and optimizing the response result based on the evaluation result to output the target file.

[0013] According to a second aspect of the present disclosure, a data processing device is provided, comprising: an acquisition module, used to obtain first input data and a response result to the first input data; an identification module, used to perform identification processing on the response result to obtain first identification data; a processing module, used to perform corresponding processing on the response result based on the first input data and the first identification data with a target processing strategy to output a target file; wherein, the target file is more in line with the user intention of the target user who generated the first input data than the response result.

[0014] According to a third aspect of the present disclosure, there is provided an electronic device, including:

[0015] at least one processor; and

[0016] a memory communicatively connected to the at least one processor; wherein,

[0017] The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the method described in the present disclosure.

[0018] According to a fourth aspect of the present disclosure, a non-transitory computer-readable storage medium storing computer instructions is provided, wherein the computer instructions are used to cause the computer to execute the method described in the present disclosure.

[0019] It should be understood that the content described in this section is not intended to identify the key or important features of the embodiments of the present disclosure, nor is it intended to limit the scope of the present disclosure. Other features of the present disclosure will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] The above and other objects, features and advantages of the exemplary embodiments of the present disclosure will become readily understood by reading the detailed description below with reference to the accompanying drawings. In the accompanying drawings, several embodiments of the present disclosure are shown in an exemplary and non-limiting manner, in which:

[0021] In the drawings, the same or corresponding reference numerals represent the same or corresponding parts.

[0022] Figure 1 A schematic diagram showing a data processing method according to an embodiment of the present disclosure Figure 1 ;

[0023] Figure 2 A schematic diagram showing a data processing method according to an embodiment of the present disclosure Figure 2 ;

[0024] Figure 3 A schematic diagram showing a data processing method according to an embodiment of the present disclosure Figure 3 ;

[0025] Figure 4 A schematic diagram showing a data processing method according to an embodiment of the present disclosure Figure 4 ;

[0026] Figure 5 A schematic diagram of a scenario of a data processing method according to an embodiment of the present disclosure is shown;

[0027] Figure 6 A schematic diagram of the structure of a data processing device according to an embodiment of the present disclosure is shown;

[0028] Figure 7 A schematic diagram of the structure of an electronic device according to an embodiment of the present disclosure is shown. DETAILED DESCRIPTION

[0029] In order to make the purpose, features, and advantages of the present disclosure more obvious and easy to understand, the technical solutions in the embodiments of the present disclosure will be clearly and completely described below in conjunction with the drawings in the embodiments of the present disclosure. Obviously, the described embodiments are only part of the embodiments of the present disclosure, not all of the embodiments. Based on the embodiments in the present disclosure, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present disclosure.

[0030] Figure 1 A schematic diagram showing a data processing method according to an embodiment of the present disclosure Figure 1 ,like Figure 1 As shown, a data processing method includes:

[0031] Step S101, obtaining first input data and a response result to the first input data.

[0032] In this embodiment, the first input data is the data input by the user to the target application. The target application may include an image generation application (such as Lenovo's creator zone), a text generation application, a video generation application, or an intelligent agent application (such as Lenovo's AI agent Xiaotian / AI Now). Among them, the intelligent agent application refers to an intelligent assistant based on artificial intelligence technology, such as an intelligent voice assistant, a device intelligent assistant, etc. The target application may also include applications that can call or integrate image generation applications, text generation applications, video generation applications, and / or intelligent agent applications, such as social media applications or album applications, which can process the input data by calling image generation applications, text generation applications, video generation applications, and / or intelligent agent applications, or, the processing of the input data can be realized through a built-in large model (such as a raw image model, a raw text model, a raw video model, a raw audio model, a raw UI model, etc.). For example, the target application is a social media application, and the social media application has a built-in text generation model and an image generation model to support the article generation function and / or image generation function. The user can choose to create an article in the social media application by providing keywords, keywords or scene description information. The text generation model can generate text that meets the user's intention by identifying and extracting the features of keywords, keywords or scene description information. The image generation model can generate images that meet the user's intention by identifying and extracting the features of keywords, keywords or scene description information. The text and image are combined to obtain an article that meets the user's intention. The first input data can be data or instructions such as text data, voice data, and image data used to express user needs; the response result to the first input data is the result generated by the target application or target processing model based on the first input data, and the response result can be an image, audio, text, video, or user interface (UI, User Interface) prototype, etc.

[0033] In one example, if the first input data is the text "Please generate an image of a lotus", the response result is an image of a lotus generated by LLM or a target application; if the first input data is the text "Please generate a word with a similar meaning to 'lotus'", the response result is a word with a similar meaning to "lotus" generated by LLM or a target application, such as "lotus", "lotus bud", etc.; if the first input data includes an image P and the text "Please change object a in this image to object b", the response result is an image P1 generated by LLM after changing object a in image P to object b.

[0034] Step S102: perform identification processing on the response result to obtain first identification data.

[0035] In this embodiment, in order to determine the content of the response result generated by the LLM, other processing models or target applications, it is necessary to perform recognition processing on the response result to obtain first recognition data representing the content or features of the response result. The response result can be correspondingly recognized based on the type of the first input data and / or the user intention represented by the first input data, and the recognition processing includes image-to-text conversion processing, image recognition processing, feature extraction processing, video-to-text processing, text summary processing or summary processing. For example, if the type of the first input data is text, and the user intention represented by it is to generate an image based on the text, then the image in the response result needs to be processed for image-to-text conversion to obtain first recognition data representing the content drawn by the image in the response result; if the type of the first input data is text, and the user intention represented by it is to generate other related texts, then the text in the response result needs to be processed for semantic recognition, intention understanding, or a summary or summary of the corresponding word expansion processing to obtain first recognition data representing the meaning expressed by the text in the response result; if the type of the first input data is text, and the user intention represented by it is to generate a video based on the text, then the video in the response result needs to be processed for video-to-text to obtain first recognition data representing the content played by the video in the response result. It should be emphasized that the type of the first recognition data and the type of the first input data may be the same or different. For example, if the type of the first input data is text and an image is generated based on the text, then after the image in the response result is processed by recognition (such as image recognition and / or feature extraction), the first recognition data obtained may be text representing the content of the image in the response result, or audio representing the content of the image in the response result.

[0036] In one example, if the type of the first input data is the text "Please generate an image of a lotus", and the user intention represented by it is to generate an image of a lotus, then it is proved that the type of the response result is an image, and the response result needs to be converted from image to text to obtain the first recognition data representing the content of the response result image, and the first recognition data may be the text "This image includes a blooming lotus"; if the type of the first input data is the text "Please generate words with similar meanings to 'lotus'", and the user intention represented by it is to generate synonyms of "lotus", then it is proved that the type of the response result is text, and the response result needs to be semantically recognized or intentionally understood to obtain the first recognition data representing the content of the response result text, and the first recognition data may be the text "These words have similar meanings and refer to the same plant, namely lotus"; if the type of the first input data is the text "Please generate a video related to 'lotus'", and the user intention represented by it is to generate a video of lotus, then it is proved that the type of the response result is a video, and the response result needs to be converted from video to text to obtain the first recognition data representing the content of the response result video, and the first recognition data may be the text "This video shows a scene of a lotus swaying in the breeze on the water".

[0037] Step S103: Based on the first input data and the first recognition data, the response result is processed accordingly with a target processing strategy to output a target file.

[0038] In this embodiment, the first input data can represent the user intention of the target user who inputs the first input data, and the first recognition data represents the content of the response result generated based on the first input data. By comparing the first input data and the first recognition data, it can be determined whether the response result meets the user intention. If it does not meet the user intention, it is necessary to perform corresponding processing on the response result with a target processing strategy, such as generation processing, image beautification processing, replacement processing, editing and modification processing, etc., so as to obtain a target file. The target file can be an image, text, video, UI prototype, etc. Compared with the response result, the target file is more in line with the user intention of the target user who generated the first input data.

[0039] In one example, different target processing strategies can be determined based on the type of response result and the comparison result between the first input data and the first recognition data. For example, if the comparison result shows that the first input data and the first recognition data are significantly different, that is, the response result does not meet the user's intention, and the response result is an image, then the image processing strategy can be used to beautify the response result, including modifying the part of the response result that does not meet the user's intention, or using other raw image models such as cloud models to generate another response result based on the first input data, compare the response results generated by LLM with those generated by other raw image models, and determine the response result that is more in line with the user's intention represented by the first input data as the target file; if the comparison result shows that the first input data and the first recognition data are significantly different, that is, the response result does not meet the user's intention, and the response result is a video, then The response result can be beautified by using image processing, video editing or beautification models, so that the beautified response result meets the user's intention, or the raw image model can be used to replace the target object or target video frame in the response result that does not meet the user's intention, or the first input data can be modified based on the response result, and the video generation process is re-performed based on the modified first input data; if the comparison result shows that the first input data and the first recognition data are significantly different, that is, the response result does not meet the user's intention, and the response result is text, the response result can be modified and beautified by using a document modification or beautification model, or directly regenerated, so that the beautified or regenerated response result meets the user's intention.

[0040] In the present disclosure, the response result generated based on the first input data is identified and processed to obtain first identification data representing the content of the response result. The comparison result between the first input data and the first identification data can determine whether the response result meets the user's intention. If it does not meet the user's intention, the response result is processed accordingly with the target processing strategy, such as generation processing, image beautification processing, replacement processing, editing and modification processing, etc., so that the target file obtained after processing meets the user's intention. As a result, the accuracy of the generation results of the LLM, other processing models or target applications for the user's first input data can be improved, thereby improving the user experience.

[0041] In another embodiment, step S101 of “obtaining first input data and a response result to the first input data” includes at least one of the following:

[0042] A first processing instruction for the first input data is obtained, and the first processing model matching the first processing instruction is used to generate and process the first input data to obtain a response result. After the first processing instruction is executed, the user intention represented by the first input data can be completed. The first processing instruction can be a picture generation instruction, a text generation instruction, a video generation instruction, a UI prototype generation instruction, and an audio generation instruction, etc. Different first processing instructions correspond to different first processing models. The first processing model can be an image generation model, a text generation model, or a video generation model, etc. The first processing instruction can be used to call the corresponding first processing model to generate and process based on the first input data, so as to obtain a response result. For example, if the first processing instruction is generated by the trigger operation of the picture generation control of the picture generation application, it is necessary to call the picture generation model to generate the picture based on the first input data; if the first processing instruction is generated on the text generation application, it is necessary to call the text generation model to perform text generation processing based on the first input data; if the first processing instruction is generated on the video generation or music generation application, it is necessary to call the video generation model or the music generation model to perform video generation or music generation processing based on the first input data.

[0043] The first target data in the first input data is generated and processed based on the type of the first input data and the user intention represented to obtain a response result. Wherein, different types of first input data have different processing methods for generating response results. For example, when the first input data is text data, the processing method may be to perform semantic analysis or intention understanding on the text data to determine the user intention, and generate text or image or video based on the user intention; when the first input data is image data, the processing method may be to convert the image data into text, determine the user intention based on the conversion result, and generate an image or video based on the user intention. Wherein, the first target data may be all or part of the first input data. For example, only specific types of words in the first input data may be generated, such as nouns, adjectives, etc.; if the first input data is an image, specific objects in the image are generated, such as only puppies or grass in the image are generated. Therefore, only the first target data in the first input data is generated, which can reduce the processor consumption in the generation process and increase the generation rate.

[0044] The first input data and the context data of its source application are integrated and processed to obtain the second input data, and the second input data is generated and processed using a generative model to obtain a response result. Among them, if the generation of the response result depends not only on the first input data, but also on the context data of the first input data, the first input data and its context data can be integrated and processed, including operations such as deduplication, weighting and splicing, to obtain the second input data, and a response result is generated based on the second input data. For example, the first input data can be text data captured from an article or news, and the context data of the first input data is text data that is scene-related to the first input data. Therefore, the scene features of the first input data can be supplemented in combination with the context data of the first input data to generate a response result. For example, in a short article, there is a description text "On a sunny afternoon, a puppy lies on the ground and plays". If the text data "a puppy" in the short article is intercepted as the first input data, the scene feature "bright afternoon" and the action feature "lying on the ground" of "a puppy" can be obtained by combining the context data of "a puppy". The scene feature "bright afternoon" and the action feature "lying on the ground" are combined with the first input data "a puppy" to generate the second input data "a puppy lies on the ground". Therefore, the response result combines the first input data and its context data, has richer meaning, and may be more in line with the user's intention.

[0045] In another embodiment, “obtaining a first processing instruction for first input data” includes at least one of the following:

[0046] In response to obtaining a trigger operation of a first type of control acting on a first type of application, a first processing instruction is generated, and the first type of application is an application that can respond to the first input data. Among them, the first type of application can be various intelligent body applications, image generation applications, text generation applications or video generation applications, among which the intelligent body application can include Lenovo Xiaotian or AI NOW, the image generation application can include Lenovo's creator zone application, the text generation application can include Lenovo's AI learning application, and the video generation application can include sora (an artificial intelligence text generation video model released by OpenAI, an American artificial intelligence research company) application. The first category of applications can respond to first input data to implement corresponding functional services, such as raw images, raw text, raw videos, AI image processing, beautification and device control, etc. The first category of controls can be raw image controls, raw text controls, raw video controls, raw music controls, generated UI controls, etc. in the first category of applications; clicking, dragging and selecting operations in the first category of controls, such as clicking on an image in the first category of controls, dragging a document into the first category of controls, or selecting images and text in the first category of controls, can generate corresponding first processing instructions, and the first processing instructions include text generation instructions, polishing processing instructions, raw image instructions, raw video instructions, generated UI prototype instructions, etc. For example, in response to clicking on an image in the first category of controls, a first processing instruction for generating a similar image based on the image is generated.

[0047] In response to obtaining a trigger operation for first input data input to a second type of application, a first processing instruction is generated, and the second type of application is an application that can respond to or cannot respond to the first input data. Among them, the second type of application can be the same as or different from the first type of application. The second type of application can be a social application or a conference application. The first input data is input into the second type of application during a chat or conference. If a trigger operation such as copying and pasting, screenshot operation, circling operation, marking operation, editing operation, etc. is performed on the first input data in the second type of application, then in response to the trigger operation, a first processing instruction is generated, and a response is performed based on the first processing instruction. If the second type of application cannot respond to the first processing instruction corresponding to the first input data, the LLM can be called to respond to the first processing instruction, and the second type of application can receive the response result generated by the LLM.

[0048] In response to obtaining the first input data input into the third category application, a corresponding first processing instruction is generated based on the application information of the third category application. The third category application may be a raw image application, a raw text application, a raw video application, etc. The corresponding first processing instruction may be generated based on the application information of the third category application input by the first input data, such as the type of the third category application, etc. For example, if the first input data is input into the raw image application, the first processing instruction is generated according to the raw image rule, and the first processing instruction represents raw image processing; if the first input data is input into the raw text application, the first processing instruction is generated according to the raw text rule, and the first processing instruction represents raw text processing.

[0049] Generate a corresponding first processing instruction based on the type of the first input data and / or the application information of the source application. The type of the first input data may be text, picture, voice, video, etc. If the first input data is text, the first processing instruction may be an instruction to instruct raw text processing; if the first input data is an image, the first processing instruction may be an instruction to instruct raw image processing or image-text conversion; the source application of the first input data is the application into which the user inputs the first input data, and the application information of the source application includes information such as the application type or function type. Based on the application information, a corresponding first processing instruction may be generated. If the application information indicates that the source application is a raw image application, the first processing instruction is an instruction to instruct raw image processing. For example, if the first input data is text data, a text enhancement instruction may be generated as the first processing instruction; if the first input data is audio data, a volume amplification instruction may be generated as the first processing instruction.

[0050] Figure 2 A schematic diagram showing a data processing method according to an embodiment of the present disclosure Figure 2 ,like Figure 2 As shown, a data processing method includes:

[0051] Step S201, obtaining first input data and a response result to the first input data.

[0052] The specific implementation details of step S201 are similar to those of step S101 and will not be repeated here.

[0053] Step S202: determining a processing model or a processing procedure corresponding to a response result based on the first input data and / or a first processing instruction for the first input data.

[0054] In this embodiment, the user intention can be determined based on the type or content of the first input data, and the processing model or processing program for processing the response result can be determined based on the user intention. For example, if the content of the first input data is the text "Please generate an image of a lotus", and the user's intention is to generate an image of a lotus, it can be determined that the image in the response result needs to be processed from image to text to identify the content of the image in the response result. Therefore, it can be determined that the processing model is a picture-to-text model; the first processing instruction includes a picture-generating instruction, a text-generating instruction, a video-generating instruction, a UI prototype-generating instruction, an audio-generating instruction, etc. Based on the first processing instruction, the processing model or processing program for processing the first input data can also be determined. For example, if the first processing instruction is a picture-generating instruction, it can also be determined that the image in the response result generated based on the first processing instruction needs to be processed from image to text to identify the content of the image in the response result. Therefore, it can be determined that the processing model is a picture-to-text model. In addition, the processing model or processing program can be determined in combination with the first input data and the first processing instruction. For example, if the type of the first input data is text and the first processing instruction is a video generation instruction, it can be determined that the video in the response result needs to be processed to generate video text in order to identify the content of the video in the response result. Therefore, the processing model or processing program can be determined to be a video text generation model or a non-model form of processing program that can implement video text generation, etc.

[0055] Step S203: performing identification processing on the response result based on the determined processing model or processing procedure to obtain first identification data.

[0056] In this embodiment, the response result can be identified and processed based on the determined processing model or processing program to obtain the first identification data representing the content of the response result. For example, if the response result generated based on the first input data is an image of a lotus, and the determined processing model or processing program is a picture-to-text model or a non-model form processing program that can realize picture-to-text, then the image in the response result can be directly identified and processed based on the picture-to-text model or the processing program that can realize picture-to-text to obtain the first identification data, and the first identification data can be the text "this image includes a blooming lotus"; if the response result generated based on the first input data is a video of a lotus, and the determined processing model or processing program is a video-to-text model or a processing program that can realize video-to-text, then the video in the response result can be directly identified and processed based on the video-to-text model or the processing program that can realize video-to-text to obtain the first identification data, and the first identification data can be the text "this video shows a scene of a lotus swaying in the breeze on the water". It should be emphasized that the type of the first recognition data and the type of the first input data may be the same or different. For example, if the type of the first input data is text and an image is generated based on the text, then after the image in the response result is recognized, the first recognition data obtained may be text representing the content drawn by the image in the response result, or may be audio representing the content drawn by the image in the response result.

[0057] Step S204: Based on the first input data and the first recognition data, the response result is processed accordingly with a target processing strategy to output a target file.

[0058] The specific implementation details of step S204 are similar to those of step S103 and will not be repeated here.

[0059] In the present disclosure, a processing model or processing procedure corresponding to a response result is determined based on first input data and / or a first processing instruction for the first input data, and the response result is identified and processed based on the determined processing model or processing procedure to obtain first identification data. The first identification data can characterize the content of the response result. By comparing the first identification data with the first input data, it is possible to determine whether the response result meets the user's intention, as well as the part of the response result that does not meet the user's intention, so as to facilitate subsequent targeted processing of the part of the response result that does not meet the user's intention, so as to output a target file that meets the user's intention, thereby improving the accuracy of the generated results for the user's first input data and the user experience.

[0060] In another embodiment, step S202 of “determining a processing model or a processing procedure corresponding to a response result based on the first input data and / or the first processing instruction for the first input data” includes at least one of the following:

[0061] Based on the type of the first input data, a second processing model or a first processing program for identifying and processing the response result is determined, so that the response result is processed into the first identification data of the same type by using the second processing model or the first processing program. Wherein, the type of the first input data includes text, image, video, audio, etc. In order to facilitate the comparison of the first identification data of the response result with the first input data, the response result can be processed into the first identification data of the same type as the first input data. Therefore, based on the type of the first input data, a second processing model or a first processing program for identifying and processing the response result can be determined, and the second processing model or the first processing program can identify the response result as the type of the first input data, and use the second processing model or the first processing program to process the response result into the first identification data of the same type. For example, if the type of the first input data is text and the type of the response result is image, the second processing model or the first processing program should be a picture-to-text model or a picture-to-text processing program that converts the image into text, and the picture-to-text model or the picture-to-text processing program can convert or reversely process the image to obtain the first identification data of the text type.

[0062] Based on the first processing model or second processing program called by the first processing instruction to process the first input data into a response result, a third processing model or third processing program for identifying the response result is determined, so as to use the third processing model or third processing program to process the response result into the first identification data. Wherein, the processing process of the first processing model or second processing program called by the first processing instruction to process the first input data into a response result and the processing process of the third processing model or third processing program for identifying the response result are reversed. For example, if the first input data is text and the response result is an image, the processing process of the first processing model or second processing program called by the first processing instruction to process the first input data into a response result is to generate an image from text, and the processing process of the third processing model or third processing program for identifying the response result is from image to text, so it can be determined that the third processing model or third processing program is a picture-to-text model or a picture-to-text processing program.

[0063] Based on the instruction content of the first processing instruction, a fourth processing model or a fourth processing program for identifying the response result is determined, so as to use the fourth processing model or the fourth processing program to process the response result into the first identification data. The instruction content represents the processing model or processing logic for obtaining the response result based on the first input data, and the processing model or processing logic is the opposite of the processing logic for identifying the response result. Therefore, the processing logic for identifying the response result can be determined by using the instruction content, and the fourth processing model or the fourth processing program can be determined based on the processing logic for identifying the response result. For example, if the instruction content is to generate an image based on text, it can be determined that the processing logic for identifying the response result is to generate text based on an image, so the fourth processing model or the fourth processing program can be a picture-to-text model or a picture-to-text processing program.

[0064] Based on the content of the first input data or the user intention represented, a fifth processing model or a fifth processing program for identifying and processing the response result is determined, so as to process the response result into the first identification data using the fifth processing model or the fifth processing program. The content of the first input data and the user intention can represent the processing logic for generating the response result based on the first input data, such as generating a video from text, and the processing logic can be reversed to obtain the processing logic for identifying and processing the response result, such as generating text from video, and then the fifth processing model or the fifth processing program can be determined to be a video-to-text model or a video-to-text processing program.

[0065] Figure 3 A schematic diagram showing a data processing method according to an embodiment of the present disclosure Figure 3 ,like Figure 3 As shown, a data processing method includes:

[0066] Step S301, obtaining first input data and a response result to the first input data.

[0067] Step S302: perform identification processing on the response result to obtain first identification data.

[0068] The specific implementation details of step S301-step S302 are similar to those of step S101-step S102 and will not be repeated here.

[0069] Step S303, obtaining a comparison result between the first input data and the first recognition data, and performing corresponding processing on the response result with a target processing strategy based on the comparison result.

[0070] In this embodiment, the comparison result between the first identification data and the first input data can reflect the difference in similarity between the response result and the user's intention. Based on the comparison result, it can be determined whether the response result meets the user's intention, and different comparison results correspond to different processing strategies. Therefore, the comparison result between the first input data and the first identification data can be obtained, and the response result can be processed accordingly based on the target processing strategy corresponding to the comparison result. For example, if the comparison result shows that the response result does not meet the user's intention, the response result is optimized so that the optimized response result meets the user's intention; if the comparison result shows that the response result meets the user's intention, the response result is determined as a target file and output. In this way, a target file that meets the user's intention can be obtained, which improves the accuracy of the generated result for the user's first input data and the user experience.

[0071] In one example, different processing strategies can be determined based on the type of response result and the comparison result. For example, in a raw image scenario, the response result can be beautified using an image processing strategy, or another response result can be regenerated based on the first input data using a cloud model or other raw image model, and the two response results can be compared to determine the response result that better meets the user's intention as the target file and output it, or the first input data can be modified based on the response result, and the raw image processing can be re-performed based on the modified first input data; in a raw video scenario, the response result can be beautified using image processing, video editing or beautification models, so that the beautified response result meets the user's intention, or the raw image model can be used to replace the target object or target video frame in the response result that does not meet the user's intention, or the first input data can be modified based on the response result, and the video generation processing can be re-performed based on the modified first input data; in a raw text scenario, the response result can be modified and beautified using a document modification or beautification model, or directly regenerated, so that the beautified or regenerated response result meets the user's intention.

[0072] In another embodiment, "processing the response result with the target processing strategy based on the comparison result" in step S303 includes at least one of the following:

[0073] When the comparison result indicates that the similarity difference between the first input data and the first recognition data is less than the first threshold, the response result is directly output as the target file, or the response result is generated to obtain the target file. The similarity difference less than the first threshold proves that the first input data and the first recognition data are substantially the same, for example, the number and content of keywords in the two are the same, or the number and content of image features are the same, etc. At this time, the response result can be directly output as the target file, or the response result can be generated again to obtain the target file, and the processing model used for the second generation process can be the same as or different from the model of the first generation process.

[0074] When the comparison result indicates that the similarity difference between the user intention represented by the first input data and the user intention represented by the first recognition data is less than the second threshold, the first input data is generated again and the response result is replaced as the target file. The second threshold may be greater than the first threshold, and the similarity difference is less than the second threshold, which proves that there is a difference between the first input data and the first recognition data. At this time, the first input data can be generated again based on the local generation model or the cloud model to obtain a new response result, and the new response result replaces the original response result as the target file. The local generation model or the cloud model can be the same or different model as the first generation process. The cloud model is generally larger, has more parameters, may be trained with more diverse data sets, and can be updated and optimized in real time. Therefore, the response result generated by the cloud model is generally better than the original response result, further improving the accuracy of the generation result for the user's first input data.

[0075] When the comparison result indicates that the similarity difference between the first input data and the first recognition data is greater than the second threshold, the response result is optimized or compared with the result of regenerating the first input data to output the target file. The similarity difference greater than the second threshold proves that there is a large difference between the first input data and the first recognition data. At this time, the response result can be optimized, such as determining the part of the response result that is different from the user's intention based on the first recognition data and the first input data, and optimizing or replacing the different part so that the response result meets the user's intention; or, using the cloud model to generate another response result based on the first input data and then compare it with the locally generated response result, and output the best of the two or both.

[0076] In another embodiment, “optimizing the response result or comparing the response result with a generated result of regenerating the first input data to output a target file” includes at least one of the following:

[0077] Based on the similarity difference and / or the type of the response result, the response result is optimized or compared with the generated result, so as to output the optimized response result as the target file, or the response result and / or the generated result after comparison is output as the target file. If the similarity difference is large, another response result can be regenerated based on the cloud model or the modified first input data, and the two response results can be output preferentially; if the similarity difference is small, the response result can be processed and optimized based on the model; if the type of the response result is an image or video, it can be optimized; if the type of the response result is text, the response result can be regenerated. For example, if the first input data is "Please generate a picture of a puppy playing with a blue ball on the grass", and the response result is a picture of a kitten on the grass, then the first recognition data of the response result and the first input data have a large difference in similarity. The cloud model can be used to regenerate another response result based on the first input data or the modified first input data (such as "Please generate a picture of a puppy playing with a blue ball on the grass, the puppy is a Chihuahua"). If the other response result is a picture of a Chihuahua playing with a blue ball on the grass, it is obvious that the response result generated by the cloud model is more in line with the user's intention than the original response result, and the response result generated by the cloud model is output; if the first input data is "Please generate a picture of a puppy playing with a blue ball on the grass", and the response result is a picture of a puppy playing with a red ball on the grass, then the first recognition data of the response result and the first input data have a small difference in similarity. The "red ball" in the response result can be adjusted to "blue ball" based on a model such as an image generation model, thereby obtaining a target result that meets the user's intention.

[0078] The target part of the response result is replaced or optimized to obtain the target file, and the target part includes the similarity difference part between the response result and the first input data. Among them, the target part is the part where the response result differs from the user's intention, and only the difference part can be partially optimized. For example, if the first input data is "please generate an image of a lotus with dewdrops on it", and the response result is an image of a lotus, but there are pearls on the lotus, then the pearl area in the response result can be optimized, and the pearls are replaced with dewdrops, so as to obtain the target file that meets the user's intention. It should be emphasized that the replacement or optimization of the target part of the response result can be achieved based on the generation of a small model, such as a small model generated by an image, etc. Compared with the LLM for generating the response result, the generated small model has obvious advantages in terms of data magnitude, resource occupation and power consumption, which can reduce resource occupation and power consumption and improve generation efficiency.

[0079] Obtain a generated result obtained by generating and processing the first input data using the sixth processing model, compare the generated result with the response result, and output the response result and / or the generated result as a target file based on the comparison result. The sixth processing model is set in an electronic device or other device different from the electronic device. Among them, the sixth processing model is a cloud model or an edge model. After obtaining the generated result based on the sixth processing model, the generated result is identified to obtain the second identification data, obtain the similarity difference between the first input data and the second identification data, compare the similarity difference between the first input data and the first identification data and the second identification data, output the generated result or response result with the smaller similarity difference, or output both the generated result and the response result, and determine the target file in response to the user's selection operation for both, thereby improving the user experience.

[0080] Figure 4 A schematic diagram showing a data processing method according to an embodiment of the present disclosure Figure 4 ,like Figure 4 As shown, a data processing method includes:

[0081] Step S401, obtaining first input data and a response result to the first input data.

[0082] Step S402: perform identification processing on the response result to obtain first identification data.

[0083] The specific implementation details of step S401-step S402 are similar to those of step S101-step S102 and are not repeated here.

[0084] Step S403, integrating the first input data and the first recognition data to obtain third input data, replacing the response result based on the generation result corresponding to the third input data to obtain the target file; and / or, evaluating the similarity of the first input data and the first recognition data based on the user portrait information of the target user and / or the current environmental parameters, optimizing the response result based on the evaluation result to output the target file.

[0085] In this embodiment, the first input data can be expanded or beautified by using the first recognition data, such as adding modifiers to keywords, or deleting the content in the first recognition data that is different from the keywords in the first input data, and splicing the deleted first recognition data with the first input data, etc., and the generated result can be obtained again by using the third input data obtained after integration, and the response result can be replaced with the generated result to obtain the target file. For example, if the first input data is "please generate an image of a lotus bud about to bloom", and the first recognition data of the response result is "this image includes a blooming lotus", then the first input data can be integrated based on the first recognition data, emphasizing that the blooming lotus should not be generated, and the integrated third input data can be "please generate an image of a lotus bud about to bloom, the lotus is not blooming", so that the accuracy of the generated result for the user's first input data can be further improved.

[0086] In this embodiment, the first input data and the response result can also be evaluated with reference to user portrait information such as user habits or preferences, and / or the current usage environment or scene parameters. For example, if the historical response results that the user is satisfied with are all images with a relatively soft style, but the image in the response result generated based on the first input data has a strong color, then it can be considered that the response result is significantly different from the first input data, and the response result needs to be optimized or corrected to obtain a target file that matches the user's intention; if the user adjusts the current usage environment or scene parameters to a usage environment or scene parameters for generating a sketch image, but the image in the response result generated based on the first input data is an oil painting, then it can be considered that the response result is significantly different from the first input data, and the response result needs to be optimized or corrected, thereby further improving the user experience.

[0087] In order to facilitate the understanding of the present disclosure, the following Figure 5 A data processing method of the present disclosure is explained as follows:

[0088] Figure 5 A schematic diagram of a data processing method according to an embodiment of the present disclosure is shown. Figure 5As shown, the first input data is "please draw a lotus with dewdrops on it", the raw image model generates a response result image A based on the first input data, and there is a lotus in image A with pearls on it; the response result image A is input into the image-generated text model for recognition processing, and a first recognition result "there is an exquisite lotus in the image with pearls on it" is obtained; the first input data and the first recognition result are input into the comparison model to obtain a comparison result, and if the comparison result shows that the similarity difference between the two is less than the first threshold, the response result image A can be directly output as the target file, and if the comparison result shows that the similarity difference between the two is less than the second threshold, the first input data can be generated again and then replaced with the response result image A as the target file; if the comparison result shows that the similarity difference is greater than the second threshold, the first input data and the first recognition result can be input into the natural language processing (NLP, Natural Language Processing) The first input data is input into a cloud model to perform differential text recognition, and the “dewdrops” in the first input data and the “pearls” in the first recognition result are obtained. The response result image A and the differential text recognition result are input into the raw image model, so that the raw image model replaces the pearls in the response result image A with dewdrops, and obtains the target file image B. There is a lotus in the image B, and there are pearls on the lotus, which meets the user's intention. If the comparison result shows that the similarity difference is greater than the second threshold value, the first input data can also be input into the cloud model, and the cloud model obtains the generated result image C. There is a lotus in the image C, and there are pearls on the lotus. Compared with image A, image C is more in line with the user's intention, and image C can be output as the target file.

[0089] Figure 6 A schematic diagram of the structure of a data processing device according to an embodiment of the present disclosure is shown. Figure 6 As shown, a data processing device includes:

[0090] The acquisition module 10 is used to obtain the first input data and the response result for the first input data; the identification module 11 is used to identify and process the response result to obtain the first identification data; the processing module 12 is used to perform corresponding processing on the response result based on the first input data and the first identification data with a target processing strategy to output a target file; wherein the target file is more in line with the user intention of the target user who generated the first input data than the response result.

[0091] In one possible implementation, the acquisition module 10 is also used to: obtain a first processing instruction for first input data, and generate and process the first input data using a first processing model that matches the first processing instruction to obtain a response result; generate and process the first target data in the first input data based on the type of the first input data and the user intent represented to obtain a response result; integrate and process the first input data and the context data of its source application to obtain second input data, and generate and process the second input data using a generative model to obtain a response result.

[0092] In one embodiment, the acquisition module 10 is also used to: generate a first processing instruction in response to obtaining a trigger operation acting on a first type of control of a first type of application, where the first type of application is an application that can respond to first input data; generate a first processing instruction in response to obtaining a trigger operation for first input data input to a second type of application, where the second type of application is an application that can or cannot respond to the first input data; generate a corresponding first processing instruction based on application information of the third type of application in response to obtaining first input data input to a third type of application; generate a corresponding first processing instruction based on the type of the first input data and / or application information of the source application.

[0093] In one possible implementation mode, the identification module 11 is further used to: determine a processing model or processing procedure corresponding to a response result based on the first input data and / or a first processing instruction for the first input data; perform identification processing on the response result based on the determined processing model or processing procedure to obtain first identification data; wherein the type of the first identification data is the same as or different from the type of the first input data.

[0094] In one possible implementation, the identification module 11 is also used to: determine a second processing model or a first processing program for identifying the response result based on the type of the first input data, so as to process the response result into first identification data of the same type using the second processing model or the first processing program; determine a third processing model or a third processing program for identifying the response result based on the first processing model or the second processing program called by the first processing instruction to process the first input data into the response result, so as to process the response result into the first identification data using the third processing model or the third processing program; determine a fourth processing model or a fourth processing program for identifying the response result based on the instruction content of the first processing instruction, so as to process the response result into the first identification data using the fourth processing model or the fourth processing program; determine a fifth processing model or a fifth processing program for identifying the response result based on the content of the first input data or the user intention represented, so as to process the response result into the first identification data using the fifth processing model or the fifth processing program.

[0095] In one embodiment, the processing module 12 is further used to: obtain a comparison result between the first input data and the first recognition data, and perform corresponding processing on the response result with a target processing strategy based on the comparison result, wherein different comparison results correspond to different processing strategies.

[0096] In one possible implementation, the processing module 12 is also used for: when the comparison result indicates that the similarity difference between the first input data and the first recognition data is less than a first threshold, directly outputting the response result as a target file, or generating the response result to obtain the target file; when the comparison result indicates that the similarity difference between the user intention represented by the first input data and the user intention represented by the first recognition data is less than a second threshold, generating the first input data again and replacing the response result as the target file; when the comparison result indicates that the similarity difference between the first input data and the first recognition data is greater than a second threshold, optimizing the response result or comparing the response result with the generated result generated again for the first input data to output the target file.

[0097] In one possible implementation, the processing module 12 is further used to: optimize the response result or compare it with the generated result based on the similarity difference and / or the type of the response result to output the optimized response result as a target file, or output the response result and / or the generated result after comparison as a target file; replace or optimize the target part of the response result to obtain a target file, the target part includes the similarity difference part between the response result and the first input data; obtain the generated result obtained by generating the first input data using the sixth processing model, compare the generated result with the response result, and output the response result and / or the generated result as a target file based on the comparison result, and the sixth processing model is set in an electronic device or other device different from the electronic device.

[0098] In one possible implementation, the processing module 12 is also used to: integrate the first input data and the first recognition data to obtain third input data, replace the response result based on the generation result corresponding to the third input data, and obtain the target file; evaluate the similarity of the first input data and the first recognition data based on the user portrait information of the target user and / or the current environmental parameters, and optimize the response result based on the evaluation result to output the target file.

[0099] According to an embodiment of the present disclosure, the present disclosure also provides an electronic device and a readable storage medium.

[0100] Figure 7A schematic block diagram of an example electronic device 800 that can be used to implement an embodiment of the present disclosure is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present disclosure described and / or required herein.

[0101] like Figure 7 As shown, the device 800 includes a computing unit 801, which can perform various appropriate actions and processes according to a computer program stored in a read-only memory (ROM) 802 or a computer program loaded from a storage unit 808 into a random access memory (RAM) 803. In the RAM 803, various programs and data required for the operation of the device 800 can also be stored. The computing unit 801, the ROM 802, and the RAM 803 are connected to each other via a bus 804. An input / output (I / O) interface 805 is also connected to the bus 804.

[0102] A number of components in the device 800 are connected to the I / O interface 805, including: an input unit 806, such as a keyboard, a mouse, etc.; an output unit 807, such as various types of displays, speakers, etc.; a storage unit 808, such as a disk, an optical disk, etc.; and a communication unit 809, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 809 allows the device 800 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0103] The computing unit 801 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the computing unit 801 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, digital signal processors (DSPs), and any appropriate processors, controllers, microcontrollers, etc. The computing unit 801 performs the various methods and processes described above, such as a data processing method. For example, in some embodiments, a data processing method may be implemented as a computer software program, which is tangibly contained in a machine-readable medium, such as a storage unit 808. In some embodiments, part or all of the computer program may be loaded and / or installed on the device 800 via ROM 802 and / or communication unit 809. When the computer program is loaded into RAM 803 and executed by the computing unit 801, one or more steps of a data processing method described above may be performed. Alternatively, in other embodiments, the computing unit 801 may be configured to perform a data processing method in any other appropriate manner (e.g., by means of firmware).

[0104] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), load programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0105] The program code for implementing the method of the present disclosure may be written in any combination of one or more programming languages. These program codes may be provided to a processor or controller of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so that the program code, when executed by the processor or controller, enables the functions / operations specified in the flow chart and / or block diagram to be implemented. The program code may be executed entirely on the machine, partially on the machine, partially on the machine and partially on a remote machine as a stand-alone software package, or entirely on a remote machine or server.

[0106] In the context of the present disclosure, a machine-readable medium may be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, device, or equipment. A machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or equipment, or any suitable combination of the foregoing. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0107] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user can provide input to the computer. Other types of devices can also be used to provide interaction with the user; for example, the feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including acoustic input, voice input, or tactile input).

[0108] The systems and techniques described herein may be implemented in a computing system that includes back-end components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes front-end components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such back-end components, middleware components, or front-end components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), and the Internet.

[0109] A computer system may include a client and a server. The client and the server are generally remote from each other and usually interact through a communication network. The relationship of client and server is generated by computer programs running on respective computers and having a client-server relationship with each other. The server may be a cloud server, a server of a distributed system, or a server combined with a blockchain.

[0110] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps recorded in this disclosure can be executed in parallel, sequentially or in different orders, as long as the desired results of the technical solutions disclosed in this disclosure can be achieved, and this document does not limit this.

[0111] In addition, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of the features. In the description of the present disclosure, the meaning of "plurality" is two or more, unless otherwise clearly and specifically defined.

[0112] The above is only a specific embodiment of the present disclosure, but the protection scope of the present disclosure is not limited thereto. Any person skilled in the art who is familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed in the present disclosure, which should be included in the protection scope of the present disclosure. Therefore, the protection scope of the present disclosure should be based on the protection scope of the claims.

Claims

1. A data processing method, comprising: Obtaining first input data and a response result to the first input data; Performing identification processing on the response result to obtain first identification data; Based on the first input data and the first identification data, the response result is processed accordingly with a target processing strategy to output a target file; Among them, the target file is more consistent with the user intention of the target user who generated the first input data than the response result.

2. The method according to claim 1, wherein: Obtaining first input data and a response result to the first input data includes at least one of the following: Obtaining a first processing instruction for the first input data, and performing generation processing on the first input data using a first processing model matching the first processing instruction to obtain the response result; generating and processing the first target data in the first input data based on the type of the first input data and the represented user intention to obtain the response result; The first input data and context data of its source application are integrated to obtain second input data, and the second input data is generatively processed using a generative model to obtain the response result.

3. The method according to claim 2, wherein: Obtaining a first processing instruction for the first input data includes at least one of the following: generating the first processing instruction in response to obtaining a trigger operation of a first type of control acting on a first type of application, wherein the first type of application is an application capable of responding to the first input data; generating the first processing instruction in response to obtaining a trigger operation for first input data input to a second type of application, the second type of application being an application that can respond to or cannot respond to the first input data; In response to obtaining first input data input to a third type of application, generating a corresponding first processing instruction based on application information of the third type of application; A corresponding first processing instruction is generated based on the type of the first input data and / or application information of a source application.

4. The method according to claim 1, wherein: Performing identification processing on the response result to obtain first identification data includes: Determine a processing model or a processing procedure corresponding to the response result based on the first input data and / or a first processing instruction for the first input data; Performing identification processing on the response result based on a determined processing model or processing procedure to obtain the first identification data; The type of the first identification data is the same as or different from the type of the first input data.

5. The method according to claim 4, wherein: Determining a processing model or a processing procedure corresponding to the response result based on the first input data and / or a first processing instruction for the first input data includes at least one of the following: Determine a second processing model or a first processing program for performing identification processing on the response result based on the type of the first input data, so as to process the response result into first identification data of the same type by using the second processing model or the first processing program; Determine a third processing model or a third processing program for performing identification processing on the response result based on the first processing model or the second processing program called by the first processing instruction to process the first input data into the response result, so as to process the response result into the first identification data by using the third processing model or the third processing program; Determine a fourth processing model or a fourth processing program for performing identification processing on the response result based on the instruction content of the first processing instruction, so as to process the response result into the first identification data by using the fourth processing model or the fourth processing program; A fifth processing model or a fifth processing program for performing recognition processing on the response result is determined based on the content of the first input data or the represented user intention, so as to process the response result into the first recognition data using the fifth processing model or the fifth processing program.

6. The method according to claim 1, wherein: Based on the first input data and the first identification data, the response result is processed accordingly with a target processing strategy, including: A comparison result between the first input data and the first recognition data is obtained, and based on the comparison result, the response result is processed accordingly with a target processing strategy, wherein different comparison results correspond to different processing strategies.

7. The method according to claim 6, wherein: Based on the comparison result, the response result is processed accordingly with a target processing strategy, including at least one of the following: When the comparison result indicates that the similarity difference between the first input data and the first recognition data is less than a first threshold, directly outputting the response result as the target file, or generating the response result to obtain the target file; When the comparison result indicates that the similarity difference between the user intention represented by the first input data and the user intention represented by the first recognition data is less than a second threshold, the first input data is generated again and then the response result is replaced as the target file; When the comparison result indicates that the similarity difference between the first input data and the first recognition data is greater than a second threshold, the response result is optimized or compared with a generated result generated by regenerating the first input data to output the target file.

8. The method according to claim 7, wherein: Optimizing the response result or comparing the response result with a generated result of regenerating the first input data to output the target file includes at least one of the following: Optimizing the response result or comparing it with the generated result based on the similarity difference and / or the type of the response result, so as to output the optimized response result as a target file, or outputting the compared response result and / or the generated result as a target file; replacing or optimizing a target portion of the response result to obtain the target file, the target portion including a similarity difference portion between the response result and the first input data; Obtain a generation result obtained by generating and processing the first input data using a sixth processing model, compare the generation result with the response result, and output the response result and / or the generation result as a target file based on the comparison result, wherein the sixth processing model is set in the electronic device or other device different from the electronic device.

9. The method according to claim 1, wherein: Based on the first input data and the first identification data, the response result is processed accordingly with a target processing strategy, and at least one of the following is also included: Integrate the first input data and the first recognition data to obtain third input data, and replace the response result based on the generated result corresponding to the third input data to obtain the target file; Based on the user portrait information of the target user and / or the current environment parameters, the first input data and the first recognition data are evaluated for similarity, and based on the evaluation result, the response result is optimized to output the target file.

10. A data processing device, comprising: An acquisition module, used for acquiring first input data and a response result to the first input data; An identification module, used to perform identification processing on the response result to obtain first identification data; A processing module, configured to process the response result accordingly with a target processing strategy based on the first input data and the first identification data, so as to output a target file; Among them, the target file is more consistent with the user intention of the target user who generated the first input data than the response result.

Images