Data processing method and device

Abstract

The invention provides a data processing method and device, and the method comprises the steps: responding to a plurality of pieces of input data of a plurality of users, and recognizing a first user intention represented by the plurality of pieces of input data; generating processing is carried out based on the first user intention, a target file is obtained, and the target file is matched with the user intention represented by at least one piece of input data.

Description

Technical Field

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

[0002] For data generation projects involving multiple people (such as raw images, raw text, raw video, raw audio, etc.), a discrete or serial working method is usually adopted, which means that participants cannot generate the same data at the same time. For example, in the text-to-image scenario, one participant first enters a text description and generates a preliminary image based on the description. The next participant then iterates and optimizes the image, and so on, until all participants optimize the image of their previous participant. However, since each participant only focuses on their own task, the correlation between them is low, which may make the generated image not fit the text description or the expectations of some participants, thus affecting the overall effect of the final work. Summary of the invention

[0003] The present disclosure provides a data processing method, device, electronic device and storage medium.

[0004] According to a first aspect of the present disclosure, there is provided a data processing method, comprising:

[0005] In response to obtaining a plurality of input data from a plurality of users, identifying a first user intention represented by the plurality of input data;

[0006] A generation process is performed based on the first user intent to obtain a target file, wherein the target file matches the user intent represented by at least one input data.

[0007] According to a second aspect of the present disclosure, there is provided a data processing device, comprising:

[0008] A first processing module, configured to, in response to obtaining a plurality of input data from a plurality of users, identify a first user intention represented by the plurality of input data;

[0009] The second processing module is used to perform generation processing based on the first user intent to obtain a target file, wherein the target file matches the user intent represented by at least one input data.

[0010] According to a third aspect of the present disclosure, there is provided an electronic device, including:

[0011] at least one processor; and

[0012] a memory communicatively connected to the at least one processor; wherein,

[0013] 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.

[0014] 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. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] 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:

[0016] In the drawings, the same or corresponding reference numerals represent the same or corresponding parts.

[0017] Figure 1 A schematic diagram of a data processing method provided by an embodiment of the present disclosure is shown;

[0018] Figure 2 A flowchart of a keyword method provided by an application embodiment of the present disclosure is shown;

[0019] Figure 3 A schematic diagram of the structure of a data processing device provided by an embodiment of the present disclosure is shown;

[0020] Figure 4 A schematic diagram of the structure of an electronic device provided by an embodiment of the present disclosure is shown. DETAILED DESCRIPTION

[0021] 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.

[0022] In the following description, reference is made to “some embodiments”, which describe a subset of all possible embodiments, but it can be understood that “some embodiments” may be the same subset or different subsets of all possible embodiments and may be combined with each other without conflict.

[0023] In the following description, the terms "first\second" involved are merely used to distinguish similar objects and do not represent a specific ordering of the objects. It can be understood that "first\second" can be interchanged with a specific order or sequence where permitted, so that the embodiments of the present disclosure described herein can be implemented in an order other than that illustrated or described herein.

[0024] Unless otherwise defined, all technical and scientific terms used in this disclosure have the same meaning as those commonly understood by those skilled in the art to which this disclosure belongs. The terms used in this disclosure are only for the purpose of describing the embodiments of this disclosure and are not intended to limit this disclosure.

[0025] It should be understood that in the various embodiments of the present disclosure, the size of the serial number of each implementation process does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present disclosure.

[0026] Figure 1 A flow chart of a data processing method provided by an embodiment of the present disclosure is shown; Figure 1 As shown, the data processing method includes:

[0027] Step 101: in response to obtaining a plurality of input data from a plurality of users, identifying a first user intention represented by the plurality of input data;

[0028] Step 102: Generate a target file based on the first user intent, wherein the target file matches the user intent represented by at least one input data.

[0029] In some embodiments, the multiple users may be multiple users using the same device, or multiple users using multiple devices, each user using one device. The users may perform the same task together, and multiple users may collaborate to complete the task, such as generating an image, generating text, or generating a video.

[0030] The device may be a computer, a smart terminal, a tablet computer or other electronic device.

[0031] The multiple input data may include at least one of multiple types such as text data, voice data, image data, drawings, operation tracks, gesture tracks, etc.

[0032] Each input data may reflect the intention of the corresponding user. By identifying multiple input data respectively input by multiple users, the first user intention represented by the multiple input data may be obtained.

[0033] The target file may be an image, text, video, audio, UI (User Interface), computer program code, etc.

[0034] In one example, to identify the first user intent represented by the multiple input data, identification can be performed first based on each input data to obtain the user intent of each input data; then, the user intent of each input data in the multiple input data can be integrated to obtain the first user intent.

[0035] In another example, multiple input data may be integrated first to obtain integrated input data, and the first user intention may be obtained based on the integrated input data.

[0036] For example, in the task of generating an image, the input data of the first user is a text description of the image content (such as drawing an animal), the input data of the second user is a text description of the image art style (such as the image style is impressionism), and the input data of the third user is a text description of the image content (such as drawing a forest with animals). In one example, the user intention of each input data can be determined first, and the user intention of the first user is that the image content includes animals, the user intention of the second user is that the image is in impressionist style, and the user intention of the third user is that the image content includes a forest with animals. In this way, the intentions are integrated, and the first user intention obtained can be: showing a forest and animals in impressionist style. In another example, the input data can also be integrated to obtain: drawing animals, drawing a forest with animals, and an image style of impressionism. The above-mentioned first user intention can also be obtained based on the integrated input data.

[0037] In another example, when identifying the first user intent represented by the multiple input data, the user priority (different priorities represent different importance of input data of different users), data generation scenario (such as generating image scenario, generating text scenario, generating video scenario, generating audio scenario, etc.), usage environment information (such as the environment in which the target file is generated), etc. can be further considered to determine the weights between user intents or processing strategies. Of course, other factors that affect the determination of user intent can also be further considered, which are not limited here.

[0038] In this way, by accurately identifying the intentions of multiple users, misunderstandings and deviations can be reduced, and the target file can be generated in a collaborative mode, ensuring that the generated target file is more in line with the ideas of multiple users; on the basis of clarifying the intentions of the first user, unnecessary iterations in the generation process can be reduced, thereby improving efficiency.

[0039] In some embodiments, based on the first user intent, a processing model that matches the user intent, such as a raw image model, a raw text model, a raw video model, a raw audio model, a raw UI model, etc., can be used for generation processing; the aforementioned artificial intelligence generated content (AIGC, Artificial Intelligence Generated Content) model can be a local model, or a cloud model, or it can be a distributed AI model distributed on various terminals for generation processing to obtain the target file.

[0040] The target file matches the user intent represented by at least one input data, indicating that the finally generated file conforms to the user intent represented by each input data, or conforms to the user intent of one or several input data due to conflict or user priority or in combination with environment, usage scenario, etc.

[0041] When generating the target file, in addition to considering generation processing on one device, you can also consider distributed generation, that is, generating on multiple devices separately, and then performing image fusion, video fusion, audio fusion, text fusion, etc.

[0042] When generating the target file, the user portrait data of each user may also be considered for generation and processing, so as to generate data that conforms to their respective habits or preferences; and / or, the needs of the demander may be considered, where the demander is the demander or recipient of the target file; and / or, corrected data or updated data may be generated after conflicts among multiple input data, etc.

[0043] In this way, the generated target file is more in line with the expectations of multiple users and / or demanders, misunderstandings and errors are reduced, and the satisfaction and experience of users and / or demanders are improved.

[0044] In some embodiments, identifying the first user intention represented by the plurality of input data may include at least one of the following exemplary solutions:

[0045] Example 1: Obtain the second user intention represented by each input data, perform corresponding processing on the multiple second user intentions represented by the multiple input data, and obtain the first user intention, wherein the first user intention includes at least one of the multiple second user intentions, or the intention content of the first user intention matches at least one of the intention contents of the multiple second user intentions.

[0046] Specifically, corresponding processing of the multiple second user intentions respectively represented by the multiple input data may include: understanding the multiple second user intentions respectively represented by the multiple input data to obtain the second user intention of each user; integrating the multiple second user intentions to obtain the first user intention.

[0047] Here, when integrating, you can sort according to different weights and then filter, or perform weighted integration, or match according to historical information, or integrate according to association relationships.

[0048] If multiple second user intentions are weighted and integrated, the first user intention may include at least one of the multiple second user intentions, indicating that the first user intention can include one or more of the second user intentions. Multiple second user intentions are weighted and integrated, and each second user intention can be weighted and integrated according to its weight to obtain a balanced result that comprehensively considers the needs of all users.

[0049] If the integration is based on the contents of multiple second user intentions, the intention content of the first user intention can match at least one of the intention contents of the multiple second user intentions, and the intention content representing the first user intention can match the intention content of one or more second user intentions.

[0050] If sorting is performed according to different weights, different weights can be assigned to each second user intention according to preset standards (such as user importance, demand urgency, etc.), and intentions with higher weights are met first.

[0051] If matching is based on historical information, the historical information may include: the user's historical behavior data and preferences, etc. The user's historical behavior data and preferences can help understand the user's needs more accurately, and based on the historical information can help determine the intention of the current input data.

[0052] If integration is based on associations, the correlations between users' input data can be analyzed to understand how different intentions influence or complement each other. When integrating, these associations can be taken into account to produce more coordinated results.

[0053] In this way, by identifying the second user intention represented by multiple input data through any of the above methods, and then synthesizing the first user intention, it is possible to understand the user's needs more comprehensively and improve the accuracy and flexibility of intent recognition.

[0054] Example 2: Targeted processing is performed on the multiple input data to obtain first input data, and a first user intention represented by the first input data is identified, wherein the first input data includes at least one of the multiple input data, or the semantic content of the first input data matches at least one of the semantic contents of the multiple input data.

[0055] Specifically, target processing may include: integrating, sorting, filtering or further generating multiple input data according to the purpose of use, etc. Target processing may also include converting different types of input data, such as converting images into text or specific vector data, or converting voice or gestures into text or semantic data, etc.

[0056] The processed first input data is obtained by integrating, sorting, filtering according to the purpose of use or further generating processing, etc., and then the first input data is identified based on the intent understanding model or the semantic analysis model to obtain the first user intent.

[0057] Here, if a plurality of input data are integrated, the first input data includes at least one of the plurality of input data, and the representation of the first input data may include one or more of the input data;

[0058] If the integration or generation is performed based on semantic content, the semantic content of the first input data matches at least one of the semantic contents of the multiple input data, indicating that the semantic content of the first input data matches the semantic content of at least one input data.

[0059] In this way, by performing targeted processing on multiple input data, it is possible to effectively filter and focus on the user's important intentions, thereby improving the accuracy of intent recognition.

[0060] Example 3: Obtain target reference information, and identify the first user intention represented by the multiple input data with a corresponding processing strategy based on the target reference information.

[0061] Specifically, the target reference information may include: the type of input data (such as text data, voice data, image data, drawings, operation trajectories, gesture trajectories, etc.), data generation scenarios (such as generating image scenarios, generating text scenarios, generating video scenarios, generating audio scenarios, etc.), user information (user location, priority, authority or association relationship, etc.), usage environment information (such as the environment in which the target file is generated), device configuration information (such as hardware resource configuration, hardware performance, configuration of each model, etc.), device resource status (such as device memory usage, load, idle computing resources, network resources, etc.), etc.

[0062] Here, the first user intention represented by the plurality of input data can be identified based on the preset rules combined with the target reference information with a corresponding processing strategy. The preset rules, various processing strategies, etc. are set by developers based on application requirements, and the specific contents of the preset rules and processing strategies are not limited.

[0063] In this way, the corresponding processing strategy is selected in combination with the target reference information, making the intent recognition process more intelligent and personalized, further optimizing the user experience and providing personalized services to better meet their needs and expectations.

[0064] In some embodiments, correspondingly processing the plurality of second user intentions respectively represented by the plurality of input data to obtain the first user intention may include at least one of the following exemplary solutions:

[0065] Example 1: Obtain configuration weights corresponding to the multiple users, and perform weighted integration on the multiple second user intentions based on the configuration weights to obtain the first user intention.

[0066] Specifically, the first user intention may be a collection of second user intentions represented by various input data.

[0067] Configuration weight can be determined based on priority weight. The higher the priority, the greater the configuration weight. Priority can be determined by multiple users through negotiation, or it can be set based on other rules without limitation. Configuration weight can also be determined based on identity, such as creator and assistant, teacher and student, party A and party B, etc. Different identities have different configuration weights.

[0068] Different representations of configuration weights consider different proportions of the second user's intention represented by the corresponding input data when integrating. Generally speaking, the higher the configuration weight, the greater the configuration proportion considered when integrating. It can be used to resolve non-conflicting choices or conflicting choices during weighted integration. For non-conflicting choices, for example, user A wants to draw a cat on the lawn, and user B wants to draw a dog, then one of the cat or dog may be selected for generation based on the configuration weight; for conflicting choices, for example, user A wants to draw two cats on the lawn, and user B wants to draw three cats on the lawn, then two cats (the corresponding configuration weight of user A is larger) may be selected instead of three cats based on the configuration weight; for another example, user A and user C both want to draw a cat, but user B wants to draw a dog, and only one animal can be generated, even if the individual configuration weights of user A and user C are lower than that of user B, a cat will be generated because the weighted weights of user A and user C are greater than that of user B.

[0069] In this way, weighted integration is performed according to the configuration weights corresponding to multiple users, so as to more accurately identify and understand the intention of the first user.

[0070] Example 2: Obtain configuration weights corresponding to the multiple users, sort the multiple second user intentions based on the configuration weights, take the second user intention in the first sequence as the first user intention, or integrate the second user intention in the first target sequence to obtain the first user intention, and the first target sequence includes at least two sequences.

[0071] Specifically, the plurality of second user intentions are sorted based on the configuration weights, and the second user intention in the first sequence can be used as the first user intention, that is, the second user intention with the highest configuration weight can be used as the main data generator, such as the main graph giving priority to the intention needs of the first-ranked user. Alternatively, the second user intentions in the first target sequence can be integrated to obtain the first user intention, and the first target sequence includes at least two sequences. For example, the second user intentions of the second sequence and the third sequence can be integrated to obtain the first user intention for data generation; or, the first three sequences, the first four sequences, and so on.

[0072] In this way, by sorting the second user intents based on the configuration weights, the most relevant intents can be prioritized, thereby optimizing the data generation process and improving the accuracy and pertinence of the results.

[0073] Example 3: Obtain a first association relationship between the multiple second user intentions, and integrate the multiple second user intentions based on the first association relationship to obtain the first user intention.

[0074] Specifically, the first association relationship may represent the similarity or degree of deviation between the second user's intended contents, such as semantically similarity.

[0075] Accordingly, integrating the multiple second user intentions based on the first association relationship to obtain the first user intention may include:

[0076] The weight of the second user intention is adjusted based on the similarity or dissimilarity, or the second user intention is deleted and edited, etc.; the first user intention is obtained according to the result of the adjustment or deletion. It is also possible to further generate associated intentions between multiple second user intentions based on the first association relationship, and combine the multiple second user intentions in combination with the associated intentions to obtain the first user intention.

[0077] In this way, by adjusting or deleting the second user intent with lower similarity, the relevance and accuracy of the first user intent finally identified can be improved.

[0078] Example 4: The multiple second user intentions are input into the first generation model for generation processing to obtain the first user intention.

[0079] Specifically, the first generation model can be pre-trained, and the first generation model is used to splice or assist in expanding the user intention, etc. Here, the generation ability of the first generation model can be directly used to assist in splicing or assist in expanding different second user intentions, etc., to obtain the first user intention. When applied, the multiple second user intentions are input into the first generation model for generation processing to obtain the output result of the first generation model, that is, the first user intention. The first generation model has the functions of reasoning and planning, extracts key information according to multiple second user intentions, splices or assists in expanding the user intentions, and fuses them to obtain the first user intention. For example, the intention of user 1 is to draw apples, and the intention of user 2 is to draw bananas. The first generation model splices or assists in diffusion of the user intentions, and generates a plate (for placing bananas and apples) according to the above two intentions. The first user intention obtained can be to draw bananas and apples placed on the plate.

[0080] In this way, by splicing or expanding the capabilities of the first generative model, multiple second user intentions can be flexibly integrated to more comprehensively meet user needs.

[0081] Example 5: Obtain a second association relationship between the multiple second user intentions, input the multiple second user intentions and the second association relationship into a second generation model for generation processing, and obtain the first user intention.

[0082] Specifically, the second generation model can be pre-trained, and the second generation model is used for splicing or auxiliary expansion based on multiple second user intentions and the second association relationship. The second association relationship can characterize the similarity program or degree of deviation between the second user intention contents, such as the semantic similarity program. The second association relationship can also characterize the spatial relationship or temporal relationship between user intentions, such as generating a cake and generating a biscuit, and generating a plate for placing the cake and biscuits; the association relationship between people and water can be people swimming in the water, people drinking water, people rowing in the water, people fishing in the water, etc. The generation process is performed based on the correlation of different intentions in spatial position or behavioral logic to supplement the correlation between the two. When applied, the multiple second user intentions and the second association relationship are input into the second generation model for generation processing to obtain the output result of the second generation model, that is, the first user intention.

[0083] In this way, the second generation model can flexibly integrate multiple second user intentions to more comprehensively meet user needs. Moreover, by simultaneously considering multiple second user intentions and their associations, information can be captured more comprehensively, thereby improving the accuracy and richness of the generated first user intentions.

[0084] In some embodiments, performing target processing on the plurality of input data to obtain the first input data may include at least one of the following exemplary solutions:

[0085] Example 1: Extract key data and perform grouping processing on the multiple input data to obtain at least two data sets of different categories, and combine the data sets of different categories according to corresponding configuration weights to obtain the first input data.

[0086] Specifically, the key data includes: keywords or key features. Accordingly, according to the keywords or key features, at least two different categories of data sets can be obtained. The data sets of different categories can be combined according to the corresponding configuration weights, and the keywords are screened from the grouped data sets according to the weight priority for reorganization to obtain the first input data. For example, multiple input data include different types of painting descriptions, and the key data are extracted and grouped to obtain color feature set 1 (such as red, blue), theme feature set 2 (such as spring), and combined according to the configuration weights (such as the weight of red is higher than the weight of blue), such as: red is the main color, there can be a small amount of blue, representing spring.

[0087] In this way, by extracting key data and grouping them, important information can be identified, the organization of data can be improved, and more accurate first input data can be generated.

[0088] Example 2: The multiple input data are input into a third generation model for generation processing to obtain the first input data.

[0089] Specifically, the third generation model can be pre-trained. For example, if the input data is text and the third generation model is a model of text generation, the multiple input data are input into the third generation model for generation processing to obtain the first input data in text form. For example, input data 1 is: a landscape painting depicting autumn, showing leaves, with a blue sky as the background; input data 2 is: a tranquil lake, reflecting the surrounding mountains and white clouds. Input data 1 and input data 2 are input into the third generation model, which is a model of text generation. The third generation model outputs combined text information based on input data 1 and input data 2, such as obtaining the first input data as: integrating autumn leaves with a tranquil lake, showing leaves and mountains against the blue sky.

[0090] In this way, the third generative model can be used to achieve complex data generation and transformation, which can enhance the diversity and depth of data and meet specific needs.

[0091] Example 3: Obtain configuration weights corresponding to the multiple users, and perform weighted integration on the multiple input data based on the configuration weights to obtain the first input data.

[0092] Specifically, the configuration weight can be determined based on the priority weight. The higher the priority, the greater the configuration weight. The priority can be determined by multiple users through negotiation, or it can be set based on other rules without limitation. The configuration weight can also be determined based on identity, such as the main creator and assistant, teacher and student, Party A and Party B, etc. Different identities have different configuration weights. For example, the weight of the main creator, teacher or Party A is A, and the weight of the assistant, student or Party B is B. A is greater than B, such as A is 60% and B is 40%. If there are multiple assistants or students, the sum of the weights of multiple assistants or students can be B, and the weight of each assistant or student can be the same or different, such as Student 1 is 20%, Student B is 10%, and Student C is 10%.

[0093] When the multiple input data are weighted and integrated based on the configuration weights, different configuration weights will consider different proportions of input data when integrating. Generally speaking, the higher the configuration weight, the greater the configuration proportion considered when integrating. When weighted integration is performed, it can solve the choice when there is no conflict or the choice when there is conflict. For the choice when there is no conflict. For example, if the input data of user A is to draw a cat on the lawn, and the input data of user B is to draw a dog, then a cat or a dog may be selected for generation based on the configuration weight; for the choice when there is a conflict, for example, if the input data of user A is to draw two cats on the lawn, and the input data of user B is to draw three cats on the lawn, then two cats (the corresponding configuration weight of user A is larger) may be selected instead of three cats based on the configuration weight; for another example, if user A and user C both want to draw a cat, but user B wants to draw a dog, and only one animal can be generated, even if the individual configuration weights of user A and user C are lower than that of user B, a cat will be generated because the weighted weights of user A and user C are greater than that of user B.

[0094] In this way, by configuring weights to perform weighted processing on the input data, relevant data can be prioritized according to importance, thereby improving the accuracy and authority of the final result.

[0095] Example 4: Obtain task information of a target task, and based on the task information, select at least one second input data from the multiple input data for conversion processing and / or weighted integration to obtain the first input data.

[0096] Specifically, the target task refers to the task currently received by the device, such as the task of generating images, the task of generating meeting notes, the task of generating videos, the task of generating audio, etc. The task information mainly includes the task type, such as raw images, raw texts, raw videos, raw audio, etc., in order to filter the key second input data from multiple input data as the first input data to generate the target file.

[0097] If the second input data is unique, a conversion process is performed, including image-text conversion or text-turning amount conversion, etc.; if there are multiple second input data, weighted fusion can be further performed to obtain relatively complete first input data.

[0098] For example, in the task of generating meeting notes, multiple input data include: input data 1, some requirements about the meeting; and input data 2, PPT pictures about the meeting; when generating meeting notes, text recognition can be performed on the PPT pictures to obtain text content, and then the partial record of the meeting in input data 1 and the text content obtained based on the PPT pictures are integrated; when integrating, weights can be further considered. For example, when integrating, content with high weights can be highlighted or used as the focus or the main theme of the meeting.

[0099] In this way, processing in conjunction with task information can ensure the pertinence of the data and improve processing efficiency and the relevance of results.

[0100] Example 5: Obtain response results of the multiple input data and guidance information for the response results, generate the first input data based on first feedback information for the guidance information, and use the guidance information to update attribute parameters of the response results.

[0101] Specifically, the response result may be a generated image, audio, text, video or UI prototype (User Interface Prototype) and the like.

[0102] The guidance information may be a guidance prompt used to optimize the response result, such as a guidance prompt for prompt input; or an optimization suggestion or design suggestion for the generated UI.

[0103] Feedback information can be the user's selection of guidance information or input other than options. For example, if the user selects the option recommended by the guidance, the selected option will be used as feedback information; or, if the user denies the recommended option, the denied input data will be used as feedback information; or, if the user re-enters a new prompt, the newly entered prompt will be used as feedback information. For example, taking the feedback of UI design as an example, the guidance information is to use a minimalist style layout in the UI interface, and the feedback information is to select the minimalist style layout. Or, the guidance information is that the color of the UI login button is green, and the feedback information is that green is not suitable and it is hoped to be changed to red to match the brand tone.

[0104] In the UI generation scenario, the feedback of users or group users on the relevant UI design suggestions, or the feedback information after testing can also be considered. It can also be further considered to combine user portrait information (including single user portrait or group user portrait), or to combine the perceived current usage scenario environment to generate the first input data, etc.

[0105] Here, the attribute parameters of the response result can be updated according to the guidance information, mainly used to optimize the display effect of the image or the display effect and interaction effect of the UI, or to optimize the display effect of the text, or the video parameters of the video, or the audio parameters of the audio, etc.

[0106] Different types of response results correspond to different configuration parameters, such as image parameters for images, audio parameters for audio, string parameters for text, video parameters for videos, and UI display parameters and interaction logic for UI.

[0107] For example, taking the feedback on UI design as an example, based on user feedback, such as the color being red, the color parameters can be adjusted. Of course, the layout parameters, element size, font and other parameters can also be adjusted. Taking the feedback on video design as an example, based on user feedback or user portrait information, requirements such as video duration, video frame rate, background music, etc. can be determined, and the video can be optimized based on the determined requirements; taking the feedback on audio design as an example, based on user feedback or user portrait information, requirements such as audio duration, number of audio channels (mono or stereo or dual-channel, etc.), music style, etc. can be determined, and the audio can be optimized based on the determined requirements; taking the feedback on text as an example, based on user feedback or user portrait information, requirements such as font, font size, paragraph format, etc. can be determined, and the text can be optimized based on the determined requirements.

[0108] In this way, data is generated based on the response results and guidance information, and data attributes can be dynamically updated and optimized, so that the generated data is more in line with needs and user feedback.

[0109] Example 6: Process the multiple input data into third input data, and generate the first input data based on the obtained adjustment operation on the third input data.

[0110] Specifically, processing the multiple input data into third input data, and generating the first input data based on the obtained adjustment operation on the third input data can include: first weighted integration of the multiple input data to obtain the third input data; and then updating the weighted integration result (i.e., the third input data) based on the adjustment operation of each user or one of the users (which can be determined based on priority) to obtain the first input data.

[0111] For example, input data 1 is: providing an image of a river, and input data 2 is: providing an image of a forest. The two images are weighted and integrated, for example, a weight of 70% can be assigned to the river image and a weight of 30% can be assigned to the forest image. Assuming that multiple users want to adjust the integration result (third input data), user A (such as the user with the highest priority) is selected to make the adjustment, such as the user wants to enhance the color of the river and slightly reduce the details of the forest. Then the final first input data can be: an image in which the color of the river is more vivid and the forest area is relatively soft.

[0112] In this way, by processing the third input data and generating based on the adjustment feedback, a more flexible data output that better meets actual needs can be achieved, thereby improving the overall effect.

[0113] In some embodiments, identifying the first user intention represented by the plurality of input data with a corresponding processing strategy based on the target reference information may include at least one of the following exemplary solutions:

[0114] Example 1: Obtain type information of the multiple input data, determine the configuration weight of each user based on the type information and user information of the multiple users, and weightedly integrate the second user intentions of the multiple input data or their representations based on the configuration weight to obtain the first user intention.

[0115] Specifically, the first user intent may be a set of user intents represented by various input data, that is, including at least one user intent.

[0116] Type information may include: the type of user input data; here, the weight of each user may be configured considering the degree of matching between the user information and the type of user input data and the target task, such as in the raw picture scene, the configuration weight of inputting a stick figure is higher than the configuration weight of inputting text; in the generated video scene, the configuration weight of inputting gesture operation or posture action is higher than that of stick figure and text; for another example, in the raw picture scene, the configuration weight of the user information being a painter is higher than that of the user information being a writer. For example, in a raw picture task, user A provides a stick figure (input type: image, specifically an image of a forest with animals), and combined with his professional background (painter), the input is given a higher weight (70%), while user B provides a descriptive text (input type: text, requiring the drawing of a river), and combined with his background (writer), the input is given a lower weight (30%). Weighted integration of the plurality of input data or the second user intention represented by them can pay more attention to the intention of user A, such as the first user intention obtained is to draw an image of a forest with multiple animals and rivers.

[0117] Example 2: Obtain user portrait information of the multiple users, determine the configuration weight of each user based on the user portrait information, and perform weighted integration on the multiple input data or the second user intentions represented by them based on the configuration weight to obtain the first user intention.

[0118] Specifically, the configuration weight of each user is determined based on the user portrait information, that is, a weight is configured for each user according to the user portrait information. For example, weights are assigned according to each person's painting style model. User input that tends to be realistic may obtain a higher weight for object detail generation.

[0119] Specifically, the weighted integration may include: including splicing, fusion, and the like. Based on the configuration weights, the weighted integration of the multiple input data or the second user intentions represented by them may be performed, and a preset model or other means may be used to splice or fuse the multiple input data or the second user intentions represented by them. For example, in a raw image task, the weight of user A is higher at 70%, and the weight of user B is lower at 30%. The second user intention of user A is to draw a forest image with animals, and the second user intention of user B is to draw a forest image with rivers. The second user intentions are spliced ​​or fused to obtain the first user intention: to draw an image of a forest with multiple animals and rivers. If the second user intention of user A is to draw an image of a forest with animals but no rivers, because the weight of user A is higher than that of user B, the first user intention obtained may be to draw an image of a forest with multiple animals but no rivers.

[0120] Example 3: Obtain device information of a target device that performs a target task, and based on the device information, identify the first user intention on a first electronic device operated by the target user, or, identify corresponding user intentions on multiple electronic devices operated by multiple users to obtain the first user intention.

[0121] Specifically, the device information may include: device configuration information and device usage information, device hardware resource configuration and usage, and device intent understanding model or large model configuration and usage. Here, the device information of the target device that performs the target task is combined to determine whether to perform intent understanding on the first electronic device or on each device. In this way, the device that performs intent understanding is selected in combination with the specific status of the device, thereby improving the efficiency of intent understanding.

[0122] For example, when performing intent understanding on a first electronic device, such as a tablet computer, the device information includes the operating system, application programs, and available hardware resources. If the current hardware resources are less, and the tablet computer is running multiple other applications at the same time, and it is believed that the current tablet computer processing efficiency may not be high, then you can choose to perform intent understanding on each device. On the contrary, if the computing resources are sufficient, you can choose to perform intent understanding on the tablet computer. Here, the choice of which device to perform intent understanding depends on the requirements of the specific task and the hardware capabilities of the device. Simple tasks can be performed on a tablet computer, while complex or computationally intensive operations can be processed by each device itself. Of course, it is also possible to combine the functions of each device. For example, some devices do not have an intent understanding model, that is, they do not have the ability to understand intent, and can be handed over to other devices with corresponding capabilities for processing.

[0123] In some embodiments, obtaining a plurality of input data of a plurality of users includes at least one of the following:

[0124] Obtaining multiple input data from multiple users on the same electronic device for the same target task;

[0125] Obtaining multiple input data from multiple users on different electronic devices for the same target task;

[0126] and / or,

[0127] Identifying a first user intention represented by the plurality of input data includes:

[0128] Extracting and grouping the multiple input data to obtain at least two different categories of key data sets;

[0129] Based on the configuration weights corresponding to the various users, target key data are filtered from the combination of various key data, and the target key data are identified to obtain the first user intention.

[0130] Specifically, the multiple input data may be multiple input data applied to the same electronic device by multiple users for the same target task, for example, the same image generation task, text generation task or video generation task.

[0131] The electronic device may be a dual-screen device or a large-screen device with multiple input areas, and each user may input data using a different input area; or, the electronic device may be connected to multiple input components, and multiple users may also input data through different input components, such as a keyboard, microphone, drawing board, etc.

[0132] Multiple input data may also be multiple input data from multiple users on different electronic devices for the same target task, for example, data input can be performed on multiple devices simultaneously in response to the target task through conference links, similar applications, etc.

[0133] In this way, multiple ways are provided to obtain input data of multiple users so as to subsequently identify the intention of the first user.

[0134] Specifically, the key data includes: keywords or key features. According to the keywords or key features, at least two different categories of data sets can be obtained.

[0135] Here, obtaining multiple input data from multiple users may include: performing key data extraction and grouping processing on the multiple input data to obtain at least two key data sets of different categories; then filtering target key data from the combination of various key data based on the configuration weights corresponding to each user, identifying the target key data to obtain the first user intention.

[0136] For example, keyword 1, keyword 2, and keyword 3 are extracted from the input data of user 1, user 2, and user 3, respectively, and after grouping, key data set 1 (including keyword 1) and key data set 2 (including keyword 2 and keyword 3) are obtained. If there is only one keyword 1 in key data set 1, keyword 1 is selected as the target key data. If there are two keywords in key data set 2, you can choose from them.

[0137] The above two keywords may conflict or not conflict, such as Figure 2 As shown, if there is a conflict, here we can select according to the configuration weights corresponding to each user (the users corresponding to keyword 2 and keyword 3 respectively) to obtain the target key data. After selecting the target keyword from keyword 2 and keyword 3, the selection result can also be presented to multiple users, and multiple users can determine whether to use this as the result. If they disagree, they can continue to enter other keywords to determine the target keyword that meets the intentions of multiple users. For example, the input data of user 1 is: draw a forest in a cartoon style; the input data of user 2 is: draw a forest with animals but no rivers; the input data of user 3 is: draw a forest with rivers and animals; the screened description of the style of the set 1 includes: cartoon style, and the description of the element set 2 includes: forest, animal, river, no river; for the conflict between the river and no river in set 3, the selection can be made according to the weights of user 2 and user 3, and the result can be fed back to user 2 and user 3, and user 2 and user 3 can determine whether to use this as the result or whether to make adjustments.

[0138] In this way, filtering key data based on user-configured weights can provide more personalized and accurate services, optimize data utilization efficiency, and enhance user experience.

[0139] In some embodiments, the generation process is performed based on the first user intention to obtain the target file, which may include at least one of the following exemplary solutions:

[0140] Example 1: Generate and process each user intent in the first user intent to obtain multiple target sub-files, and concatenate or merge the target sub-files to obtain the target file.

[0141] Specifically, the first user intent may be a set of user intents represented by various input data, that is, including at least one user intent.

[0142] Here, each user intention may be generated separately to obtain multiple target sub-files, and then the multiple target sub-files may be spliced ​​or fused to obtain the target file. For example, each object in an image, or each paragraph in a text, or each image frame in a video may be generated first; and then each object, each paragraph, or each image frame may be fused into one target file.

[0143] Here, whether to stitch or fuse can be decided based on the type and association of the sub-files, such as images can be stitched, texts can be fused, videos can be stitched, and so on.

[0144] In this way, by processing each user's intention separately, generating multiple target sub-files and then splicing or merging them, different types of data can be integrated more flexibly. The splicing or fusion method is selected according to the characteristics of the sub-files, so that the final target file is more in line with user needs and the consistency and coherence of the content are improved.

[0145] Example 2: Obtain second feedback information on the generation processing result of the first user intention, and update the generation processing result based on the second feedback information to obtain the target file.

[0146] Here, an optimization function can be provided for the generated results. For example, the second feedback information can be an optimization suggestion for the raw image effect, or an optimization suggestion for the generated video. The second feedback information can also be an adjustment to the image parameters of the image or the corresponding position generated in the image; the result is optimized according to the second feedback information to obtain the final target file.

[0147] For example, the generated processing result obtained according to the first user's intention is an image or audio or video or UI, etc., which is displayed to multiple users, and the users provide feedback, such as: the style of the image, the element layout of the image, the length of the video, the special effects of a certain section of the video, the audio channel, the audio accompaniment, the color of the UI, the layout of the UI, etc. The generated processing result is updated according to the specific feedback to obtain the target file. In this way, providing an optimization function can effectively improve the quality of the generated results. Users can make specific improvement suggestions for raw images, audio, video, or UI, etc., to achieve a final effect that is more in line with user expectations. This iterative optimization process not only improves user satisfaction, but also enhances the intelligence level of the system.

[0148] Example 3: Generate a third user intent based on the third feedback information obtained for the first user intent, input the third user intent into the corresponding fourth generation model, and obtain the target file.

[0149] Here, a function of updating user intent based on user feedback is provided. Here, the third feedback information can also reflect the updated user intent. An updated third user intent is generated according to the third feedback information. A final target file, such as an image or video, audio, etc., is generated according to the updated third user intent. The fourth generated model can be a local AI model or a cloud model.

[0150] For example, the first user intention is to generate a comic-style landscape painting; the third feedback information is to increase the brightness of the picture and add some elements, such as flowers; the third user intention is combined with the third user intention to generate a third user intention: to generate a bright landscape painting and add elements of flowers. Or, the third feedback information is to change the comic style to a watercolor painting, then the first user intention is adjusted in combination with the third feedback information to obtain the third user intention: a watercolor-style landscape painting.

[0151] The third feedback information may be one or more, that is, one user may provide feedback multiple times, or multiple users may provide feedback once or multiple times respectively; the first user intention may also be adjusted according to multiple third feedback information to obtain the third user intention.

[0152] In this way, by updating user intent through feedback, we can ensure that the generated content is always consistent with the user's latest needs; updating intent based on real-time user feedback improves personalized service capabilities, making the final generated text, image or video more in line with user expectations.

[0153] In some embodiments, the generation process is performed based on the first user intention to obtain the target file, which may include at least one of the following exemplary solutions:

[0154] Example 1: Obtain device information of a target device, and based on the device information, perform the generation processing operation on a first electronic device operated by a target user, or on multiple electronic devices operated by multiple users, to obtain the target file.

[0155] Specifically, considering the influence of the device status on the effect and efficiency of generating the target file, the device information can be obtained, and the device information can reflect the idle state of the device (such as idle computing resources, idle memory resources, etc.), and the idle state of the device is used to determine whether to perform the image generation operation on the first electronic device or to perform the collaborative generation processing operation on multiple devices. For example, if the idle computing resources, idle memory resources, etc. of the first electronic device meet the computing resources and memory resource requirements required to generate the target file, the generation processing operation can be performed on the first electronic device, and if the idle computing resources, idle memory resources, etc. of the first electronic device do not meet the computing resources and memory resource requirements required to generate the target file, the collaborative generation processing operation can be performed on multiple devices with generation capabilities.

[0156] In this way, the operation of generating the target file is reasonably arranged according to the idle state of the device, so as to avoid the problems caused by resource competition while ensuring the effect of generating the target file, thereby improving efficiency.

[0157] Example 2: Obtain demand information of a target demander, perform generation processing based on the demand information and the first user intention, and obtain the target file. The target demander is the recipient of the target file.

[0158] Specifically, the demand information of the demander can be obtained, and the target file can be generated according to the demand information. Here, the demander is the demander or receiver of the target file, and the demand information can be image style, video format, video parameters, text theme style, etc.

[0159] Here, specific demand information can be obtained from the demander in advance, such as the required format, style and content, and the demand information is analyzed as an input data with the demander having the highest weight, and the fourth user intent is obtained by combining the first user intent, and the target file is obtained by generating and processing according to the fourth user intent. Alternatively, it is determined whether the first user intent conflicts with the demand information, and if not, the first user intent is continued to be used, and if so, the first user intent is adjusted, and the target file is obtained by generating and processing based on the adjusted first user intent.

[0160] In this way, generating a target file based on the demand information of the demander can accurately meet the demander's specific requirements for images, videos, texts, etc., and improve the pertinence and applicability of the target file.

[0161] Example 3: Obtain user portrait information of the multiple users, and update the generation result of the first user intention based on the user portrait information to obtain the target file.

[0162] Specifically, the user portrait information can reflect the habits or preferences of different users. After generating the generation result of the first user intention, the generation result can be adjusted in combination with the habits or preferences of different users to obtain a target file that meets the user preferences. For example, after generating an image, the image can be processed again based on the habits or preferences of different users to obtain an image that meets the user's habits.

[0163] Here, after obtaining the generated result, the generated result of the first user intention can be updated based on the user portrait information of multiple users to obtain the target file. For example, the generated result is an image, and based on the user portrait information of multiple users, it is determined that there are more users (or users with higher weight) who like warm colors (such as orange, yellow, pink). After the image is generated, it is detected whether the image tone is a warm tone or whether it includes the tone in the user portrait information. If not, the image tone can be adjusted to a warm tone. For another example, the generated result is a video, which contains the basic elements of the seaside sunset, such as the beach, the sea, the setting sun, etc., combined with the user preference information, it is determined that there are a considerable number of users (or users with higher weight) who have specific preferences when watching the seaside sunset video, such as the sea water is clearer and more transparent, and the background music of the video is the gentle sound of waves mixed with the cry of seagulls, then the video can be updated according to the user portrait information, the clarity of the sea water is enhanced by the color correction algorithm, and the background music is replaced with the gentle sound of waves mixed with the cry of seagulls, and finally the target file is obtained.

[0164] Here, after obtaining the generated result, the generated result of the first user intention can also be updated according to the user portrait information of each user to obtain the target file corresponding to each user. For example, if the generated result is a travel video, user A likes rock music, and user B likes country music, the background music of the travel video can be changed to rock music for user A, and the background music of the travel video can be changed to rock music and country music for user B.

[0165] In this way, by updating the generated results through user portrait information, personalized, accurate and dynamically adapted content to user needs can be provided.

[0166] In some embodiments, the method further comprises at least one of the following:

[0167] In the case where the target file is an image file, displaying identification information of a target image object on the image, the identification information including identification information of a target user;

[0168] When the user information of the plurality of users changes, the output parameter of the target file is updated.

[0169] Here, if the image file includes multiple image objects, such as cats, trees, objects, etc., identification information can be added to the above target image objects. The identification information can be the name, title, etc. of the creator corresponding to each image object.

[0170] The target user is the provider of input data required to generate the target image object. The identification information of one user can be displayed, or the identification information of all users can be displayed. The information can also be provided as a watermark according to the user's choice or willingness.

[0171] Here, after a user leaves or is replaced, the image output parameters and display parameters may be updated, such as display style, regenerating a new image, and so on.

[0172] In this way, the identification information enables users to quickly identify content related to themselves, which improves the personalization and targeting of images. In addition, the identification information can also be used to help better manage and track user preferences and behaviors, providing data support for subsequent optimization. In addition, the output parameters are updated in a timely manner according to changes in user information, ensuring that the generated content always meets the latest needs and improves the relevance of the content.

[0173] Figure 3 A schematic diagram of the structure of a data processing device provided by an embodiment of the present disclosure is shown; Figure 3 As shown, the data processing device comprises:

[0174] A first processing module, configured to, in response to obtaining a plurality of input data from a plurality of users, identify a first user intention represented by the plurality of input data;

[0175] The second processing module is used to perform generation processing based on the first user intent to obtain a target file, wherein the target file matches the user intent represented by at least one input data.

[0176] In some embodiments, the first processing module is used to perform at least one of the following:

[0177] Obtaining a second user intention represented by each input data, and performing corresponding processing on the multiple second user intentions respectively represented by the multiple input data to obtain the first user intention, wherein the first user intention includes at least one of the multiple second user intentions, or the intention content of the first user intention matches at least one of the intention contents of the multiple second user intentions;

[0178] Performing target processing on the multiple input data to obtain first input data, and identifying a first user intention represented by the first input data, wherein the first input data includes at least one of the multiple input data, or the semantic content of the first input data matches at least one of the semantic contents of the multiple input data;

[0179] Target reference information is obtained, and a first user intention represented by the plurality of input data is identified based on the target reference information using a corresponding processing strategy.

[0180] In some embodiments, the first processing module is used to perform at least one of the following:

[0181] Obtaining configuration weights corresponding to the multiple users, and performing weighted integration on the multiple second user intentions based on the configuration weights to obtain the first user intention;

[0182] Obtaining configuration weights corresponding to the multiple users, sorting the multiple second user intentions based on the configuration weights, taking the second user intentions in the first sequence as the first user intention, or integrating the second user intentions in the first target sequence to obtain the first user intention, wherein the first target sequence includes at least two sequences;

[0183] Obtaining a first association relationship between the plurality of second user intentions, and integrating the plurality of second user intentions based on the first association relationship to obtain the first user intention;

[0184] Inputting the plurality of second user intentions into a first generation model for generation processing to obtain the first user intention;

[0185] A second association relationship between the multiple second user intentions is obtained, and the multiple second user intentions and the second association relationship are input into a second generation model for generation processing to obtain the first user intention.

[0186] In some embodiments, the first processing module is used to perform at least one of the following:

[0187] Extracting key data and performing grouping processing on the multiple input data to obtain at least two data sets of different categories, and combining the data sets of different categories according to corresponding configuration weights to obtain the first input data;

[0188] Inputting the plurality of input data into a third generation model for generation processing to obtain the first input data;

[0189] Obtaining configuration weights corresponding to the multiple users, and performing weighted integration on the multiple input data based on the configuration weights to obtain the first input data;

[0190] Obtaining task information of a target task, and selecting at least one second input data from the plurality of input data based on the task information to perform conversion processing and / or weighted integration to obtain the first input data;

[0191] Obtaining response results of the plurality of input data and guidance information for the response results, and generating the first input data based on first feedback information for the guidance information, wherein the guidance information is used to update attribute parameters of the response results;

[0192] The plurality of input data are processed into third input data, and the first input data is generated based on the obtained adjustment operation for the third input data.

[0193] In some embodiments, the first processing module is used to perform at least one of the following:

[0194] obtaining type information of the plurality of input data, determining a configuration weight of each user based on the type information and user information of the plurality of users, and weightedly integrating the plurality of input data or the second user intentions represented by the plurality of input data based on the configuration weight to obtain the first user intention;

[0195] Obtaining user portrait information of the multiple users, determining a configuration weight of each user based on the user portrait information, and weighted integrating the multiple input data or the second user intentions represented by them based on the configuration weights to obtain the first user intention;

[0196] Obtain device information of a target device that performs a target task, and based on the device information, identify the first user intention on a first electronic device operated by a target user, or, identify corresponding user intentions on multiple electronic devices operated by multiple users to obtain the first user intention.

[0197] In some embodiments, the first processing module is used to perform at least one of the following:

[0198] Obtaining multiple input data from multiple users on the same electronic device for the same target task;

[0199] Obtaining multiple input data from multiple users on different electronic devices for the same target task;

[0200] and / or,

[0201] Identifying a first user intention represented by the plurality of input data includes:

[0202] Extracting and grouping the multiple input data to obtain at least two different categories of key data sets;

[0203] Based on the configuration weights corresponding to the various users, target key data are filtered from the combination of various key data, and the target key data are identified to obtain the first user intention.

[0204] In some embodiments, the second processing module is used to perform at least one of the following:

[0205] Generate each user intention in the first user intention to obtain multiple target sub-files, and concatenate or merge the target sub-files to obtain the target file;

[0206] Obtaining second feedback information on a generation result of the first user intention, and updating the generation result based on the second feedback information to obtain the target file;

[0207] A third user intent is generated based on the third feedback information obtained for the first user intent, and the third user intent is input into a corresponding fourth generation model to obtain the target file.

[0208] In some embodiments, the second processing module is used to perform at least one of the following:

[0209] Obtaining device information of a target device, and performing the generation processing operation on a first electronic device operated by a target user, or on multiple electronic devices operated by multiple users, based on the device information, to obtain the target file;

[0210] Obtaining demand information of a target demander, and performing generation processing based on the demand information and the first user intention to obtain the target file, wherein the target demander is a recipient of the target file;

[0211] User portrait information of the multiple users is obtained, and the generation result of the first user intention is updated based on the user portrait information to obtain the target file.

[0212] In some embodiments, the second processing module may also be used to perform at least one of the following:

[0213] In the case where the target file is an image file, displaying identification information of a target image object on the image, the identification information including identification information of a target user;

[0214] When the user information of the plurality of users changes, the output parameter of the target file is updated.

[0215] It is understandable that, when the data processing device provided in the above embodiment implements the corresponding data processing method, the above processing can be assigned to different program modules as needed to complete all or part of the above-described processing. In addition, the device provided in the above embodiment and the embodiment of the corresponding method belong to the same concept, and the specific implementation process is detailed in the method embodiment, which will not be repeated here.

[0216] An embodiment of the present disclosure provides a computer-readable storage medium storing executable instructions, wherein the executable instructions are stored. When the executable instructions are executed by a processor, the processor will be triggered to execute the data processing method provided by the embodiment of the present disclosure.

[0217] In some embodiments, the computer-readable storage medium may be a ferroelectric random access memory (FRAM), a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), a flash memory, a magnetic surface memory, an optical disk, or a CD-ROM; it may also be various devices including one or any combination of the above memories.

[0218] In some embodiments, executable instructions may be in the form of a program, software, software module, script or code, written in any form of programming language (including compiled or interpreted languages, or declarative or procedural languages), and may be deployed in any form, including as a stand-alone program or as a module, model, subroutine or other unit suitable for use in a computing environment.

[0219] By way of example, executable instructions may be deployed to be executed on one computing device, or on multiple computing devices located at one site, or on multiple computing devices distributed across multiple sites and interconnected by a communication network.

[0220] An embodiment of the present disclosure provides a computer program product, which includes a computer program / instructions. When the computer program / instructions are executed by a processor, the data processing method described in the present disclosure is implemented.

[0221] Figure 4 FIG. 1 shows a schematic diagram of the structure of an electronic device provided by an embodiment of the present disclosure. Figure 4As shown, the electronic device 40 includes: a processor 401 and a memory 402 for storing a computer program that can be run on the processor; when the processor 401 is used to run the computer program, it executes the data processing method provided by the embodiment of the present disclosure.

[0222] In actual application, the electronic device 40 may further include: at least one network interface 403. The various components in the electronic device 40 are coupled together via a bus system 404. It is understood that the bus system 404 is used to achieve connection and communication between these components. In addition to the data bus, the bus system 404 also includes a power bus, a control bus, and a status signal bus. However, for the sake of clarity, Figure 4 In the figure, various buses are labeled as bus system 404. There may be at least one processor 401. The network interface 403 is used for wired or wireless communication between the electronic device 40 and other devices.

[0223] The memory 402 in the embodiment of the present disclosure is used to store various types of data to support the operation of the electronic device 40 .

[0224] The method disclosed in the above embodiment of the present disclosure can be applied to the processor 401, or implemented by the processor 401. The processor 401 may be an integrated circuit chip with signal processing capabilities. In the implementation process, each step of the above method can be completed by the hardware integrated logic circuit in the processor 401 or the instruction in the form of software. The above processor 401 can be a general processor, a digital signal processor (DSP, DiGital Signal Processor), or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components, etc. The processor 401 can implement or execute the disclosed methods, steps and logic block diagrams in the embodiment of the present disclosure. The general processor can be a microprocessor or any conventional processor, etc. The steps of the method disclosed in the embodiment of the present disclosure can be directly embodied as a hardware decoding processor to execute, or the hardware and software modules in the decoding processor can be combined to execute. The software module can be located in a storage medium, which is located in the memory 402. The processor 401 reads the information in the memory 402 and completes the steps of the above method in combination with its hardware.

[0225] In some embodiments, the electronic device 40 can be implemented by one or more application-specific integrated circuits (ASICs), DSPs, programmable logic devices (PLDs), complex programmable logic devices (CPLDs), field-programmable gate arrays (FPGAs), general-purpose processors, controllers, microcontrollers (MCUs), microprocessors, or other electronic components to execute the aforementioned method.

[0226] 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.

[0227] 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.

[0228] 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: In response to obtaining a plurality of input data from a plurality of users, identifying a first user intention represented by the plurality of input data; A generation process is performed based on the first user intent to obtain a target file, wherein the target file matches the user intent represented by at least one input data.

2. The method according to claim 1, wherein: Identifying a first user intention represented by the plurality of input data comprises at least one of the following: Obtaining a second user intention represented by each input data, and performing corresponding processing on the multiple second user intentions respectively represented by the multiple input data to obtain the first user intention, wherein the first user intention includes at least one of the multiple second user intentions, or the intention content of the first user intention matches at least one of the intention contents of the multiple second user intentions; Performing target processing on the multiple input data to obtain first input data, and identifying a first user intention represented by the first input data, wherein the first input data includes at least one of the multiple input data, or the semantic content of the first input data matches at least one of the semantic contents of the multiple input data; Target reference information is obtained, and a first user intention represented by the plurality of input data is identified based on the target reference information using a corresponding processing strategy.

3. The method according to claim 2, wherein: Correspondingly processing the plurality of second user intentions respectively represented by the plurality of input data to obtain the first user intention includes at least one of the following: Obtaining configuration weights corresponding to the multiple users, and performing weighted integration on the multiple second user intentions based on the configuration weights to obtain the first user intention; Obtaining configuration weights corresponding to the multiple users, sorting the multiple second user intentions based on the configuration weights, taking the second user intentions in the first sequence as the first user intention, or integrating the second user intentions in the first target sequence to obtain the first user intention, wherein the first target sequence includes at least two sequences; Obtaining a first association relationship between the plurality of second user intentions, and integrating the plurality of second user intentions based on the first association relationship to obtain the first user intention; Inputting the plurality of second user intentions into a first generation model for generation processing to obtain the first user intention; A second association relationship between the multiple second user intentions is obtained, and the multiple second user intentions and the second association relationship are input into a second generation model for generation processing to obtain the first user intention.

4. The method according to claim 2, wherein: The target processing is performed on the plurality of input data to obtain first input data, including at least one of the following: Extracting key data and performing grouping processing on the multiple input data to obtain at least two data sets of different categories, and combining the data sets of different categories according to corresponding configuration weights to obtain the first input data; Inputting the plurality of input data into a third generation model for generation processing to obtain the first input data; Obtaining configuration weights corresponding to the multiple users, and performing weighted integration on the multiple input data based on the configuration weights to obtain the first input data; Obtaining task information of a target task, and selecting at least one second input data from the plurality of input data based on the task information to perform conversion processing and / or weighted integration to obtain the first input data; Obtaining response results of the plurality of input data and guidance information for the response results, and generating the first input data based on first feedback information for the guidance information, wherein the guidance information is used to update attribute parameters of the response results; The plurality of input data are processed into third input data, and the first input data is generated based on the obtained adjustment operation for the third input data.

5. The method according to claim 2, wherein: Identifying the first user intention represented by the plurality of input data based on the target reference information with a corresponding processing strategy includes at least one of the following: obtaining type information of the plurality of input data, determining a configuration weight of each user based on the type information and user information of the plurality of users, and weightedly integrating the plurality of input data or the second user intentions represented by the plurality of input data based on the configuration weight to obtain the first user intention; Obtaining user portrait information of the multiple users, determining a configuration weight of each user based on the user portrait information, and weighted integrating the multiple input data or the second user intentions represented by them based on the configuration weights to obtain the first user intention; Obtain device information of a target device that performs a target task, and based on the device information, identify the first user intention on a first electronic device operated by a target user, or, identify corresponding user intentions on multiple electronic devices operated by multiple users to obtain the first user intention.

6. The method according to claim 1, wherein: Obtaining multiple input data from multiple users, including at least one of the following: Obtaining multiple input data from multiple users on the same electronic device for the same target task; Obtaining multiple input data from multiple users on different electronic devices for the same target task; and / or, Identifying a first user intention represented by the plurality of input data includes: Extracting and grouping the multiple input data to obtain at least two different categories of key data sets; Based on the configuration weights corresponding to the various users, target key data are filtered from the combination of various key data, and the target key data are identified to obtain the first user intention.

7. The method according to claim 1, wherein: Generating and processing based on the first user intention to obtain a target file includes at least one of the following: Generate each user intention in the first user intention to obtain multiple target sub-files, and concatenate or merge the target sub-files to obtain the target file; Obtaining second feedback information on a generation result of the first user intention, and updating the generation result based on the second feedback information to obtain the target file; A third user intent is generated based on the third feedback information obtained for the first user intent, and the third user intent is input into a corresponding fourth generation model to obtain the target file.

8. The method according to claim 1 or 7, wherein: Generating and processing based on the first user intention to obtain a target file includes at least one of the following: Obtaining device information of a target device, and performing the generation processing operation on a first electronic device operated by a target user, or on multiple electronic devices operated by multiple users, based on the device information, to obtain the target file; Obtaining demand information of a target demander, and performing generation processing based on the demand information and the first user intention to obtain the target file, wherein the target demander is a recipient of the target file; User portrait information of the multiple users is obtained, and the generation result of the first user intention is updated based on the user portrait information to obtain the target file.

9. The method according to claim 1, further comprising at least one of the following: In the case where the target file is an image file, displaying identification information of a target image object on the image, the identification information including identification information of a target user; When the user information of the plurality of users changes, the output parameter of the target file is updated.

10. A data processing device, comprising: A first processing module, configured to, in response to obtaining a plurality of input data from a plurality of users, identify a first user intention represented by the plurality of input data; The second processing module is used to perform generation processing based on the first user intent to obtain a target file, wherein the target file matches the user intent represented by at least one input data.

Images