Interaction content generation method and device, storage medium and electronic equipment

Abstract

The invention provides an interactive content generation method and device, a storage medium and electronic equipment, and the method comprises the steps: obtaining a user inquiry; sending the user inquiry to a cloud; acquiring a data identifier fed back by the cloud; the data identifier is a data identifier corresponding to the user inquiry word and determined by the cloud through a data warehouse model; obtaining first data corresponding to the data identifier from a home terminal device, and generating interaction content for replying to the user inquiry based on the first data; and outputting the interaction content. Through the method, the interaction experience of the user can be improved.

Description

Technical Field

[0001] The technical solution disclosed herein relates to the field of human-computer interaction technology, and in particular to an interactive content generation method, device, storage medium and electronic device. Background Art

[0002] When users use electronic devices such as mobile phones and tablets, they sometimes ask simple questions through the user terminal. After the user terminal receives the user's inquiry, it will extract the keywords, and then search for the corresponding answer based on the keywords to answer the user, completing an interaction. For example, if a user asks "how much battery is left in the phone", the user terminal will search for the phone's battery data based on the keyword "battery" and then feedback to the user.

[0003] In actual use, it is found that this interaction method has the following problems: when the user's query is complex or does not contain valid keywords, the user end cannot give the answer the user wants. For example, when the user asks "How much time did I use my phone in the past week?", the user end may respond "Sorry, I don't understand" or display the answer retrieved from the Internet to the user. This will obviously reduce the user experience. Summary of the invention

[0004] In view of this, the present disclosure provides an interactive content generation method, device, storage medium and electronic device to improve the user's interactive experience.

[0005] According to a first aspect of an embodiment of the present disclosure, there is provided a method for generating interactive content, which is applied to a user terminal, and the method includes:

[0006] Get the user's query;

[0007] Sending the user query to the cloud;

[0008] Acquire the data identifier of the cloud feedback; the data identifier is the data identifier corresponding to the user query determined by the cloud through a data warehouse model;

[0009] Acquire first data corresponding to the data identifier from the local device, and generate interactive content for answering the user's inquiry based on the first data;

[0010] The interactive content is output.

[0011] In combination with any embodiment provided in the present disclosure, the method further includes:

[0012] Acquire the second data fed back by the cloud, where the second data is analysis data generated by the first model pre-trained in the cloud based on the first data periodically reported by the local device, and the first model is further used to: create or update the data warehouse model based on the user query;

[0013] Generating interactive content for answering the user's inquiry based on the first data includes:

[0014] Interactive content is generated based on the first data and the second data.

[0015] In combination with any implementation manner provided in the present disclosure, the method is applied to a user end; the user query language is used to query the usage of the local device;

[0016] The first data includes: real-time data generated during the use of the local device.

[0017] In combination with any embodiment provided in the present disclosure, the user terminal is provided with a pre-trained second model;

[0018] The output is used to respond to the interactive content of the user's query, including:

[0019] The first data and the user query language are input into the second model to obtain the interaction content to be output; the interaction content also includes: usage suggestions for the local device.

[0020] In combination with any embodiment provided in the present disclosure, the method further includes:

[0021] Obtain the data obtained by distilling the first model from the cloud;

[0022] Model parameters of the second model are optimized based on the data.

[0023] In combination with any implementation manner provided in the present disclosure, sending the user query to the cloud includes:

[0024] If no data identifier associated with the user query is found in the local device, the user query is sent to the cloud;

[0025] The method further comprises:

[0026] The user query and the data identifier are associated and stored.

[0027] According to a second aspect of an embodiment of the present disclosure, a method for generating interactive content is provided, the method being applied in a cloud, comprising:

[0028] Get the user query sent by the user end;

[0029] Obtaining a data identifier corresponding to the user query based on a data warehouse model;

[0030] The data identifier is sent to the user end, so that the user end obtains the first data corresponding to the data identifier from the local device, generates interactive content for replying to the user's inquiry based on the first data, and outputs the interactive content.

[0031] In combination with any embodiment provided in the present disclosure, the method further includes:

[0032] A first model is obtained by training based on preset data; the preset data at least includes historical user query words and data relationships between data identifiers corresponding to the historical user query words;

[0033] The first model is controlled to create or update the data warehouse model based on the user query language, wherein the data warehouse model includes an association relationship between the user query language and the data identifier.

[0034] In combination with any implementation manner provided in the present disclosure, the preset data further includes: first data, and analysis data of the first data;

[0035] The method further comprises:

[0036] Acquire first data periodically reported by the user terminal;

[0037] Inputting the first data into the first model to obtain second data obtained by the first model based on analysis of the first data;

[0038] The second data is sent to the user terminal, so that the user terminal generates interactive content based on the first data and the second data.

[0039] According to a third aspect of an embodiment of the present disclosure, there is provided an interactive content generation device, which is applied to a cloud, and the device includes:

[0040] An input acquisition unit, used to acquire a user's query language;

[0041] A sending unit, used for sending the user query to the cloud;

[0042] An identifier acquisition unit, used to acquire a data identifier of the cloud feedback; the data identifier is a data identifier corresponding to the user query determined by the cloud through a data warehouse model;

[0043] a content generating unit, configured to obtain first data corresponding to the data identifier from a local device, and generate interactive content for answering the user's inquiry based on the first data;

[0044] An output unit is used to output the interactive content.

[0045] According to a fourth aspect of an embodiment of the present disclosure, there is provided an interactive content generating device, the device comprising:

[0046] An acquisition unit, used for acquiring a user query sent by a user terminal;

[0047] A query unit, used to obtain a data identifier corresponding to the user query based on a data warehouse model;

[0048] A sending unit is used to send the data identifier to the user end, so that the user end obtains the first data corresponding to the data identifier from the local device, generates interactive content for replying to the user's inquiry based on the first data, and outputs the interactive content.

[0049] According to a fifth aspect of an embodiment of the present disclosure, there is provided an electronic device, including:

[0050] processor;

[0051] a memory for storing processor-executable instructions;

[0052] The processor is configured to implement the steps of the interactive content generation method of any one of the first aspect or the second aspect by running the executable instructions.

[0053] According to a sixth aspect of an embodiment of the present disclosure, a non-temporary computer-readable storage medium is provided, on which a computer program is stored, and when the program is executed by a processor, the steps of the interactive content generation method described in any one of the first aspect or the second aspect are implemented.

[0054] The technical solution provided by the embodiments of the present disclosure may have the following beneficial effects:

[0055] By sending the acquired user query to the cloud, the cloud data warehouse model can analyze the intent of the user query and quickly locate the data identifier corresponding to the user query. Furthermore, since the data identifier points to the first data used to reply to the user query, the user end can quickly find the first data in the local device based on the data identifier, and then generate and output the corresponding interactive content according to the first data.

[0056] Through the above solution, the interaction quality can be improved while ensuring the response speed.

[0057] It is to be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present disclosure. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present disclosure and, together with the description, serve to explain the principles of the present disclosure.

[0059] Figure 1 is an interactive interface display diagram shown in accordance with an exemplary embodiment of the present disclosure;

[0060] Figure 2 is a flow chart of a method for generating interactive content according to an exemplary embodiment of the present disclosure;

[0061] Figure 3 is a structural diagram of generating a data warehouse model according to an exemplary embodiment of the present disclosure;

[0062] Figure 4 is a flowchart of another interactive content generating method according to an exemplary embodiment of the present disclosure;

[0063] Figure 5 is a flowchart of another interactive content generating method according to an exemplary embodiment of the present disclosure;

[0064] Figure 6 is a structural schematic diagram of an interactive content generating device according to an exemplary embodiment of the present disclosure;

[0065] Figure 7 is a structural schematic diagram of another interactive content generating device according to an exemplary embodiment of the present disclosure;

[0066] Figure 8 It is a schematic structural diagram of an electronic device according to an exemplary embodiment of the present disclosure. DETAILED DESCRIPTION

[0067] It should be understood that the term "and / or" used herein refers to and encompasses any or all possible combinations of one or more of the associated listed items.

[0068] It should be understood that although the terms first, second, third, etc. may be used in the present disclosure to describe various information, such information should not be limited to these terms. These terms are only used to distinguish the same type of information from each other. For example, without departing from the scope of the present disclosure, the first information may also be referred to as the second information, and similarly, the second information may also be referred to as the first information. Depending on the context, the word "if" as used herein may be interpreted as "at the time of" or "when" or "in response to determining".

[0069] When using electronic devices such as mobile phones and tablets, if users want to know the usage status of the device (such as lifespan, memory usage, wear and tear, battery performance, etc.), they usually need to query through the user end.

[0070] For example, the user can find the relevant data by himself in the "settings" option of the electronic device.

[0071] This solution is obviously not intuitive enough, and users are likely to not be able to find where the data is stored.

[0072] In order to solve the above problems, the following solutions are provided:

[0073] Provide users with a user end of the intelligent assistant, allowing users to freely ask questions of interest through the intelligent assistant.

[0074] For example, a user asks, "How much battery is left on my phone?" At this point, the smart assistant will search for the corresponding battery data based on the keyword "battery" and then provide feedback to the user. However, in actual applications, it is found that when the user's query is more complex or does not contain valid keywords, the smart assistant cannot give the answer the user wants. For example, when a user asks, "How much time did I use my phone in the past week?", the smart assistant may respond with "Sorry, I don't understand," or display the answer retrieved from the Internet to the user.

[0075] It can be seen that the answers obtained in this way are often not accurate enough.

[0076] Based on this, a second solution is provided: providing users with a user terminal (such as a mobile phone security center) for specifically querying the usage of electronic devices.

[0077] By using the interactive page of the mobile security center (such as Figure 1 A limited number of options are displayed in the user interface, so that when the user selects an option, the user is fed back the answer to the usage of the option. For example, when the user selects the option for querying the battery performance of the mobile phone, the user terminal will run a pre-set program to obtain interactive content that feeds back the specific situation of the battery performance of the mobile phone.

[0078] It can be seen that this solution does not allow users to freely input questions they are interested in. That is, when the interactive page of the mobile security center does not display the options that the user is interested in, the user still needs to search for them in the "settings" option of the electronic device, or get inaccurate answers through the smart assistant in the first solution.

[0079] Therefore, there is still a need to provide a solution that can improve the user interaction experience.

[0080] The following briefly introduces another better interactive content generation solution provided by the present disclosure:

[0081] The user query language input by the user is obtained and sent to the cloud. In this way, the data warehouse model in the cloud can be used to determine the data identifier corresponding to the user query language. After the user end obtains the data identifier fed back by the cloud, the corresponding first data can be found in the local device according to the data identifier, thereby generating interactive content for replying to the user query language according to the first data, and outputting the interactive content.

[0082] In the method embodiment of the present disclosure, the data warehouse model has a multi-layer data structure and can store and manage a large amount of structured data. Therefore, no matter what user query language the user enters, the data identifier for answering the user query language can be obtained faster and more accurately. On this basis, the user end can quickly find the first data stored in the local device according to these data identifiers, so as to answer the user's question more accurately according to the first data, thereby improving the user's interactive experience.

[0083] The following describes the solution of the present disclosure in detail by taking the user end of an electronic device such as a mobile phone as an example:

[0084] Figure 2 is a flowchart of a method for generating interactive content according to an exemplary embodiment of the present disclosure. Figure 2 As shown, the method comprises the following steps:

[0085] Step 101: Obtain user query language.

[0086] The user end obtains the natural language input by the user through the interactive interface or interactive device of the mobile phone and uses it as the user query language. The interactive device includes but is not limited to: camera, recording device, keyboard. The user input method includes but is not limited to: text input, voice input.

[0087] Step 102: Send the user query to the cloud.

[0088] When answering user inquiries, you need to respond in conjunction with relevant data.

[0089] The method embodiment of the present disclosure can be applied to any inquiry scenario. For example, the user's inquiry is: How is the weather today? The user terminal needs to obtain the weather information of the day (for example: cloudy) and then feedback it to the user. The "cloudy" in this scenario is the relevant data required to answer the user's inquiry, and "weather" can be understood as the data identifier required to find the specific data of "cloudy".

[0090] For the convenience of description, the following text mainly introduces the use scenario of user query language to query the usage of the local device. For example, if the user query language is: how much battery is left in the mobile phone, then it is necessary to obtain the value of the battery, such as "60% battery", and then feedback to the user: the current remaining battery is 60%. Here, "60%" is the relevant data required to answer the user's query, and "battery" can be understood as the data identifier required to find this data.

[0091] The user query language input by the user is unstructured, so it is easy to have various problems such as: the description is too complicated, the description is relatively brief, the description content has no focus, or too many elements are involved. Therefore, it is difficult to find relevant data for answering this user query language, or to find data identifiers for quickly locating these relevant data.

[0092] Based on this, the present disclosure adopts a data warehouse model set up in the cloud to determine the data identifier corresponding to the user query. On the one hand, since the data warehouse model has multiple advantages such as multi-layer data structure, efficient index, rich metadata, data integration and consistency, it is possible to find data identifiers faster and more accurately through the data warehouse model. On the other hand, setting the data warehouse model in the cloud ensures that users can obtain data identifiers using this data warehouse model through different user terminals. Therefore, after executing step 101, the user query is sent to the cloud.

[0093] Step 103, obtaining the data identifier of the cloud feedback; the data identifier is a data identifier corresponding to the user query determined by the cloud through a data warehouse model.

[0094] After receiving the user's query, the cloud inputs it into the data warehouse model, obtains the identifier output by the data warehouse model, and sends it to the user end as the data identifier corresponding to the user's query. In this process, if the cloud's data warehouse model does not have the natural language recognition function, the cloud will convert the user's query into the corresponding machine query instruction before inputting it into the data warehouse model.

[0095] Step 104: Acquire first data corresponding to the data identifier from the local device, and generate interactive content for replying to the user's inquiry based on the first data.

[0096] For example, assume that the data identifier is: power. After obtaining the data identifier "power", the user end queries the first data corresponding to "power" from the local device. For example, if the first data is "60%", the interactive content for answering the user query "how much power is left" is generated based on the first data: 60% power remaining.

[0097] It should be noted that when the method embodiment of the present disclosure is applied to the user end, and the user query is used to query the usage of the local device, the first data includes: real-time data generated during the use of the local device (such as the current power, current memory usage, etc.). In this way, the latest and most accurate first data can be obtained based on the real-time data collected from the local device, ensuring the validity of the interactive content.

[0098] Step 105: output the interactive content.

[0099] The interactive content sorted according to step 104 is output.

[0100] The data warehouse model in the embodiment of the disclosed method has a multi-layer data structure and can store and manage a large amount of structured data. On this basis, users can be allowed to input user queries of interest, and by sending the user queries to the cloud, the data warehouse model in the cloud can parse the intent of the user queries and quickly locate the data identifier corresponding to the user queries. Furthermore, since the data identifier points to the first data used to reply to the user query, the user end can quickly find the first data in the local device based on the data identifier, and then generate and output the corresponding interactive content based on the first data. Through the above scheme, the quality of interaction can be improved while ensuring the response speed.

[0101] In an optional embodiment, the method further includes:

[0102] The second data fed back by the cloud is obtained, where the second data is analysis data generated by the first model pre-trained in the cloud based on the first data periodically reported by the local device.

[0103] When a user uses an electronic device, the user terminal will periodically report at least one first data (battery level, loss level, memory data, application usage, health data, etc.) to leave a trace in the cloud, so that the user can recover this data from the cloud when needed.

[0104] For example, the first data reported by the user terminal in the last period includes data one and data two.

[0105] Data 1: The current battery level of the mobile phone is 30%.

[0106] Data 2: Currently, 80% of mobile phone memory is used.

[0107] Then, the second data obtained by analyzing the first data includes:

[0108] The second data obtained by analyzing the first data: the current battery power is low and it is expected to be used for 2 hours.

[0109] The second data obtained from the analysis of data 2: the memory usage is relatively high, and it is expected to reach 95% after 60 days (or on a certain day of a certain month).

[0110] When the cloud receives the user's query "how much battery is left", in addition to sending the data identifier (such as "battery") to the user, the cloud will also send the previous analysis data of data one "the current battery is low and it is expected to be used for 2 hours" to the user. It should be noted here that since the cloud obtains the second data based on the analysis of the first data previously uploaded by the user, this second data is not real-time. In order to further improve the accuracy of the second data, the cloud can modify the data content of the second data appropriately according to the actual situation after the user sends the user query, and then send the modified second data to the user. For example, the second data "expected to be used for 2 hours" is analyzed at 12:00, and the user sends the user query at 12:30, so it can be modified to "expected to be used for 1.5 hours" (or intelligent prediction, by introducing other data to obtain more accurate values).

[0111] The second data here is obtained by analyzing the cloud based on the pre-trained first model.

[0112] The first model is also used to: create or update the data warehouse model based on the user query; and convert the user query into a machine query instruction, and input the machine query instruction into the data warehouse model to obtain a data identifier corresponding to the user query.

[0113] Taking the query scenario of mobile phone health analysis as an example, the first model automatically establishes / updates the data warehouse model and queries the corresponding data identifier based on the data warehouse model:

[0114] 1. Assume that a user has a smartphone and wants to check the health of his device regularly.

[0115] Equipment health may involve multiple dimensions, such as:

[0116] (1) Hardware status:

[0117] Battery health: You can use system tools to detect the battery's usage status, usually measured by the number of charge and discharge cycles and battery capacity. Storage space: The percentage of remaining storage space. Hardware damage: Screen damage, frame damage, etc.

[0118] (2) Software performance:

[0119] System Fluency: Tests the phone's response time and processing speed by running various apps and games. App Status: Checks whether pre-installed and user-installed apps are up to date, damaged, or unable to launch.

[0120] (3) System update status:

[0121] System version: Check whether the phone is running the latest version of the operating system and how updated it is.

[0122] (4) Safety status:

[0123] Virus Scan: Scan for viruses and malware via security software. Privacy Protection: Check your phone's privacy settings, app permissions, etc.

[0124] (5) Usage habits:

[0125] Usage frequency: the frequency and method of daily use of the mobile phone. Maintenance habits: the user's maintenance and care of the mobile phone.

[0126] For example, the mobile quality score (MQS) data may be used to describe the health of the mobile phone. The MQS data includes but is not limited to:

[0127] Mobile phone performance data: such as CPU usage, memory usage, storage space, etc.

[0128] User App usage habits: such as frequently used applications, usage frequency, usage time, etc.

[0129] Battery health: such as battery capacity, number of charges, battery life, etc.

[0130] Phone temperature: such as temperature changes while the device is running.

[0131] Other relevant data: such as network connection quality, screen brightness, etc.

[0132] 2. In response to the above content, users may raise the following query requirements:

[0133] “What’s wrong with my recent phone habits?”.

[0134] "View records of battery health levels below 80% in the past month".

[0135] "Query the relationship between the phone's storage health and system performance."

[0136] “Show me trends on my device health, including changes in battery, storage, and system performance.”

[0137] 3. In this case, the cloud needs to deal with the following issues:

[0138] The first is natural language understanding and intent recognition: the cloud needs to understand the content of the user's query and identify the query's target, conditions, time range, etc. In order to more accurately understand the user's needs, the cloud obtains the first model that can understand natural language by training models such as large language models (LLMS or LLM).

[0139] Next, the first model needs to complete: data indicator and dimension analysis, and table structure design between each layer of the data warehouse. Specifically, the corresponding database structure will be automatically generated based on the query requirements to ensure that the data warehouse can store various health indicators that users are concerned about.

[0140] For example, assume that the following core tables need to be established in the data warehouse model:

[0141] (1) Device table: records the basic information of each device, such as device ID, device model, operating system version, etc.

[0142] (2) Battery Health: records the battery health data of each device, including the ratio of the maximum battery capacity to the current capacity.

[0143] (3) Storage Health: records the usage of device storage, such as remaining storage and percentage of used storage.

[0144] (4) System Performance Table (SystemPerformance): records the system performance indicators of the device, such as CPU usage, memory usage, etc.

[0145] The first model will automatically generate these tables by analyzing the user's query requirements and ensure that the design of the data table can meet the query requirements. Figure 3 To understand:

[0146] The first model needs to automatically create a data warehouse model based on user needs and existing data sources, and extract useful data identifiers from it. After receiving the user's query, the first model generates:

[0147] (1) Data table structure: represents the structure of the original data set, including metadata information such as table name, column name and its type.

[0148] (2) Data table information description: Provide detailed information about the purpose and function of the data table to facilitate subsequent analysis and understanding.

[0149] (3) Data indicator system: including business North Star indicators, Ratings, etc. A series of key performance indicators (KPIs) are defined, which are crucial for evaluating business performance.

[0150] Next, the first model integrates the information from (1) to (3) above and forms a more comprehensive data view through language description generation. Then, the first model further processes this information into a higher-level table structure, the so-called "hierarchical table". Finally, by analyzing these hierarchical tables, we can derive data identifiers for "user query language" suitable for user queries, thereby completing the construction process of the entire data warehouse model.

[0151] Finally, after the data warehouse structure is established, the first model needs to generate appropriate machine query instructions based on the user's query needs. That is, understand the user's arbitrary query language input by the user and convert it into a machine query instruction (such as an SQL statement) that can be used to query the data identifier in the data warehouse model, so as to enhance the interaction between the user and the data warehouse through natural language processing (NLP). As the query requirements change, the first model also needs to be able to automatically adjust the data warehouse model. For example, if the user starts to query the health trends between multiple devices, the model may automatically add a new "device health trend table" or add new fields to store the health analysis results of different devices. In addition, if a query demand is very frequent, the first model can optimize the efficiency of database query execution and reduce query time. For example, the first model can speed up queries by creating indexes, partition tables, etc.

[0152] According to the above content, the first model can understand user queries, automatically infer the structure of the data warehouse model, including entities, attributes and relationships, and can understand the relationship between user queries and the data warehouse model (especially the various data tables in the data warehouse model), and automatically generate data mapping and conversion logic, thereby completing multiple tasks such as automatically establishing / optimizing the data warehouse model, querying data identifiers, and obtaining the second data.

[0153] After the cloud sends the data identifier and the second data to the user end, the user end executes step 104 and step 105 .

[0154] At this time, executing step 104 to generate interactive content for answering the user's inquiry based on the first data includes:

[0155] Interactive content is generated based on the first data and the second data.

[0156] Exemplarily, the interactive content may be: the battery power is 30% remaining and the device is expected to shut down in 2 hours.

[0157] In an optional embodiment, the user terminal is provided with a pre-trained second model.

[0158] Executing step 104 to generate interactive content for answering the user's inquiry based on the first data includes:

[0159] The first data and the user query language are input into the second model to obtain the interaction content to be output; the interaction content also includes: usage suggestions for the local device.

[0160] For example, the second model can also be obtained by training a model such as LLMS, so that the second model can analyze the intention of the user's query language, and then combine the first data to obtain interactive content for feedback to the user. The interactive content can also include usage suggestions for the local device.

[0161] For example, the user's query is: How much battery power is left? The interactive content is: 30% battery power is left (generated based on the first data), and it is recommended to enter the power saving mode (a usage suggestion for the local device).

[0162] For another example, if the first data and the second data are obtained from the cloud, the interaction content may also be:

[0163] The battery is 30% remaining and the device is expected to shut down in 2 hours. It is recommended to enter power saving mode.

[0164] Or, if the battery is 30% remaining and the device is expected to shut down in 2 hours, it is recommended to close the following applications that consume more power: Application A and Application B.

[0165] In an optional embodiment, the method further includes:

[0166] Obtain data obtained by distilling the first model in the cloud; and optimize model parameters of the second model based on the data.

[0167] A large amount of data is stored in the database in the cloud, so the first model can be continuously optimized and updated based on the data in the database, and then the model parameters of the first model can be updated. On this basis, the model parameters of the second model can be optimized by directly obtaining the data obtained by distilling the first model in the cloud, that is, injecting the knowledge of the first model into the second model, making the second model more intelligent. Here, the first model in the cloud can also be understood as a teacher model, and the second model on the user side can be understood as a student model, and the student model on the user side can be updated through the teacher model in the cloud.

[0168] In an optional embodiment, executing step 102 to send the user query to the cloud includes:

[0169] If no data identifier associated with the user query is found in the local device, the user query is sent to the cloud.

[0170] The method further comprises:

[0171] The user query and the data identifier are associated and stored.

[0172] That is, after executing step 101 and acquiring the user query, it is first checked in the local device whether a data identifier corresponding to the user query is stored.

[0173] If the data identifier is found, the first data can be directly obtained by using the data identifier to perform steps 104 and 105. If the corresponding data identifier is not found, steps 102 and 103 need to be performed to obtain the data identifier from the cloud.

[0174] In this way, when the user asks this user query regularly, the user end does not need to query from the cloud every time, but can directly query from the local device and give the corresponding interaction content, thereby improving the interaction speed while ensuring accuracy.

[0175] It should be noted that if the user changes the user query next time, but the second model recognizes that the new user query has the same meaning as the (historical) user query stored in the local device, the data identifier corresponding to the (historical) user query stored in the local device will also be used.

[0176] In an alternative embodiment, Figure 4 FIG. 1 is a flowchart of another interactive content generation method according to an exemplary embodiment of the present disclosure, which is applied to the cloud and includes:

[0177] Step 201: Obtain a user query sent by a user terminal.

[0178] Step 202: obtaining a data identifier corresponding to the user query based on a data warehouse model.

[0179] Step 203, sending the data identifier to the user end, so that the user end obtains the first data corresponding to the data identifier from the local device, generates interactive content for replying to the user's inquiry based on the first data, and outputs the interactive content.

[0180] In an optional embodiment, the interactive content generating method further includes:

[0181] The first model is trained based on preset data; the preset data at least includes historical user queries and data relationships between data identifiers corresponding to the historical user queries. The first model is controlled to create or update the data warehouse model based on the user queries, wherein the data warehouse model includes data relationships between the user queries and the data identifiers.

[0182] Wherein, obtaining the data identifier corresponding to the user query based on the data warehouse model includes:

[0183] The user query is input into the first model to obtain a machine query instruction; the machine query instruction is input into a data warehouse model to obtain a data identifier corresponding to the user query.

[0184] Exemplarily, the cloud can train the first model in the following manner:

[0185] 1. Data collection:

[0186] Preset data: A large amount of preset data is collected from the Internet, books, news or databases. These data at least include historical user queries and data relationships between data identifiers corresponding to these queries.

[0187] For example:

[0188] Historical user queries: records of questions that users have asked in the past when inquiring about device usage, such as "How much battery power is on my phone?", "How much memory is left on my phone?", etc.

[0189] Data identifier: Specific data identifier associated with historical user queries, such as battery power, memory usage, network connection status, etc.

[0190] 2. Data preprocessing:

[0191] Cleaning and standardization: Clean the collected data to remove invalid or erroneous data, and perform standardization to ensure that the data format is consistent.

[0192] Label data: Label historical user queries with corresponding data identifiers and establish a mapping relationship between query terms and data identifiers.

[0193] 3. Model training:

[0194] Initial model: Use high-performance computing resources (such as GPU clusters) to train the initial large language models (LLMs). The initial model has learned the basic structure and characteristics of the language through a large amount of text data.

[0195] Fine-tuning: Fine-tuning the initial model using annotated historical user queries and data identifiers, so that it can understand user queries and convert them into reasonable machine query instructions.

[0196] Self-supervised learning: Through self-supervised learning methods, the model learns the data relationship between historical user queries and data identifiers, such as hierarchical relationships, data table relationships, etc.

[0197] 4. Data relationship learning:

[0198] Mapping relationship: The model learns the mapping relationship between historical user queries and data identifiers. For example, the corresponding data identifier for "What is the battery level of my phone?" is "battery_level".

[0199] Data table relationships: The model learns the relationship between different data identifiers, such as the association between "battery_level" and "charging_status".

[0200] 5. Data warehouse model establishment:

[0201] Automated modeling: Through the learned data relationships, the model can automatically build a data warehouse model. The data warehouse model defines the structure and relationships of the data so that the data can be efficiently queried and managed.

[0202] Hierarchical relationships: The model is able to recognize and establish hierarchical relationships of data, such as "device" -> "battery" -> "battery_level".

[0203] Data table relationships: The model can identify and establish relationships between different data tables, such as the foreign key relationship between the "device_info" table and the "battery_info" table (the association between the device information table and the battery information table).

[0204] In summary, through cloud training, the first model can learn how to understand historical user queries, how to convert them into reasonable machine query instructions, and how to build and manage data warehouse models. In this way, users can more conveniently query and manage the health status of their devices, and cloud services can provide more efficient and accurate services.

[0205] It should be noted that after the training is completed, the second model on the user side will continue to collect user queries that the user is interested in. If it is found that a certain type of user query is asked more frequently, it will also be uploaded to the cloud as preset data, so that the cloud can further optimize the model parameters of the first model based on these updated preset data.

[0206] In an optional embodiment, the preset data further includes: first data, and analysis data of the first data. The method further includes:

[0207] Acquire first data periodically reported by the user terminal.

[0208] The first data is input into the first model to obtain second data obtained by the first model based on analysis of the first data.

[0209] The second data is sent to the user terminal, so that the user terminal generates interactive content based on the first data and the second data.

[0210] Among them, the contents involved in the above steps have been explained in the above method embodiments and will not be repeated here.

[0211] Exemplary, combined Figure 5 A more preferred embodiment provided by the present disclosure is introduced, such as Figure 5 As shown, the method includes:

[0212] Step 501: Obtain user query language.

[0213] Step 502: Access the cache area of ​​the local device.

[0214] Step 503, check whether the data identifier associated with the user query is stored. If not, execute steps 504-507. If yes, execute step 508.

[0215] Step 504: Send the user's query to the cloud.

[0216] Step 505: Establish / optimize the data warehouse model in the cloud.

[0217] Step 506: The cloud obtains the data identifier from the data warehouse model and stores it in the cloud (here it can be stored in a user-specific storage area to protect the user's privacy).

[0218] Step 507: The cloud sends the data identifier to be stored in the user terminal device (the local device of the user terminal), and returns to step 502 to associate the data identifier with the user query and store it in the cache area of ​​the local device.

[0219] Step 508: The user terminal obtains the data identifier.

[0220] Step 509: The user terminal finds the first data according to the data identifier.

[0221] Step 510: Generate and output interactive content.

[0222] The contents involved in the above steps 501-510 have been explained in the above method embodiment and will not be repeated here.

[0223] For the aforementioned method embodiments, for the sake of simplicity of description, they are all expressed as a series of action combinations, but those skilled in the art should know that the present disclosure is not limited by the described order of actions, because according to the present disclosure, certain steps can be performed in other orders or simultaneously.

[0224] Corresponding to the aforementioned application function realization method embodiment, the present disclosure also provides an embodiment of an interactive content generation device and a corresponding terminal.

[0225] Optional, Figure 6 is a structural diagram of an interactive content generating device according to an exemplary embodiment of the present disclosure; Figure 6 As shown, the interactive content generating device may include:

[0226] Input acquisition unit 601, used to obtain the user query language.

[0227] The sending unit 602 is used to send the user query to the cloud.

[0228] The identifier acquisition unit 603 is used to acquire the data identifier of the cloud feedback; the data identifier is the data identifier corresponding to the user query determined by the cloud through the data warehouse model.

[0229] The content generating unit 604 is configured to obtain the first data corresponding to the data identifier from the local device, and generate interactive content for answering the user's query based on the first data;

[0230] The output unit 605 is used to output the interactive content.

[0231] Optionally, the device further comprises:

[0232] A data acquisition unit is used to acquire second data fed back by the cloud, where the second data is analysis data generated by a first model pre-trained in the cloud based on first data periodically reported by the local device.

[0233] The first model is also used to: create or update the data warehouse model based on the user query; and convert the user query into a machine query instruction, and input the machine query instruction into the data warehouse model to obtain a data identifier corresponding to the user query.

[0234] The content generating unit 604 is used to generate interactive content for answering the user's query based on the first data, including:

[0235] Interactive content is generated based on the first data and the second data.

[0236] Optionally, the device is applied to a user end; the user inquiry language is used to query the usage of the local device; and the first data includes: real-time data generated during the use of the local device.

[0237] Optionally, the user terminal is provided with a pre-trained second model.

[0238] The content generating unit 604 is used to generate interactive content for answering the user's query based on the first data, including:

[0239] The first data and the user query language are input into the second model to obtain the interaction content to be output; the interaction content also includes: usage suggestions for the local device.

[0240] Optionally, the device further comprises:

[0241] The model optimization unit is used to obtain data obtained by distilling the first model in the cloud; and optimize the model parameters of the second model based on the data.

[0242] Optionally, the sending unit 602 is used to send the user query to the cloud, including:

[0243] If no data identifier associated with the user query is found in the local device, the user query is sent to the cloud.

[0244] The device also includes:

[0245] A storage unit is used to store the user query language and the data identifier in association with each other.

[0246] Optional, Figure 7 is a schematic diagram of the structure of another interactive content generating device according to an exemplary embodiment of the present disclosure, such as Figure 7 As shown, the interactive content generating device may include:

[0247] The acquisition unit 701 is used to acquire the user query language sent by the user terminal.

[0248] The query unit 702 is used to obtain a data identifier corresponding to the user query based on a data warehouse model.

[0249] The sending unit 703 is used to send the data identifier to the user end, so that the user end obtains the first data corresponding to the data identifier from the local device, generates interactive content for replying to the user's inquiry based on the first data, and outputs the interactive content.

[0250] Optionally, the device further comprises:

[0251] A training unit is used to obtain a first model through training based on preset data; the preset data at least includes historical user query words and data relationships between data identifiers corresponding to the historical user query words.

[0252] A modeling unit is used to control the first model to create or update the data warehouse model based on the user query language, and the data warehouse model includes a data relationship between the user query language and the data identifier.

[0253] Optionally, the query unit 702 is used to obtain a data identifier corresponding to the user query based on a data warehouse model, including:

[0254] The user query is input into the first model to obtain a machine query instruction; the machine query instruction is input into a data warehouse model to obtain a data identifier corresponding to the user query.

[0255] Optionally, the preset data also includes: first data, and analysis data of the first data.

[0256] The device also includes:

[0257] A prediction unit is used to obtain first data periodically reported by the user terminal; input the first data into the first model to obtain second data obtained by the first model based on the analysis of the first data; and send the second data to the user terminal so that the user terminal generates interactive content based on the first data and the second data.

[0258] As for the device embodiment, since it basically corresponds to the method embodiment, the relevant parts may refer to the partial description of the method embodiment.

[0259] Accordingly, an embodiment of the present disclosure provides an electronic device, comprising: a processor; and a memory for storing processor executable instructions; wherein the processor is configured to implement the steps of any of the above-mentioned interactive content generation methods by running the executable instructions.

[0260] Figure 88 is a schematic diagram of the structure of an electronic device according to an exemplary embodiment of the present disclosure. For example, the electronic device 800 may be a user device, which may specifically be a mobile phone, a computer, a digital broadcast terminal, a message transceiver device, a game console, a tablet device, a medical device, a fitness device, a personal digital assistant, a wearable device such as a smart watch, a smart glasses, a smart bracelet, a smart running shoe, etc.

[0261] Reference Figure 8 , the electronic device 800 may include one or more of the following components: a processing component 802 , a memory 804 , a power component 806 , a multimedia component 808 , an audio component 810 , an input / output (I / O) interface 812 , a sensor component 814 , and a communication component 816 .

[0262] The processing component 802 generally controls the overall operation of the electronic device 800, such as operations associated with display, phone calls, data communications, camera operations, and recording operations. The processing component 802 may include one or more processors 820 to execute instructions to complete all or part of the steps of the above-mentioned method. In addition, the processing component 802 may include one or more modules to facilitate the interaction between the processing component 802 and other components. For example, the processing component 802 may include a multimedia module to facilitate the interaction between the multimedia component 808 and the processing component 802.

[0263] The memory 804 is configured to store various types of data to support operations on the device 800. Examples of such data include instructions for any application or method operating on the electronic device 800, contact data, phone book data, messages, pictures, videos, etc. The memory 804 can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk or optical disk.

[0264] The power supply component 806 provides power to the various components of the electronic device 800. The power supply component 806 may include a power management system, one or more power supplies, and other components associated with generating, managing, and distributing power to the electronic device 800.

[0265] The multimedia component 808 includes a screen that provides an output interface between the electronic device 800 and the user. In some embodiments, the screen may include a liquid crystal display (LCD) and a touch panel (TP). If the screen includes a touch panel, the screen may be implemented as a touch screen to receive input signals from the user. The touch panel includes one or more touch sensors to sense touch, slide, and gestures on the touch panel. The touch sensor may not only sense the boundaries of the touch or slide action, but also detect the duration and pressure associated with the touch or slide operation. In some embodiments, the multimedia component 808 includes a front camera and / or a rear camera. When the electronic device 800 is in an operating mode, such as a shooting mode or a video mode, the front camera and / or the rear camera may receive external multimedia data. Each front camera and rear camera may be a fixed optical lens system or have a focal length and optical zoom capability.

[0266] The audio component 810 is configured to output and / or input audio signals. For example, the audio component 810 includes a microphone (MIC), and when the electronic device 800 is in an operating mode, such as a call mode, a recording mode, and a voice recognition mode, the microphone is configured to receive an external audio signal. The received audio signal can be further stored in the memory 804 or sent via the communication component 816. In some embodiments, the audio component 810 also includes a speaker for outputting audio signals.

[0267] I / O interface 812 provides an interface between processing component 802 and peripheral interface modules, such as keyboards, click wheels, buttons, etc. These buttons may include but are not limited to: a home button, a volume button, a start button, and a lock button.

[0268] The sensor assembly 814 includes one or more sensors for providing various aspects of status assessment for the electronic device 800. For example, the sensor assembly 814 can detect the open / closed state of the electronic device 800, the relative positioning of the components, such as the display and keypad of the electronic device 800, and the sensor assembly 814 can also detect the position change of the electronic device 800 or a component of the electronic device 800, the presence or absence of contact between the user and the electronic device 800, the orientation or acceleration / deceleration of the electronic device 800, and the temperature change of the electronic device 800. The sensor assembly 814 may include a proximity sensor configured to detect the presence of nearby objects without any physical contact. The sensor assembly 814 may also include an optical sensor, such as a CMOS or CCD image sensor, for use in imaging applications. In some embodiments, the sensor assembly 814 may also include an acceleration sensor, a gyroscope sensor, a magnetic sensor, a pressure sensor, or a temperature sensor.

[0269] The communication component 816 is configured to facilitate wired or wireless communication between the electronic device 800 and other devices. The electronic device 800 can access a wireless network based on a communication standard, such as WiFi, 4G or 5G, 4G LTE, 5G NR or a combination thereof. In an exemplary embodiment, the communication component 816 receives a broadcast signal or broadcast-related information from an external broadcast management system via a broadcast channel. In an exemplary embodiment, the communication component 816 also includes a near field communication (NFC) module to facilitate short-range communication. For example, the NFC module can be implemented based on radio frequency identification (RFID) technology, infrared data association (IrDA) technology, ultra-wideband (UWB) technology, Bluetooth (BT) technology and other technologies.

[0270] In an exemplary embodiment, the electronic device 800 may be implemented by one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors, or other electronic components to perform the above methods.

[0271] In an exemplary embodiment, a non-temporary computer-readable storage medium is also provided, such as a memory 804 including instructions, which, when the instructions in the storage medium are executed by a processor 820 of an electronic device 800, enables the electronic device 800 to execute the method in any of the above embodiments.

[0272] The non-transitory computer-readable storage medium may be a ROM, a random access memory (RAM), a CD-ROM, a magnetic tape, a floppy disk, an optical data storage device, and the like.

Claims

1. A method for generating interactive content, characterized in that: include: Get the user's query; Sending the user query to the cloud; Obtaining a data identifier of the cloud feedback; The data identifier is a data identifier corresponding to the user query determined by the cloud through a data warehouse model; Acquire first data corresponding to the data identifier from the local device, and generate interactive content for answering the user's inquiry based on the first data; The interactive content is output.

2. The method according to claim 1, characterized in that The method further comprises: Acquire second data fed back by the cloud, where the second data is analysis data generated by the first model pre-trained in the cloud based on the first data periodically reported by the local device; The first model is also used to: create or update the data warehouse model based on the user query; and convert the user query into a machine query instruction, and input the machine query instruction into the data warehouse model to obtain a data identifier corresponding to the user query; Generating interactive content for answering the user's inquiry based on the first data includes: Interactive content is generated based on the first data and the second data.

3. The method according to claim 1, characterized in that The method is applied to the user end; the user query language is used to query the usage of the local device; The first data includes: real-time data generated during the use of the local device.

4. The method according to claim 3, characterized in that The user terminal is provided with a pre-trained second model; The generating, based on the first data, interactive content for answering the user's inquiry language comprises: The first data and the user query language are input into the second model to obtain the interaction content to be output; the interaction content also includes: usage suggestions for the local device.

5. The method according to claim 4, further comprising: Obtain the data obtained by distilling the first model from the cloud; Model parameters of the second model are optimized based on the data.

6. The method according to claim 1, characterized in that Sending the user query to the cloud includes: If no data identifier associated with the user query is found in the local device, the user query is sent to the cloud; The method further comprises: The user query and the data identifier are associated and stored.

7. A method for generating interactive content, characterized in that: include: Get the user query sent by the user end; Obtaining a data identifier corresponding to the user query based on a data warehouse model; The data identifier is sent to the user end, so that the user end obtains the first data corresponding to the data identifier from the local device, generates interactive content for replying to the user's inquiry based on the first data, and outputs the interactive content.

8. The method according to claim 7, characterized in that The method further comprises: A first model is obtained by training based on preset data; the preset data at least includes historical user query words and a data relationship between data identifiers corresponding to the historical user query words; The first model is controlled to create or update the data warehouse model based on the user query, wherein the data warehouse model includes a data relationship between the user query and the data identifier.

9. The method according to claim 8, characterized in that Obtaining a data identifier corresponding to the user query based on a data warehouse model includes: Inputting the user query into the first model to obtain a machine query instruction; The machine query instruction is input into a data warehouse model to obtain a data identifier corresponding to the user query.

10. The method according to claim 8, characterized in that The preset data also includes: first data, and analysis data of the first data; The method further comprises: Acquire first data periodically reported by the user terminal; Inputting the first data into the first model to obtain second data obtained by the first model based on analysis of the first data; The second data is sent to the user terminal, so that the user terminal generates interactive content based on the first data and the second data.

11. An interactive content generating device, characterized in that: The device comprises: An input acquisition unit, used to acquire a user's query language; A sending unit, used for sending the user query to the cloud; An identifier acquisition unit, used to acquire a data identifier of the cloud feedback; the data identifier is a data identifier corresponding to the user query determined by the cloud through a data warehouse model; a content generating unit, configured to obtain first data corresponding to the data identifier from a local device, and generate interactive content for answering the user's inquiry based on the first data; An output unit is used to output the interactive content.

12. An interactive content generating device, characterized in that: The device comprises: An acquisition unit, used for acquiring a user query sent by a user terminal; A query unit, used to obtain a data identifier corresponding to the user query based on a data warehouse model; A sending unit is used to send the data identifier to the user end, so that the user end obtains the first data corresponding to the data identifier from the local device, generates interactive content for replying to the user's inquiry based on the first data, and outputs the interactive content.

13. An electronic device, characterized in that: include: processor; a memory for storing processor-executable instructions; The processor is configured to implement the steps of any one of the methods of claims 1 to 10 by running the executable instructions.

14. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the steps of any method described in claims 1 to 10 are implemented.

Images