DATA PROCESSING METHODS AND ELECTRONIC EQUIPMENT

Abstract

A data processing procedure comprises obtaining a target input, obtaining initial information in response to the target input fulfilling a condition, and outputting secondary information based on the initial information. The secondary information specifies the time required to obtain a target output generated by a target model based on the target input, and this time is related to the initial information.

Description

REFERENCE TO RELATED REGISTRATION

[0001] This application claims priority over Chinese patent application No. 202411750356.7, filed on November 29, 2024, the entire contents of which are hereby incorporated by reference. TECHNICAL AREA

[0002] The present disclosure relates generally to the field of artificial intelligence and in particular to a data processing method and an electronic device. TECHNICAL BACKGROUND

[0003] With the advancement of artificial intelligence technology, model-based data processing systems are becoming increasingly popular. Current data processing systems have a single mode for outputting response information after receiving user input, which is less intelligent. OVERVIEW OF THE INVENTION

[0004] According to the disclosure, a data processing method is provided which includes obtaining a target input, obtaining initial information in response to the target input fulfilling a condition, and outputting secondary information based on the initial information. The secondary information specifies the time required to obtain a target output generated by a target model based on the target input, and this time is related to the initial information.

[0005] The disclosure also provides a data processing method that includes obtaining a target input, obtaining initial information related to an input object in response to the target input satisfying a condition, and outputting second information based on the initial information. The second information comprises a first output or a second output generated by a target model based on the target input, and the first output is different from the second output.

[0006] Also according to the disclosure, a data processing procedure is provided which includes obtaining a target input, obtaining information in response to the target input satisfying a condition, and determining a target policy for generating a target output based on the information.The target policy includes: selecting a target model that corresponds to the information from multiple models and generating a target output based on the target model; determining a number of calls to the target model based on the information; determining, based on the information, one or more steps by which the target model generates the target output based on the target input and target information; determining, based on the information, a depth of an information search in a predefined information source or a selected knowledge base based on the target input; and / or determining, based on the information, a breadth of an information search in the predefined information source or the selected knowledge base based on the target input. BRIEF DESCRIPTION OF THE DRAWINGS

[0007] The above and other features, advantages, and aspects of embodiments of the present disclosure will become clearer from the following detailed description in conjunction with the accompanying drawings. In all drawings, the same or similar reference numerals are used to refer to the same or similar elements. The figures are schematic, and elements and components are not necessarily drawn to scale. Fig. Figure 1 is a schematic flowchart of a data processing procedure provided by the present disclosure. Fig. 2A is a schematic flowchart of obtaining first information and outputting second information based on the first information according to the present disclosure. Fig. Figure 2B is a schematic diagram showing an interaction example of obtaining initial information after obtaining a target input and outputting second information based on the initial information, according to the present disclosure. Fig. 3A is another schematic flowchart of obtaining first information and outputting second information based on the first information according to the present disclosure. Fig. Figure 3B is a schematic diagram showing another interaction example of obtaining initial information after obtaining a target input and outputting second information based on the initial information, according to the present disclosure. Fig. 4A is another schematic flowchart of obtaining first information and outputting second information based on the first information according to the present disclosure. Fig. Figure 4B is a schematic diagram showing another interaction example of obtaining initial information after obtaining a target input and outputting second information based on the initial information, according to the present disclosure. Fig. 5A is another schematic flowchart of a further implementation of obtaining first information and outputting second information based on the first information according to the present disclosure. Fig. Figure 5B is a schematic diagram showing another interaction example of obtaining the first information after obtaining the target input and outputting the second information based on the first information according to the present disclosure. Fig. Figure 6 is a diagram showing an interaction example of obtaining initial information after obtaining a target input and outputting third information based on the initial information, according to the present disclosure. Fig. Figure 7A is another schematic flowchart of the data processing procedure according to the present disclosure. Fig. Figure 7B is a schematic diagram showing an interaction example for obtaining first information after obtaining the target input through the first input object A and outputting fourth information based on the first information according to the present disclosure. Fig. Figure 7C is a schematic diagram showing another interaction example for obtaining first information after obtaining the target input through the second input object B and outputting fourth information based on the first information according to the present disclosure. Fig. Figure 7D is a schematic diagram of another interaction example showing obtaining the first information after obtaining the target input through the first input object A and outputting the fourth information based on the first information according to the present disclosure. Fig. Figure 7E is a schematic diagram showing another interaction example for obtaining first information after obtaining the target input through the first input object A and outputting fourth information based on the first information according to the present disclosure. Fig. Figure 7F is a schematic diagram of another interaction example for obtaining first information and outputting fourth information based on the first information after obtaining a target input of an input object input according to the present disclosure. Fig. Figure 7G is a schematic diagram of another interaction example for obtaining first information after obtaining a target input of an input object input and outputting fourth information based on the first information according to the present disclosure. Fig. 7H is a schematic diagram showing an output of sixth information according to the present disclosure. Fig. Figure 8 is another schematic flowchart of a data processing procedure according to the present disclosure. Fig. Figure 9 is a schematic structure diagram of an electronic device according to the present disclosure. DETAILED DESCRIPTION OF THE EXECUTION FORMS

[0008] Embodiments of the present disclosure are described below with reference to the drawings. The terms used in the embodiments of the present disclosure serve only to explain the specific embodiments of the present disclosure and are not intended to limit the present disclosure. It is known to those skilled in the art that with technical progress and the emergence of new scenarios, the technical solutions provided in the embodiments of the present disclosure may also be applicable to similar technical problems.

[0009] The terms “first,” “second,” etc., in the description and claims of this application and the drawings mentioned above are used to distinguish similar elements and are not necessarily used to describe a specific order or sequence. The terms used in this way may be interchangeable under appropriate circumstances, and this merely represents one way of distinguishing objects with the same attributes in the embodiments of this disclosure. Furthermore, the terms “comprise” and “have” and all variations thereof are intended to cover non-exclusive inclusion, so that a process, method, system, product, or apparatus having a number of elements is not necessarily limited to those elements but may also have other elements not expressly listed or that are not inherent to such a process, method, product, or apparatus.are inherent to such an institution.

[0010] The data processing method provided by the embodiments of the present disclosure can be used for an electronic device, wherein the electronic device is equipped with a machine learning model, for example, a neural network model, and the neural network model can include, among others, a convolutional neural network (CNN), a deep convolutional network (DCN), a cyclic neural network (RNN), a large model, etc. The large model can include, among others, a large language model (LLM), a multimodal large model (MLLM), etc.

[0011] A flowchart of a data processing procedure provided by the present disclosure is included in Fig. 1 shown, which is described below.

[0012] A target input is obtained in S101.

[0013] Target input refers to interactive content entered by an input object (e.g., a user) and can include, among other things, text, file, image content and / or the like.

[0014] In S102, initial information is obtained when the target input fulfills a first condition.

[0015] The first condition is used to assess whether it is necessary to output second information to specify the time at which the input object achieves the target output based on the target input.

[0016] That the target input fulfills a first condition can include the fact that the target input is a complex problem.

[0017] In one example, the complexity of the target input can be determined. If the complexity is greater than the target complexity, it is determined that the target input is a complex problem. Otherwise, it is determined that the target input is not a complex problem. Determining the complexity of the target input can involve the following:

[0018] A semantic analysis is performed on the target input to determine the amount of semantics it contains (i.e., the amount of distinct semantic information it contains). The complexity of the target input correlates positively with the amount of semantics it contains; that is, the greater the amount of semantics in the target input, the greater its complexity.

[0019] Semantic analysis is performed on the target input to determine the correlation between all semantic information contained within it. The complexity of the target input correlates positively with the correlation between all semantic information contained within it; that is, the greater the correlation between all semantic information contained within the target input, the greater the complexity of the target input.

[0020] In one example, the target input can be processed by a pre-trained classification model to obtain a classification result that corresponds to the target input, where the classification result indicates whether the target input is a complicated problem.

[0021] The initial information may include information related to an input object, an input environment, or an electronic device.

[0022] In S103, secondary information is output based on the initial information. This secondary information is used to specify an initial time for obtaining a target output. The target output is generated by a target model based on the target input, and the initial time is related to the initial information.

[0023] According to the present disclosure, before the target model generates the target output based on the target input, the time to obtain the target output is estimated based on the first information, and then the second information is output so that the input object can get the returned time (i.e., the first time) of the target output.

[0024] According to the data processing method provided by the embodiments of the present disclosure, after obtaining the target input, if the target input satisfies the first condition, the second set of information is output based on the first set of information to inform the input object of the timeframe within which the target output can be obtained. The target model is then used to generate and output the target output based on the target input. In this way, the output mode is modified to provide the input object with a clear understanding of the specific situation of the interactive system that obtains the target output based on the target input. This increases the variety of modes of interaction with the input object, thereby improving the intelligence of the large-scale model-based data processing system and fulfilling the input object's needs for intelligent interaction.

[0025] In some embodiments, such as in Fig. As shown in Figure 2A, the above-mentioned process of obtaining the first information and outputting the second information based on the first information can include the following.

[0026] Initial sub-information is obtained in S201.

[0027] The first sub-information is contained within the first information; that is, the first information includes the first sub-information. The first sub-information characterizes the processing power of an electronic device executing the target model. The first sub-information may include, among other things, at least some information about the number of cores, main frequency, and cache size of the processor; memory capacity and frequency; hard drive type and storage capacity; video memory size; number of stream processors; bus width; performance of the installed software; and so on.

[0028] The processing power of the electronic device correlates positively with the number of cores of the processor, i.e., the more cores the processor has, the higher the processing power of the electronic device.

[0029] The processing power of the electronic device correlates positively with the main frequency of the processor, i.e., the higher the main frequency of the processor, the faster the operating speed and the higher the processing power of the electronic device.

[0030] The processing power of the electronic device correlates positively with the cache size of the processor; that is, the larger the cache size of the processor, the faster the data processing speed of the processor and the higher the processing power of the electronic device.

[0031] The processing power of the electronic device correlates positively with the storage capacity of the electronic device; that is, the larger the storage capacity of the electronic device, the higher the processing power.

[0032] The processing power of the electronic device correlates positively with the storage frequency of the electronic device, i.e., the higher the storage frequency of the electronic device, the faster the data transfer of the storage and the higher the processing power of the electronic device.

[0033] Different types of hard drives have different read and write speeds, and electronic components have different processing power. The processing power of the electronic component correlates positively with the hard drive's read / write speed; that is, the higher the hard drive's read / write speed, the higher the processing power of the electronic component.

[0034] The processing power of the electronic device correlates positively with the storage capacity of the hard drive, i.e., the larger the storage capacity of the hard drive, the higher the processing power of the electronic device.

[0035] The processing power of the electronic device correlates positively with the size of the video memory; that is, the larger the video memory, the greater the ability of the video card to process images and videos, and the higher the processing power of the electronic device.

[0036] The stream processor is a central processing unit in the graphics card, and the processing power of the electronic device correlates positively with the number of stream processors in the graphics card; that is, the higher the number of stream processors, the better the processing power of the graphics card and the higher the processing power of the electronic device.

[0037] The wider the bus width, the faster the data transmission and the higher the processing power of the electronic device.

[0038] The higher the performance of the software called by the electronic device, the higher the processing power of the electronic device.

[0039] In S202, an initial time (i.e., the time the target model needs to produce the target output based on the target input) is determined based on the first sub-information.

[0040] In some embodiments, the initial time required by the target model to generate the target output based on the target input is shorter the higher the processing power of the electronic device represented by the initial sub-information; conversely, the initial time required by the target model to generate the target output based on the target input is longer the lower the processing power of the electronic device represented by the initial sub-information.

[0041] S203 outputs a second piece of information. This second piece of information contains a first time, which was determined based on the first set of sub-information.

[0042] Fig. Figure 2B shows an interaction example for obtaining initial information after receiving a target input and outputting secondary information based on the initial information, as disclosed herein. In this example, the target input is "Help me keep up with some of the mainstream ideas recently expressed by famous experts from the three leading companies in the field of patent applications," and the secondary information is "OK, this question is quite complicated. Depending on the machine's performance, it will take me about 3 minutes to answer you."

[0043] In some embodiments, a further flowchart is used to obtain the first information and output the second information based on the first information. Fig. 3A shows which includes the following.

[0044] Second sub-information is obtained in S301.

[0045] The second set of sub-information is contained within the first set of information; that is, the first set of information encompasses the second set of sub-information. The second set of sub-information characterizes the allocation of available processing resources of the electronic device by a non-target model.

[0046] The processing resources of the electronic device may include, among other things, processors, memory, etc.

[0047] The available processing resources of the electronic device are the available processing resources of the electronic device when the electronic device receives the target input.

[0048] The non-target model can include other applications running on the electronic device alongside the target model. The non-target model can be determined based on the input object's appointment information. For example, if the user's calendar information shows that the user is in a meeting from 3:00 PM to 4:00 PM, then the non-target model could include meeting software.

[0049] The allocation of available processing resources by non-target models can include, among other things, the start time at which the available processing resources are allocated by the non-target model, the duration for which the available processing resources are allocated by the non-target model, the proportion of available processing resources allocated by the non-target model, the priority of the non-target model, etc.

[0050] In S302, a first time (i.e., the time the target model needs to produce the target output based on the target input) is determined based on the second set of sub-information.

[0051] In some embodiments, the target model and the non-target model occupy the available processing resources simultaneously before the non-target model begins to occupy them if the target model cannot generate the target output based on the target input. The second set of sub-information characterizes the non-target model's allocation of available processing resources. The longer the duration characterized by the second set of sub-information, the longer the initial period. Conversely, the larger the proportion characterized by the second set of sub-information, the smaller the proportion of available processing resources occupied by the target model, and thus the longer the initial period.The higher the priority of the non-target model, as characterized by the second set of sub-information, the greater the proportion of available processing resources occupied by the non-target model, the smaller the proportion of available processing resources occupied by the target model, and the longer the initial time.

[0052] In some embodiments, before the non-target model begins to occupy the available processing resources, if the target model is capable of generating a target output based on the target input, an initial time at which the target model generates the target output based on the target input is determined based on the currently available processing resources (i.e., if there are applications occupying the processing resources, the target model, without suspending the applications currently occupying the processing resources, only occupies a portion of the processing resources of an electronic device).

[0053] In some embodiments, before the non-target model begins to occupy the available processing resources, if the target model occupies all the processing resources of the electronic device (i.e., if there are applications currently occupying processing resources, these applications are suspended) and is capable of generating a target output based on the target input, the first time the target model generates the target output based on the target input is determined based on all the processing resources of the electronic device. This results in a shorter time to obtain the target output.In this case, interaction information can also be output to query whether the input object agrees to allocate all processing resources of the electronic device to generate the target output before the non-target model begins allocating those resources. If the input object chooses to agree to suspend the applications currently consuming the processing resources and allocate all processing resources of the electronic device to execute the target model, then the target model can rapidly generate a target output based on the target input. If the input object chooses not to agree, then the application program currently consuming the processing resources continues to execute, and the initial time is determined based on the two embodiments described above.

[0054] S303 outputs a second set of information. This second set of information contains the first time, which was determined based on the second set of sub-information.

[0055] Another example interaction diagram of obtaining initial information and outputting second information based on the initial information after obtaining the target input is in Fig. 3B is shown. In this example, the goal input is “Help me keep up with some of the mainstream ideas recently voiced by renowned experts from the three leading companies in the field of patent filing,” and the second piece of information is “Your question is quite complex. I see you have a meeting in 15 minutes. However, if you are willing to have me utilize all the facility's resources, I can probably calculate the answer to your question in 10 minutes. Would you like me to do this?”

[0056] In some embodiments, a further flowchart is used to obtain the first information and output the second information based on the first information. Fig. 4A shows which includes the following.

[0057] In S401, third-order sub-information is obtained. This third-order sub-information characterizes the input object's understanding of the domain to which the target input belongs.

[0058] The third set of sub-information is contained within the first set of information; that is, the first set of information encompasses the third set of sub-information. The third set of sub-information can include, among other things, the input object's browsing history, job title, years of professional experience, and so on. The more information about the field to which the target input belongs that is viewed by the input object and characterized by its browsing history, the better the input object's understanding of that field. The higher the rank of the input object's job title, the better the understanding of the field to which the target input belongs. The longer the input object's professional experience, the better the input object's understanding of the field to which the target input belongs.

[0059] In S402, based on the third set of sub-information, an initial time (i.e., the time the target model needs to produce the target output based on the target input) is determined.

[0060] The better the input object understands the domain to which the target input belongs, the more in-depth the response should be. Higher-quality responses require more time, so the initial response time is longer. Otherwise, the initial response time is shorter.

[0061] S403 outputs a second piece of information. This second piece of information contains the first time, which was determined based on the third set of sub-information.

[0062] In some embodiments, the first information includes the first sub-information, the second sub-information and / or the third sub-information.

[0063] If the initial information includes at least two of the three sub-information described above, the initial time (i.e., the time the target model needs to produce the target output based on the target input) can be determined by considering the at least two sub-information within the initial information. The associative relationship between the initial time and various sub-information can be found in the above revelation.

[0064] Another interaction example diagram of obtaining the first information after obtaining the target input, where the second information is output based on the first information, is in Fig. 4B is shown. In this example, the target input is “Help me keep up with some of the mainstream ideas recently expressed by renowned experts from the three leading companies in the field of patent applications.” and the second piece of information is “Your question is quite complex. Considering that you are a professional in the patent field, I need to search and organize the relevant materials in great depth and will reply to you in approximately 12 minutes.”

[0065] In some embodiments, a further flowchart of obtaining the first information and outputting the second information based on the first information is shown. Fig. 5A shows which includes the following.

[0066] In S501, fourth sub-information is obtained.

[0067] The fourth set of sub-information is contained within the first set of information; that is, the first set of information includes the fourth set of sub-information. The fourth set of sub-information relates to the input object and / or the input environment.

[0068] In some embodiments, the fourth set of sub-information relating to the input object can characterize the input object's personality. In this disclosure, the fourth set of sub-information can include a user personality, determined based on the input object's daily behavior, where the user personality characterizes whether the input object is patient. For example, after entering information, the input object often performs other operations before the target output is fully displayed; or the input object often moves the mouse aimlessly or clicks the mouse, etc., while waiting for the target output. All of these behaviors indicate that the input object has an impatient personality. In this case, a label "impatient" can be added to the user personality to indicate that the user has an impatient personality.

[0069] In some embodiments, the fourth set of sub-information relating to the input environment may include the time and location where the input object is currently located.

[0070] In S502, a second time is determined based on the fourth set of sub-information. This second time characterizes the time the input object can wait.

[0071] In some embodiments, if the input object has a more impatient personality, it can only wait a comparatively short time, and if the input object has a more patient personality, it can wait longer.

[0072] If the input object is currently in the first target time period (e.g., 9:00 AM to 10:00 PM) and is located at the first target location (e.g., company), then the input object is considered to still be at work and hopes to quickly achieve the target output; that is, the input object can only wait a relatively short time.

[0073] If the input object is not currently in the first target time period and is located at a second target location (e.g., at home), then it is assumed, for example, that the input object may need sleep and can wait a comparatively long time.

[0074] S503 outputs secondary information based on the first and second timestamps. This secondary information is also used to instruct the input object to trigger execution of the target model or to perform other tasks.

[0075] That is, according to the present disclosure, before the target output is generated by the target model based on the target input, in addition to obtaining the first time at which the target model generates the target output based on the target input, the second time that the input object can wait is also obtained.

[0076] The second set of information provided not only indicates the first time to obtain the target output, but also prompts the input object to trigger an execution of the target model in order to output the target output within the second or third time, or prompts the input object to perform other tasks.

[0077] The input object can perform other tasks using the electronic device (e.g., it can continue to use the electronic device while the target model simultaneously generates a target output based on the target input in the background) or it can perform no other tasks using the electronic device (e.g., it can sleep, read books, exercise, etc.).

[0078] Another interaction example for obtaining the first piece of information after obtaining the target input and outputting the second piece of information based on the first piece of information is in Fig. 5B is shown. In this example, the target input is “Help me keep up with some of the mainstream ideas recently voiced by renowned experts from the three leading companies in the field of patent filing.” The second piece of information is “Your question is quite complicated. I need more time. It seems you are going to bed now. Can I give you the answer tomorrow morning?”

[0079] If the fourth set of sub-information characterizes the input object as having a more patient personality, the second set of information prompts the input object to trigger an execution of the target model, so that it produces a target output at a second time.

[0080] If the fourth set of sub-information indicates that the input object has a more patient personality and its waiting time is shorter than a certain time threshold, the second set of information prompts the input object to trigger an execution of the target model, causing it to output the target output at a third time. This third time is determined based on the second time and is longer than the second time, with the difference between the third and second times being smaller than the preset difference.The input object is comparatively impatient: If the second set of instructions instructs the input object to trigger execution of the target model to output the target output within the second time frame, the input object may be impatient to see the output result when its wait time is about to reach the second time frame—at which point the full result has not yet been output. However, if the second set of instructions instructs the input object to trigger execution of the target model to output the target output within the third time frame, the target output may be obtained before the input object's wait time reaches the third time frame, thus improving the user experience.

[0081] If the first information includes third sub-information, in some embodiments the third sub-information characterizes the input object's understanding of a domain to which the target input belongs, and if the input object's understanding of the domain to which the target input belongs satisfies a second condition, the third information can be output based on the first, the third information being used to prompt the input object to select the knowledge base.

[0082] For example, the understanding of a domain to which the target input belongs by the input object includes the second condition that the degree of understanding of the domain to which the target input belongs by the input object is greater than a degree threshold.

[0083] If the input object's understanding of the field to which the target input belongs meets the second condition, this indicates that the input object has a comparatively better understanding of the field to which the target input belongs and is a professional in that field, making it necessary to provide a more in-depth target output. To provide a more in-depth target output, reference can be made to more professional knowledge bases. More professional knowledge bases may be subject to a fee. At this point, the third piece of information can be provided to ask the input object if it is willing to pay for access to a more professional knowledge base.

[0084] Fig. Figure 6 is a schematic diagram of an interaction example for obtaining initial information and outputting third information based on the initial information after obtaining a target input according to an embodiment of the present disclosure. In this example, the target input is "Help me keep up with some of the mainstream ideas recently expressed by renowned experts from the three leading companies in the patent application field." The third information is: "Your question is quite complex. Considering that you are a person skilled in the patent field, I need to search and organize the relevant materials in great depth. If you allow me to use a paid knowledge base, I can provide a higher-quality answer. Would you consider using the paid knowledge base?"

[0085] If the user chooses to pay to select a more professional knowledge base, then the target model references the more professional knowledge base to generate a target output based on the target input; otherwise, it does not reference the more professional knowledge base.

[0086] Another implementation flowchart of the data processing method provided by the embodiments of the present disclosure is shown in Fig. 7A shows which includes the following.

[0087] A target input is obtained in S701.

[0088] The target input is interactive content entered by an input object (e.g., a user) and can include content in the formats of text, files, images, and / or the like.

[0089] In S702, initial information is obtained when the target input fulfills the first condition. This initial information is related to the input object.

[0090] The first condition is used to determine whether personalized output is needed for the input object.

[0091] In some embodiments, the requirement that the target input fulfills the first condition may include that the target input is a complex question. To determine whether the target input is a complex question, reference can be made to the embodiments described above.

[0092] In some embodiments, the initial information may include characteristic information of the input object and / or the environment in which the input object is located. The characteristic information of the input object may include, among other things, identity information, personality, browsing history, job title, years of professional experience, and / or other properties of the input object. The environment in which the input object is located may include, among other things, the time of day, the location of the input object, and / or the like.

[0093] In S703, fourth pieces of information are output based on the first pieces of information. These fourth pieces of information comprise either a first output or a second output. The first output or the second output is generated by the object model based on the object input, and the first output is different from the second output.

[0094] This means that for different initial pieces of information obtained, the fourth pieces of information generated by the object model based on the object input will differ.

[0095] In some embodiments, a knowledge base can be determined based on the initial information. Relevant knowledge is then searched for in this knowledge base using the target input, and further information is generated by the target model based on the target input and the sought-after relevant knowledge.

[0096] Depending on the initial information, the specific knowledge base can vary. If the specific knowledge base differs, the related knowledge being sought will also differ. Therefore, the fourth pieces of information generated by the target model based on the target input and the related knowledge being sought will also differ.

[0097] In some embodiments, the related knowledge can be searched for in a predefined knowledge base based on the target input (i.e., the related knowledge is searched for in the same knowledge base independently of the initial information), and the fourth piece of information can be generated by the target model based on the target input, using the searched related knowledge and the initial information. Therefore, if the initial information differs, the fourth piece of information generated by the object model based on the searched related knowledge and the initial information, based on the object input, will also differ.

[0098] According to the data processing method provided in the embodiments of this disclosure, after obtaining the target input, if the target input satisfies the first condition, the first information relating to the input object is obtained, and the fourth information is output based on the first information. The fourth information comprises the first output or the second output. The first output or the second output is generated by the target model based on the target input, and the first output is different from the second output. That is, the different outputs are obtained based on the different first information relating to the input object, and the output mode is changed so that the input object receives the personalized output related to the first information.In this way, the interaction mode with the input object is increased, the intelligence of the data processing system based on the large model is improved, and the input object's requirement for interaction intelligence is met.

[0099] In some embodiments, outputting the fourth piece of information based on the first piece of information includes the following.

[0100] If the first pieces of information characterize the input object as a first input object, the fourth pieces of information, comprising the first output, are output. As in Fig. As shown in Figure 7B, after obtaining the target input of the first input object A, the first piece of information is retrieved, and the fourth piece of information is output based on the first piece of information. In this example, the target input is "Help me keep up with some of the mainstream ideas recently voiced by famous experts from the three leading companies in the field of patent filing," and the first output is "The leading companies in the field of patent filing typically include..."

[0101] If the initial information identifies the input object as a second input object, the fourth piece of information, comprising the initial output, is displayed. The first input object differs from the second input object. Another example is in Fig. Figure 7C illustrates this: After the target input is obtained by the second input object B, the first piece of information is retrieved, and the fourth piece of information is output based on the first piece of information, according to the embodiments of this disclosure. In this example, the object input is “Help me keep up with some of the mainstream ideas recently expressed by renowned experts from the three leading companies in the field of patent applications,” and the second output is “Understanding mainstream ideas in the patent field can help you better understand the significance and commercial value of intellectual property rights…”.

[0102] This means that if different input objects provide the same target input, the initial information associated with the different input objects is obtained, and the initial information associated with the different input objects is generally different, and thus the target outputs generated by the target model based on the target inputs are also different.

[0103] In some embodiments, outputting the fourth piece of information based on the first piece of information includes the following.

[0104] If the first piece of information indicates that the time the input object can wait satisfies the second condition, the fourth piece of information, comprising the first output, is displayed. The quality of the first output is higher than the quality of the second output. As in Fig. As shown in Figure 7D, after the first input object A receives the target input, the first piece of information is obtained, and the fourth piece of information is output based on this initial information. In this example, the target input is "Why is global climate change leading to an increase in extreme weather events?" and the first output is "Global climate change is a complex and multidimensional problem that...". In this example, the first input object A has enough time to wait for the model to generate a response; therefore, the content of the first output is richer and more in-depth.

[0105] If the time the input object can wait meets the second condition, it indicates that the input object can wait longer and the target model has enough time to produce higher-quality output. If the time the input object can wait does not meet the second condition, it indicates that the input object can wait less time and the target model does not have enough time to produce higher-quality output; therefore, only lower-quality output can be obtained. As in Fig. As shown in Figure 7E, after the first input object A receives the target input, the first piece of information is retrieved, and the fourth piece of information is output based on the first piece of information. In this example, the target input is "Why is global climate change leading to an increase in extreme weather events?" and the second output is "Considering that you have a meeting in 5 minutes, for the sake of brevity, the following is provided...". In this example, the first input object A only has a few minutes to wait for the model to generate a response; therefore, the contents of the second output are comparatively short.

[0106] That is, if the first information associated with the first input object characterizes that the time the first input object can wait satisfies the second condition, and the first information associated with the second input object characterizes that the time the second input object can wait does not satisfy the second condition, then the quality of the first output is higher than the quality of the second output.

[0107] If the first pieces of information associated with the first input object indicate that the waiting time of the first input object satisfies the second condition, then the fourth set of information, comprising the first output, is displayed. Conversely, if the first pieces of information associated with the first input object indicate that the waiting time of the second input object does not satisfy the second condition, then the fourth set of information, comprising the second output, is displayed. The quality of the first output is higher than the quality of the second output.

[0108] In some embodiments, outputting the fourth piece of information based on the first piece of information includes the following.

[0109] If the initial information indicates that the input object's understanding of the domain to which the target input belongs fulfills the third condition, the fourth piece of information, comprising the first output, is displayed. The quality of the first output is higher than the quality of the second output. Another interaction example is in Fig. Figure 7F shows that after the input object receives the target input, the fourth piece of information is output based on the first piece of information, according to the embodiments of this disclosure. In this example, the target input is "How are applicable laws and jurisdictions determined in an international business dispute involving multinational laws?" and the first output is "In an international business dispute involving multinational laws, the determination of relevant laws and jurisdictions is a complex and crucial issue. Below are some specific procedures and suggestions based on knowledge and practical experience in the field of law...". In this example, the input objects are legal professionals.Therefore, the answers given are comparatively comprehensive (some content was omitted due to space limitations) and of comparatively high quality.

[0110] The input object's understanding of the field to which the target input belongs fulfills the third condition, indicating that the input object is a person skilled in the art in that field and that a more in-depth output of the content is necessary. If the input object's understanding of the field to which the target input belongs does not fulfill the third condition, this indicates that the input object has less knowledge of the field to which the target input belongs, and a superficial and understandable output of the content is required. As in Fig. As shown in Figure 7G, the fourth piece of information is output based on the first piece of information after obtaining the target input from the input object, according to the embodiments of this disclosure. In this example, the target input is "How are applicable laws and jurisdictions determined in an international business dispute involving multinational laws?" and the second output is "In an international business dispute involving multinational laws, determining the relevant laws and jurisdictions is a complex and crucial issue. Some basic guidelines and procedures follow...". In this example, the input object is not a person skilled in the art and has little knowledge of law; therefore, the answer given is simple (part of it has been omitted for brevity) and of comparatively low quality.

[0111] That is, if the first information associated with the first input object characterizes that the first input object's understanding of the domain to which the target input belongs satisfies the third condition, and the first information associated with the second input object characterizes that the second input object's understanding of the domain to which the target input belongs does not satisfy the third condition, then the quality of the first output is higher than the quality of the second output.

[0112] In some embodiments, outputting the fourth piece of information based on the first piece of information includes the following.

[0113] If the initial information characterizes that the input object is a first input object and is located in a first scene, the fourth piece of information, comprising the first output, is output as shown in Fig. 7D is shown.

[0114] If the initial information indicates that the input object is a first input object and is located in a second scene, the fourth piece of information, comprising a second output, is displayed. The first scene differs from the second scene, as described in Fig. 7E is shown.

[0115] This means that if the same input object provides a target input in different scenes and different initial information associated with the input object is obtained, the target model will generate different target outputs based on the target input.

[0116] The first and second scenes can represent the different input environments in which the input object is located. For example, the first scene involves the input object being located in a primary target time period (e.g., 9:00 AM to 10:00 PM) and in a primary target location (e.g., the office). The second scene involves the input object being located in a non-primary target time period (e.g., 10:00 PM to 9:00 AM the following day) and in a non-primary target location (e.g., home).

[0117] If the input object is in different scenes, the time the input object can wait may vary, and the quality of the first and second outputs may differ.

[0118] In some embodiments, the output of the fourth piece of information based on the first piece of information may further include the following.

[0119] Fifth and / or sixth information will be provided.

[0120] The fifth piece of information includes at least one piece of source data and an association display relationship between the source data and an output contained in the fourth piece of information. The source data can include, but is not limited to: articles, software (also referred to as applications), a knowledge base, and / or the like.

[0121] This means that when the fourth piece of information is output, the associated source data is also output. This data is referenced when the target model generates the target output (first output or second output) based on the target input. The association between the fourth piece of information and the source data can include, among other things, the association of each segment in the fourth piece of information with the source data.

[0122] Furthermore, after obtaining the selection instruction for one of the segments in the fourth set of information, the source data associated with one of the segments are highlighted, so that the input object knows that the selected segment was created based on the highlighted source data.

[0123] The sixth piece of information comprises source data with a label that characterizes the input object as associated with the source data. As in Fig. The 7H diagram refers to five documents to obtain the fourth piece of information. The second document, "2. COP26 | Interpretation of the Climate Report: Extreme weather events are a major threat from global warming," is accompanied by an eye pattern. associated, which indicates that the input object has viewed this document.

[0124] This means that if at least one piece of source data contains source data associated with the input object, for example, source data that was viewed by the input object or source data that was created by the input object, then when outputting the source data associated with the input object, an identifier is added to the source data so that the input object knows that it is associated with the identified source data.

[0125] If the association between the input object and the source data differs, the labeling of the source data can also differ. For example, if the source data has a first label, this indicates that the input object has viewed the source data. If the source data has a second label, this indicates that the input object created the source data.

[0126] In some embodiments, a flowchart of a further implementation of the data processing method provided by the present disclosure is shown in Fig. 8 shown, which includes the following.

[0127] A target input is obtained in S801.

[0128] The target input is interactive content entered by an input object (e.g., a user) and can include content in the formats of text, files, images, and / or the like.

[0129] In S802, the first information is obtained when the target input meets the first condition.

[0130] The first condition is used to determine whether an output policy should be selected.

[0131] In some embodiments, the requirement that the target input fulfills the first condition may include that the target input is a complex problem. Reference may be made to the disclosure above to determine whether the target input is a complex problem.

[0132] In some embodiments, the first information may include information that relates to at least one of the following: an input object, an input environment, and an electronic device.

[0133] In some embodiments, the information related to the input object and / or the input environment may include, among other things, the identity of the input object, personality, time of day, location, browser history, job title, years of professional experience, and / or the like.

[0134] The information associated with the electronic device may include, but is not limited to: the number of cores, the main frequency and cache size of the processor, the storage capacity and frequency, the hard disk type and storage capacity, the video memory size, the number of stream processors, the bus width, the software performance and / or the like.

[0135] In S803, a target policy is determined based on the initial information.

[0136] If the initial information differs, then the specific target policy will also differ.

[0137] The target policy characterizes a policy for generating a target output, which may include the following.

[0138] A model that matches the initial information is selected as the target model from several models based on the initial information, and a target output is generated based on the target model.

[0139] The number of calls to the target model is determined based on the initial information.

[0140] The steps for generating the target output by the target model based on the target input are determined using the initial information.

[0141] The depth of an information search (i.e., the information search depth) in a preset information source (such as a website or a freely available knowledge base, etc.) or a selected knowledge base based on a target input is determined from the initial information.

[0142] The breadth of an information search (i.e., the information search width) in a preset information source or a selected knowledge base based on a target input using the initial information.

[0143] Any two models from the multiple models are different models, and the differences between the models may include: different publishers of the models, different types of models (e.g., models that are good at abstracting, models that are good at inferring, models that are good at classifying, etc.), different versions of the models, and / or the like.

[0144] According to the preset correspondence relationship between the initial information and the models, the model among the models that corresponds to the obtained initial information can be used as the model that corresponds to the initial information, i.e., the target model.

[0145] Or each of the multiple models can be used as a category to categorize the initial information, to determine which of the multiple models the initial information belongs to, and the specific category to which the initial information belongs can be determined as the model that corresponds to the initial information, i.e., the target model.

[0146] Alternatively, the waiting time of the input object and / or the input object's understanding of the target input domain can be determined based on the initial information. According to the predetermined match between the waiting time and / or the degree of understanding of the target input domain and the model, the model that corresponds to the input object's waiting time and / or degree of understanding of the target input domain is determined to be the model that corresponds to the initial information, i.e., the target model.

[0147] Alternatively, the requirements of the input object can first be determined based on the target input. After the waiting time of the input object and / or the level of understanding of the input object for the target input domain has been determined using the initial information, the model corresponding to the waiting time of the input object and / or the level of understanding of the input object for the target input domain can be identified as the model corresponding to the initial information, i.e., the target model, which is the model capable of realizing the requirements of the input object with respect to the predefined correspondence relationship between the waiting time and / or the level of understanding of the input object for the target input domain and the model.

[0148] There can be only one object model or multiple object models. If multiple object models exist, they can be similar models from different publishers or similar models from different versions. Target outputs can be generated by each model based on the target input, and then the final target output can be selected from the outputs generated by different target models, or the outputs generated by different target models can be combined to obtain the final target output.

[0149] In some embodiments, after determining the target model, the depth of an algorithm for generating the target output can also be determined by the target model based on the target input; that is, the number of calls to the target model based on the initial information. For example, if the initial information indicates that the input object can wait longer or has a better understanding of the domain to which the target input belongs, the number of calls to the target model can be greater to provide a more accurate and in-depth output; otherwise, the number of calls to the target model can be smaller to provide a more accurate and understandable output.

[0150] In some embodiments, after defining the target model, the steps for generating the target output based on the target input can be determined by the target model using the initial information. For example, the number of steps required for generating the target output can be determined based on the target input. The number of steps required can be determined based on the behavioral data of the input object. In one example, the input object wants the large model to help recommend an Asian hot pot. The general process is for the large model to determine the user's requirements based on the target input, then gather the relevant data, and then select the hot pot restaurant that meets the input object's requirements based on the gathered relevant data to recommend to the input object.According to the present disclosure, if the input object is interested in a particular firepot based on its behavior prior to the query to the large model, for example, if the input object operates certain software and views a particular firepot in the software several times, or if the chat history characterizes that the input object is interested in a particular firepot, the large model can directly recommend the firepot that was viewed several times or showed interest in the chat history to the input object without collecting any associated data.If the input object has not used any software before the question to the large model, or if the chat history shows no interest in a particular fire pot, the user requirement can be determined according to the target input, and then the associated data is collected, and then the fire pot that meets the requirements of the input object is selected according to the collected associated data and recommended to the input object.

[0151] In some embodiments, after determining the target model based on the initial information, an information search depth is determined in a predefined information source or a knowledge base selected by the input object, based on the target input. For example, if the initial information indicates that the input object possesses a higher level of knowledge in the area to which the target input belongs, the information search depth in a predefined information source or a knowledge base selected by the input object can be increased; otherwise, the information search depth can be reduced.

[0152] In some embodiments, after determining the target model based on the initial information, the information search breadth is determined in a preset information source or a knowledge base selected by the input object, based on the target input. For example, if the initial information indicates that the understanding of the input object encompasses several areas, the information search breadth in the knowledge base selected by the preset information source or in a knowledge base selected by the target input can be increased; otherwise, the information search breadth can be reduced.

[0153] According to the data processing method provided in this disclosure, upon receipt of the target input, if the target input fulfills the first condition, the initial information is obtained. Based on this initial information, the target policy is determined, and the target policy is a policy that characterizes the generation of the target output. According to this disclosure, different policies for generating the target output are determined based on the different initial information, thus changing the manner in which the target output is displayed. In this way, the input object obtains the personalized output associated with the initial information, the interaction mode with the input object is enhanced, and the input object's requirement for intelligent interaction is fulfilled.

[0154] The embodiments of the present disclosure also provide an electronic device. The in Fig. The electronic device shown in Figure 9 is only an example and is not intended to limit the functions and scope of use of the embodiments of the present disclosure.

[0155] As in Fig. As shown in Figure 9, the electronic device includes a processing unit (e.g., a central processing unit, a graphics processing unit, etc.) 901, which can perform various suitable actions and processes according to programs stored in a read-only memory (ROM) 902 or programs loaded by the storage device 908 into a random-access memory (RAM) 903. When the electronic device is switched on, various programs and data required for its operation are also stored in the RAM 903. The processing unit 901, the ROM 902, and the RAM 903 are interconnected via a bus 904. An input / output port (I / O port) 905 is also connected to the bus 904.

[0156] In general, the following devices can be connected to the I / O port 905: an input device 906, including touchscreen, touchpad, keyboard, mouse, camera, microphone, accelerometer, gyroscope, etc.; an output device 907, including liquid crystal display (LCD), speaker, vibrator, etc.; a storage device 908, including memory cards, hard disks, etc.; and communication devices 909. The communication device 909 can enable the electronic device to communicate wirelessly or via wired connections with other devices to exchange data. While it shows Fig. 9. An electronic device with various features; however, not all of the features shown need to be implemented or provided. More or fewer features may be implemented or provided.

[0157] The processing unit 901 contained in the electronic device is used to output the second piece of information, and the second piece of information is used to specify an initial time for obtaining a target output. The processing unit 901 is further used to provide an environment for executing the target model so that the target model generates the target output based on the target input. In some embodiments, when outputting the second piece of information, the processing unit 901 is used to obtain a target input, obtain the first piece of information if the target input satisfies the first condition, and output the second piece of information based on the first piece of information.

[0158] And / or the processing device 901 contained in the electronic device is used to obtain a target input, obtain first information if the target input satisfies a first condition, wherein the first information is related to an input object, output the fourth information based on the first information, wherein the fourth information comprises a first output or a second output and the first output or the second output is generated based on the target input by a target model and the first output is different from the second output.

[0159] And / or the processing device 901 contained in the electronic device is used to obtain a target input, obtain initial information if the target input satisfies an initial condition, and determine a target policy based on the initial information, and the target policy characterizes a policy for producing a target output, which may include the following.

[0160] A model that matches the initial information is selected as the target model from several models based on the initial information, and a target output is generated based on the target model.

[0161] The number of calls to the target model is determined based on the initial information.

[0162] The steps for generating the target output by the target model based on the target input and target information are determined based on the initial information.

[0163] The information search depth in a preset information source or the selected knowledge base is determined based on the initial information obtained from the target input.

[0164] The information search range in the preset information source or the selected knowledge base is determined based on the target input and the initial information.

[0165] Embodiments of the present disclosure also provide a computer program product comprising computer-readable instructions. When the computer-readable instructions are executed on an electronic device, the electronic device can implement one of the data processing methods provided in the embodiments of the present disclosure.

[0166] The embodiments of the present disclosure also provide a computer-readable storage medium on which one or more computer programs are stored, and when the one or more computer programs are executed by the electronic device, the electronic device can implement any of the data processing methods provided by the embodiments of the present disclosure.

[0167] The devices described above are for illustrative purposes only, and the units described as separate may or may not be physically separate, and units represented as units may or may not be physical units, may be located in a single location, or may be distributed across multiple network units. Some or all of the modules may be selected according to actual needs to achieve the solution of this example. Furthermore, in the drawings of the embodiments of the device provided by this disclosure, the connection between the modules represents that the modules have a communication link, and this may specifically be implemented as one or more communication buses or signal lines.

[0168] In the embodiments of the present disclosure, the claims, the different embodiments and the features can be combined to solve the aforementioned technical problems.

[0169] From the above description of the embodiments, it will be apparent to the person skilled in the art that the present disclosure can be implemented by devices with software plus the necessary general-purpose hardware, or of course also by devices with special-purpose hardware including application-specific integrated circuits, special-purpose CPUs, special-purpose memory, special-purpose components, etc. In general, functions performed by computer programs can be easily implemented by appropriate hardware, and specific hardware structures for implementing the same functions can vary, for example, analog circuits, digital circuits, or dedicated circuits. However, for the present disclosure, a software program implementation is a preferred embodiment in many other cases.Based on such an understanding, the technical solution of the present disclosure may be embodied, substantially or in a prior art part, in the form of a software product stored on a readable storage medium such as a floppy disk, a USB storage device, a removable hard disk, a ROM, a RAM, a magnetic disk or an optical disk of a computer, etc., and comprising several instructions for causing a computer device (which may be a personal computer, a training device or a network device, etc.) to execute the method according to the embodiments of the present disclosure.

[0170] In the embodiments described above, the implementation can be wholly or partially achieved through software, hardware, firmware, or any combination thereof. If implemented in software, it can be wholly or partially in the form of a computer program product. The person skilled in the art can implement the described functionality in different ways for each specific implementation; however, such implementation decisions should not be interpreted as resulting in a departure from the scope of this disclosure.

[0171] The computer program product comprises one or more computer instructions. When loaded onto a computer and executed, it fully or partially establishes a process or function according to embodiments of the present disclosure. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or another programmable device. The computer instructions may be stored on a computer-readable storage medium or transferred from one computer-readable storage medium to another; for example, the computer instructions may be transmitted by wired connection (e.g., via coaxial cable, optical fiber, DSL (Digital Subscriber Line)) or wireless connection (e.g., via infrared, radio, microwave, etc.) from one website, computer, training facility, or data center to another website, computer, training facility, or data center.to another data center. The computer-readable storage medium can be any available medium that can be stored by a computer or a data storage facility such as a training facility, a data center, or the like, which incorporates one or more available media. The usable medium can be a magnetic medium (e.g., floppy disk, hard disk, tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., solid-state storage (SSD)), etc.

[0172] In the present description, each embodiment is described in a progressive manner, and each embodiment is mainly described with respect to one point that differs from other embodiments, and identical and similar parts of embodiments are all sufficient to refer to each other.

[0173] The foregoing description of the disclosed embodiments is provided to enable the person skilled in the art to manufacture or use the present disclosure. Various modifications of these embodiments are obvious to the person skilled in the art, and the generic principles defined herein can be applied to other embodiments without departing from the spirit or scope of the present disclosure. Thus, the present disclosure is not intended to be limited to the embodiments shown herein, but is to be given the broadest scope compatible with the principles and novel features disclosed herein. QUOTES INCLUDED IN THE DESCRIPTION

[0000] This list of documents cited by the applicant was automatically generated and is included solely for the reader's convenience. The list is not part of the German patent or utility model application. The DPMA accepts no liability for any errors or omissions. Cited patent literature

[0000] CN 202411750356.7

[0001]

Claims

[1] Data processing methods, including: Obtaining a target input; Obtaining initial information in response to the target input fulfilling a condition; and Output of second information based on the first information, where the second information specifies a time to obtain a target output generated by a target model based on the target input, and the time is related to the first information. [2] Method according to claim 1, wherein: The initial information includes: first sub-information that characterizes the processing power of an electronic device executing the target model, second sub-information that characterizes the allocation of available processing resources of the electronic device by a non-target model, and / or third sub-information that characterizes an input object's understanding of a domain to which the target input belongs; and This includes outputting the second piece of information based on the first piece of information, determining the time based on the first sub-information, the second sub-information and / or the third sub-information, and outputting the second piece of information. [3] Method according to claim 1, wherein: the time is a first time; The first information includes sub-information related to an input object and / or input environment; and Outputting the second piece of information based on the first piece of information includes: Determining a second time based on the sub-information, where the second time characterizes a time that the input object is capable of waiting; and Outputting the second piece of information based on the first time and the second time, the second piece of information further requesting the input object to trigger an execution of the target model or to perform another task. [4] Method according to claim 1, where: the condition is a first condition; and The initial information includes sub-information that characterizes an input object's understanding of an area to which the target input belongs; and the procedure further includes: Determining the level of understanding of the input object for the domain based on the sub-information; and Output of third-party information based on time in response to the level of understanding fulfilling a second condition, with the third-party information prompting the selection of a knowledge base. [5] Electronic device comprising: at least one memory in which one or more instructions are stored; and at least one processor configured to execute the one or more instructions to perform the method according to claim 1. [6] Electronic device according to claim 5, wherein: The initial information includes: first sub-information that characterizes the processing power of an electronic device executing the target model, second sub-information that characterizes the allocation of available processing resources of the electronic device by a non-target model, and / or third sub-information that characterizes an input object's understanding of a domain to which the target input belongs; and which at least one processor is further configured to execute one or more instructions to determine the time based on the first sub-information, the second sub-information and / or the third sub-information when outputting the second information based on the first information and to output the second information. [7] Electronic device according to claim 5, wherein: the time is a first time; The first information includes sub-information related to an input object and / or input environment; and which at least one processor is further configured to execute one or more instructions in order to output the second piece of information based on the first piece of information: based on the sub-information, to determine a second time, where the second time characterizes a time that the input object is capable of waiting; and Based on the first time and the second time, the second information is to be output, the second information further requesting the input object to trigger an execution of the target model or to perform another task. [8] Electronic device according to claim 5, wherein: the condition is a first condition; and The initial information includes sub-information that characterizes an input object's understanding of an area to which the target input belongs; which at least one processor is further configured to execute one or more instructions in order to: to determine the input object's level of understanding of the area based on the sub-information; and In response to the level of understanding fulfilling a second condition, third information is provided based on time, with the third information prompting the user to select a knowledge base. [9] Non-volatile computer-readable storage medium on which one or more instructions are stored which, when executed by a processor, cause an electronic device comprising the processor to perform the method according to claim 1. [10] Data processing methods, including: Obtaining a target input; Obtaining initial information in response to the target input fulfilling a condition, where the initial information is related to an input object; and Outputting second information based on the first information, wherein the second information comprises a first output or a second output generated by a target model based on the target input, and the first output is different from the second output. [11] The method of claim 10, wherein the output of the second information comprises: Outputting the first output in response to the initial information characterizing the input object as a first input object; and Outputting the second output in response to the first information characterizing that the input object is a second input object that is different from the first input object. [12] The method of claim 10, wherein the output of the second information comprises: Outputting the first output in response to the initial information indicating that the input object is in an initial scene; and Outputting the second output in response to the initial information indicating that the input object is located in a second scene that is different from the first scene. [13] Method according to claim 10, wherein: the condition is a first condition; and Outputting the second piece of information based on the first piece of information includes: Outputting the first output in response to the initial information indicating that a time the input object is capable of waiting satisfies a second condition; and / or Outputting the first output in response to the initial information characterizing that the input object's level of understanding of a domain to which the target input belongs satisfies a third condition, and that the quality of the first output is higher than the quality of the second output. [14] The method of claim 10, which, when outputting the second information based on the first information, further comprises: Output of third-party and / or fourth-party information; where: The third set of information includes source data and an association display relationship between the source data and an output contained in the second set of information; and The fourth piece of information includes source data with a label that characterizes that the input object has an association with the source data. [15] Electronic device comprising: at least one memory in which one or more instructions are stored; and at least one processor configured to execute the one or more instructions to perform the method according to claim 10. [16] Non-volatile computer-readable storage medium on which one or more instructions are stored which, when executed by a processor, cause an electronic device comprising the processor to perform the method according to claim 10. [17] Data processing methods, including: Obtaining a target input; Obtaining information in response to the target input fulfilling a condition; and Determine, based on the information, a target policy to generate a target output; the target policy includes: Selecting a target model that matches the information from multiple models and generating a target output based on the target model; Determining the number of calls to the target model based on the information; Determine, based on the information, one or more steps by which the target model generates the target output based on the target input and target information; Determine, based on the information, a depth of an information search in a preset information source or a selected knowledge base based on the target input; and / or Determine, based on the information, a breadth of an information search in the preset information source or the selected knowledge base based on the target input. [18] Electronic device comprising: at least one memory in which one or more instructions are stored; and at least one processor configured to execute the one or more instructions to perform the method according to claim 17. [19] Non-volatile computer-readable storage medium on which one or more instructions are stored which, when executed by a processor, cause an electronic device comprising the processor to perform the method according to claim 17.

Images