Data processing method and apparatus, electronic device and storage medium

Abstract

Embodiments of the present disclosure provide a data processing method and apparatus, an electronic device and a storage medium. The data processing method comprises: displaying table data, the table data comprising at least one data object and at least one piece of basic statistical information corresponding to the data object; in response to a trigger operation for a preset control, presenting a dialogue component window on a page of the table data; and in response to receiving, by means of the dialogue component window, a first message comprising a target demand, displaying a reply message in the dialogue component window, the reply message comprising conclusion content and / or a recommended demand template for the target demand, wherein the conclusion content comprises at least one of a text, a table and a chart. Once the table data is displayed, the table data is analyzed on the basis of the target demand and a language model to obtain the reply message and display same, thereby realizing rapid processing and analysis of the table data.

Description

Data processing method, device, electronic device and storage medium

[0001] This application claims priority to Chinese Patent Application No. 202311560886.0 filed on November 21, 2023, and the contents of the above-mentioned Chinese patent application disclosure are hereby cited in their entirety as a part of this application. Technical Field

[0002] The embodiments of the present disclosure relate to a data processing method, device, electronic device, and storage medium. Background Art

[0003] A spreadsheet is a commonly used data carrier that records information in the form of rows and columns. The resulting tabular data is widely used in various scenarios such as data management and data analysis.

[0004] Currently, the processing and analysis of tabular data usually requires users to write dedicated analysis programs, which leads to low processing efficiency and poor accuracy of analysis results. Summary of the Invention

[0005] The embodiments of the present disclosure provide a data processing method, device, electronic device, and storage medium to overcome the problems of low efficiency in table data processing and poor accuracy of analysis results.

[0006] In a first aspect, an embodiment of the present disclosure provides a data processing method, including:

[0007] Displaying tabular data, wherein the tabular data includes at least one data object and at least one basic statistical information corresponding to the data object; presenting a dialog component window on the page of the tabular data in response to a triggering operation on a preset control; displaying a reply message in the dialog component window in response to receiving a first message containing a target requirement through the dialog component window, wherein the reply message includes conclusion content and / or a recommended requirement template for the target requirement; wherein the conclusion content includes at least one of text, a table and a chart.

[0008] In a second aspect, an embodiment of the present disclosure provides a data processing device, including:

[0009] a display unit, configured to display tabular data, wherein the tabular data includes at least one data object and at least one basic statistical information corresponding to the data object;

[0010] A receiving unit, configured to present a dialog component window on the page of the table data in response to a triggering operation on a preset control;

[0011] A processing unit is used to display a reply message in the dialogue component window in response to receiving a first message containing a target requirement through the dialogue component window, wherein the reply message includes conclusion content and / or a recommended requirement template for the target requirement; wherein the conclusion content includes at least one of text, table and chart.

[0012] In a third aspect, an embodiment of the present disclosure provides an electronic device, including: a processor and a memory;

[0013] The memory stores computer-executable instructions;

[0014] The processor executes the computer-executable instructions stored in the memory, so that the at least one processor executes the data processing method described in the first aspect and various possible designs of the first aspect.

[0015] In a fourth aspect, an embodiment of the present disclosure provides a computer-readable storage medium, in which computer execution instructions are stored. When a processor executes the computer execution instructions, the data processing method described in the first aspect and various possible designs of the first aspect is implemented.

[0016] In a fifth aspect, an embodiment of the present disclosure provides a computer program product, including a computer program, which, when executed by a processor, implements the data processing method described in the first aspect and various possible designs of the first aspect. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solutions in the embodiments of the present disclosure, a brief introduction will be given below to the drawings required for use in the description of the embodiments. Obviously, the drawings described below are some embodiments of the present disclosure. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0018] FIG1 is a diagram of an application scenario of the data processing method provided by an embodiment of the present disclosure;

[0019] FIG2 is a flowchart of a data processing method according to an embodiment of the present disclosure;

[0020] FIG3 is a schematic diagram of a process of a first terminal device responding to a triggering operation according to an embodiment of the present disclosure;

[0021] FIG4 is a schematic diagram of displaying alternative requirement texts provided by an embodiment of the present disclosure;

[0022] FIG5 is a schematic diagram of displaying second prompt information provided by an embodiment of the present disclosure;

[0023] FIG6 is a schematic diagram of a process for displaying extended text according to an embodiment of the present disclosure;

[0024] FIG7 is a schematic diagram of displaying a reply message and chart information provided by an embodiment of the present disclosure;

[0025] FIG8 is a second flow chart of a data processing method according to an embodiment of the present disclosure;

[0026] FIG9 is a flowchart of a specific implementation of step S204 in the embodiment shown in FIG8 ;

[0027] FIG10 is a schematic diagram of a system structure for implementing the method of the present embodiment provided by the present disclosure;

[0028] FIG11 is a flowchart of a specific implementation of step S205 in the embodiment shown in FIG8 ;

[0029] FIG12 is a flowchart of a specific implementation of step S2051 in the embodiment shown in FIG11 ;

[0030] FIG13 is a schematic structural diagram of a first intelligent agent provided by an embodiment of the present disclosure;

[0031] FIG14 is a flowchart of a specific implementation method of step S2052 in the embodiment shown in FIG11 ;

[0032] FIG15 is a structural block diagram of a data processing device provided by an embodiment of the present disclosure;

[0033] FIG16 is a schematic structural diagram of an electronic device provided by an embodiment of the present disclosure; and

[0034] FIG17 is a schematic diagram of the hardware structure of an electronic device provided in an embodiment of the present disclosure. DETAILED DESCRIPTION

[0035] To make the objectives, technical solutions, and advantages of the embodiments of the present disclosure more clear, the technical solutions in the embodiments of the present disclosure will be clearly and completely described below in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present disclosure, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present disclosure without making any creative efforts shall fall within the scope of protection of the present disclosure.

[0036] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this disclosure are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with the relevant laws, regulations and standards of relevant countries and regions, and provide corresponding operation entrances for users to choose to authorize or refuse.

[0037] The following explains the application scenarios of the embodiments of the present disclosure:

[0038] Figure 1 illustrates an application scenario for the data processing method provided by an embodiment of the present disclosure. The data processing method provided by an embodiment of the present disclosure can be applied to applications with table data processing capabilities, and more specifically, to application scenarios based on online spreadsheet data processing. The execution entity of this embodiment can be a terminal device running the aforementioned application with online spreadsheet functionality and program, a server hosting the server corresponding to the aforementioned application, or other electronic devices that perform similar functions. Taking a terminal device as an example, an application with online spreadsheet functionality is running within the terminal device. As shown in FIG1 , the application's interactive interface displays a data table, sheet_1. This data table, sheet_1, for example, has N data rows and M data columns (N and M are positive integers), where each data row corresponds to a data object. As shown in the figure, the first column in sheet_1 records the object identifier of the data object. For example, the object identifier of the data object corresponding to the first row is "Material Data #1," and the object identifier of the data object corresponding to the second row is "Material Data #2." Statistical information corresponding to each data object is recorded in the remaining data columns, for example, the first column records the statistical quantity of the data object, the second column records the statistical time of the data object, and so on. After obtaining the above-mentioned table data, the terminal device can further run a pre-programmed data analysis program to process the table data and obtain data processing results, thereby achieving data analysis, information mining, and other data processing purposes.

[0039] The analysis process for tabular data like the one above can be based on user experience, meaning users can write corresponding data processing programs to process and analyze the data. However, data processing and analysis have a high operational threshold, and the accuracy of the analysis results depends primarily on the user's data analysis experience and business knowledge. Furthermore, because data processing programs must be written in advance, the rules for data analysis and reasoning cannot be flexibly and in real time modified. This leads to problems such as low data processing efficiency and poor analysis accuracy during the analysis and processing of tabular data.

[0040] The embodiments of the present disclosure provide a data processing method to solve the above problems.

[0041] Referring to FIG2 , FIG2 is a flow chart of a data processing method according to an embodiment of the present disclosure. The method of this embodiment can be applied in a terminal device, and the data processing method includes:

[0042] Step S101: displaying table data, where the table data includes at least one data object and at least one basic statistical information corresponding to the data object.

[0043] For example, referring to the application scenario diagram shown in FIG1 , the execution subject of the information display method provided by the embodiment of the present disclosure can be a terminal device, such as a personal computer, running an application with a spreadsheet function in the terminal device. The terminal device provides an operation entry for the user by displaying the interactive interface of the application. The user operates the application through the operation entry, can load and open the table data in the application, and display the table data in the interactive interface of the application. Among them, the interactive interface can include a main window and a dialog component window. The main window and the dialog component window can be arranged in a left-right or top-down layout in the interactive interface. The main window is used to display the original table data, and the dialog component window is used to display relevant interactive information for the table data, for example, for providing an editable text box to receive text entered by the user; for another example, for displaying intermediate process information during the interaction process, such as prompt information and reply messages. The content displayed in the dialog component window will be described in detail in the subsequent embodiment steps in conjunction with the embodiment scheme, and will not be repeated here. Furthermore, the main window and the dialog component window can correspond to windows and pages displayed in the interactive interface. For example, a first window for displaying tabular data is provided in the main window position; a second window for displaying related interactive information and reply messages is provided in the conversation component window position. The specific implementation scheme for displaying content in the main window and conversation component windows can be set as needed and is not limited here.

[0044] Furthermore, the application can have an online electronic spreadsheet function, that is, the user can edit the spreadsheet data online and synchronize it across multiple terminals. Therefore, the spreadsheet data can be stored locally on the terminal device and directly loaded locally by the terminal device based on user instructions; it can also be stored in the cloud and obtained by the terminal device accessing the cloud server. The details will not be discussed here.

[0045] Furthermore, referring to the table data in the application scenario diagram shown in FIG1 , the table data includes at least one data object and at least one basic statistical information corresponding to the data object, that is, the table data is a data storage medium for data storage based on the data object. Each data object can correspond to the same amount of basic statistical information. Specifically, for example, taking the application scenario of storing "book information" in the table data as an example, the data object corresponds to the name of the book, such as "XX Autobiography" and "XXX Adventures". The basic statistical information corresponding to each data object includes, for example, the author of the book, the time of publication of the book, the sales volume of the book, etc., that is, in the table data, the author of the book "XX Autobiography" and the time of publication of the book, and the sales volume of the book are included respectively.

[0046] Among them, the way of displaying tabular data in the interactive interface, in one possible implementation method, can be to display data objects by row, that is, each row displays one data object, and corresponding one or more basic statistical information, which is the display method of tabular data shown in Figure 1; in another possible implementation method, can also be to display data objects by column, that is, each column displays one data object, and corresponding one or more basic statistical information. The specific implementation method can be set as needed and is not specifically limited here.

[0047] Step S102: In response to a triggering operation on a preset control, a dialog component window is presented on the page of the table data.

[0048] Step S103: In response to receiving a first message containing a target requirement through the dialogue component window, a reply message is displayed in the dialogue component window, the reply message including conclusion content and / or a recommended requirement template for the target requirement; wherein the conclusion content includes at least one of text, table and chart.

[0049] Exemplarily, an input component for receiving user input text, such as an editable text box (edit_text), is provided within the interactive interface. After the user enters text through the input component, the terminal device obtains the user input content, i.e., the target requirement, through the input component. In one possible implementation, based on the description of the steps in the previous embodiment, the input component can be set within the dialog component window to avoid affecting the display effect of the table data in the main window. Furthermore, before displaying the reply message, the method further includes: generating a reply message based on the target requirement, wherein the reply message is generated based on the language model and includes the target data object obtained based on the inference rule. The target requirement is used to represent the inference rule for inferring the table data based on advanced statistical information, and the advanced statistical information is generated based on the basic statistical information. For example, in one example, the content of the target requirement includes: "Find the ten books with the fastest reading growth rate." In this target requirement, the advanced statistical information is "reading volume growth rate", and the advanced statistical information "reading volume growth rate" is obtained based on the basic statistical information "reading volume". This instructs the terminal device (language model) to reason about the table data based on the high-level statistical information about "reading volume growth rate," thereby deriving the logic for "the ten books with the fastest reading volume growth," or inference rules. These inference rules, as represented by the target requirements, guide the language model through subsequent data processing steps, ultimately generating a response message.

[0050] Furthermore, based on the different contents of the target requirements, the inference rules represented by them have multiple implementation methods. For example, the inference rules representing the target requirements include at least one of the following: determining the target data object based on the average value of the data object's advanced statistical information within the target time interval; determining the target data object based on the change in the data object's advanced statistical information within the target time interval; determining the target data object based on the rising and / or falling values ​​of the data object's advanced statistical information within the target time interval; determining the target data object based on the change trend of the data object's advanced statistical information within the target time interval. Based on the target requirements entered by the user, the tabular data can be personalized analyzed from multiple dimensions, thereby obtaining effective and high-value data analysis results.

[0051] Furthermore, the process of generating a reply message is generated based on a language model, which is a model for task processing based on natural language, such as a large language model (LLM). In this embodiment, the language model can be used to generate corresponding action instructions in combination with target requirements to call an external data analysis module, thereby realizing automated analysis and processing of the above-mentioned table data, thereby generating corresponding reply messages. Furthermore, in the process of generating a reply message based on a language model, prompt words can be used to guide the content generated by the language model. The above-mentioned prompt words are generated based on target requirements and a preset prompt word template. The generation control of the above-mentioned reply message is realized by setting the prompt word template.

[0052] By displaying tabular data, the tabular data includes at least one data object and at least one basic statistical information corresponding to the data object; in response to a triggering operation of a preset control, a dialogue component window is presented on the page of the tabular data; in response to receiving a first message containing a target requirement through the dialogue component window, a reply message is displayed in the dialogue component window, the reply message including conclusion content and / or a recommended requirement template for the target requirement; wherein the conclusion content includes at least one of text, a table, and a chart. After displaying the tabular data, by receiving the target requirement input by the user, and analyzing the tabular data based on the target requirement and a language model, obtaining a reply message and displaying it, rapid processing and analysis of the tabular data is achieved, without the need to run an additional data analysis program for data processing, thereby improving the data processing efficiency and data analysis accuracy of the tabular data.

[0053] In one possible implementation, after displaying the table data (step S101), the terminal device can execute subsequent steps by responding to a trigger operation input by the user. FIG3 is a schematic diagram of the process of the first terminal device responding to a trigger operation according to an embodiment of the present disclosure. As shown in FIG3, the table data is displayed in the main window of the interactive interface of the terminal device. When the user clicks the interactive component in the interactive interface (the name of the interactive component is "smart assistant"), the terminal responds to the trigger operation and displays an editable text box edit_text in the dialogue component window in the interactive interface. At the same time, a preset explanatory message is displayed in the second window, such as "Hello, I am a table data intelligent analysis assistant. Please enter your requirements." After the user enters the target requirements through the editable text box edit_text, the terminal device obtains the target requirements and executes subsequent data processing steps to obtain a reply message corresponding to the table data and displays it in the interactive interface.

[0054] Optionally, in one possible implementation, the first message includes an indicator and a first requirement text, wherein the first requirement text is used to indicate target data in the table data: the specific implementation method of step S103 includes: in response to the indicator received through the dialogue component window, obtaining the target data indicated by the first requirement text, and displaying a reply message in the dialogue component window based on the target data.

[0055] Exemplarily, the first message includes an indicator and a first requirement text, where the indicator can be a preset special symbol, letter, or number, such as the "@" symbol. The first requirement text is used to indicate target data within the tabular data, such as a row number, column number, or a combination of row and column numbers, thereby locating the target data within the tabular data. The target data indicated by the first requirement text is then used as the target requirement or a portion of the target requirement for subsequent language model processing steps, thereby generating and displaying a reply message.

[0056] Optionally, before step S102, the method further includes:

[0057] Step S100A: Displaying at least one candidate requirement text in the dialog component window, where the candidate requirement text represents a requirement template for the target requirement;

[0058] Step S100B: In response to the selection instruction for the candidate requirement text, the target requirement is obtained.

[0059] Correspondingly, step S102 is implemented by obtaining the target demand obtained in step S100B.

[0060] FIG4 is a schematic diagram of displaying alternative requirement texts provided by an embodiment of the present disclosure. As shown in FIG4 , the descriptive information displayed in the second window may further include alternative requirement texts. For example, as shown in the figure, the descriptive information displayed in the second window also includes the following alternative requirement texts:

[0061] "I can help you complete the following tasks: Processing Function I; Processing Function II; Processing Function III."

[0062] Afterwards, in response to the selection instruction for the alternative requirement text, a target requirement is determined. The specific implementation method of the selection instruction includes, for example, clicking on the text "processing function I" or "processing function II" or "processing function III" in the alternative requirement text to generate a corresponding selection instruction. The specific implementation method of the selection instruction also includes, for example, entering "processing function I" or "processing function II" or "processing function III" in an editable text box to generate a corresponding selection instruction. For example, as shown in the figure, when the alternative requirement text "processing function I" is selected, the terminal device determines "processing function I" as the target requirement. Afterwards, in one possible implementation method, the terminal device can directly execute subsequent steps based on the target requirement determined above and generate a reply message using a language model. In another possible implementation method, the terminal device can further interact with the user for the target requirement to further determine the implementation method of the inference rule represented by the target requirement (such as processing function I). Exemplarily, in the second window, at least one implementation text corresponding to the selected target requirement is further displayed, and the implementation text is used to represent the implementation method of the inference rule corresponding to the target requirement. As shown in the figure, the target requirement "Processing Function 1" corresponds to three implementation texts. In the second window, implementation texts 1, 2, and 3 corresponding to "Processing Function 1" are displayed, representing the implementation methods of the three target requirements. Specifically, for example, if the content of Processing Function 1 is "Find the three most valuable books," its corresponding implementation text 1 includes: "Measure the value of books by their sales volume"; implementation text 2 includes: "Measure the value of books by their reading volume"; and implementation text 3 includes: "Measure the value of books by their ratings." The user then selects implementation text 2, identifying it as the target requirement. Based on the target requirement generated by the target requirement, subsequent steps are executed, generating a reply message using the language model. For example, implementation text 2 may contain: "Help me analyze the books that are ranked in the top ten in terms of reading volume but not in the top ten in terms of sales volume." In this embodiment, a multi-level interactive approach is used to further determine more detailed target requirements, thereby improving the accuracy of the inference rules indicated by the user and ensuring that the resulting streaming theoretical data better meets user needs.

[0063] Optionally, before step S102, the method further includes:

[0064] Step S100C: Displaying capability description information of the dialog component window in the dialog component window.

[0065] For example, the capability description information of a dialog component window refers to the data processing capabilities enabled by the dialog component window input, i.e., the data processing capabilities of the language model that implements the dialog component window. Specifically, examples include data classification processing, data insight, data prediction, and data visualization. By displaying the capability description information within the dialog component window, the data processing capabilities of the dialog component window are demonstrated, guiding users to perform correct operations and improving interaction efficiency.

[0066] Optionally, in a possible implementation, after step S103, the method further includes:

[0067] In response to a second message received through the dialog component window, the conclusion content is inserted into the table data to generate modified table data. For example, after generating and displaying a reply message, the user can further input a second message into the dialog component window, causing the terminal device to respond to the second message and insert the generated conclusion content into the initial table data to obtain updated table data, thereby updating the table data. The conclusion content includes one or more of text, tables, and charts, i.e., the text, tables, and charts generated in the dialog component window are inserted into the table data.

[0068] Optionally, in a possible implementation, after step S103, the method further includes:

[0069] Step S104: displaying at least one second prompt information, where the second prompt information is generated based on the reply message, and the second prompt information is used to represent a generation rule of target advanced statistical information in the reply message.

[0070] Exemplarily, after or at the same time as generating a reply message and displaying it in the dialog component window, the terminal device may further display at least one second prompt message. FIG5 is a schematic diagram of displaying the second prompt message provided by an embodiment of the present disclosure. As shown in FIG5 , the conclusion content is displayed in the second window located in the dialog component window in the interactive interface, and its content is, for example: "The three books with the highest book ratings are: 1. "XX Autobiography"; 2. "XX Adventures"; 3. "XX Sports"". Afterwards, upon receiving the instruction text "Please explain the scoring rules for the book ratings in the conclusion content" entered by the user through an input component (e.g., an editable text box, edit_text shown in the figure), the terminal device calls the language model based on the instruction text to generate a second prompt message for "book rating", i.e., the second prompt message. The second prompt message is used to describe how the "book rating" (target advanced statistical information) in the conclusion content is generated based on the primary statistical information, that is, the generation rules for the advanced statistical information in the conclusion content. For example, as shown in the figure, the content of the second prompt message displayed in the second window includes:

[0071] "Book ratings are determined using the following formula:

[0072] Value=0.3*sale+0.3*read+0.4*conm

[0073] Among them, Value is the book rating, sale is the book's sales volume in a week, read is the book's reading volume in a week, and conm is the book's rating in a week.

[0074] Of course, in another possible implementation method, the terminal device can automatically generate the above-mentioned second prompt information without receiving the instruction text entered by the user, and display it or not in the interactive interface. The specific implementation method can be set based on specific needs and will not be repeated here.

[0075] Furthermore, in a possible implementation, the reply message includes target advanced statistical information, and after step S103, the following is further included:

[0076] Step S105: display at least one extended text; in response to a selection instruction for the extended text, obtain an updated target requirement, and generate an updated reply message based on the updated target requirement; display the updated reply message in the dialog component window; wherein, the extended text is generated based on the reply message, and the extended text is used to represent the inference rules for inferring the tabular data based on the target high-level statistical information.

[0077] Exemplarily, after or simultaneously generating a reply message and displaying it in a dialog component window (e.g., a second window), the terminal device may further display at least one extended text. The extended text serves the same purpose as the alternative requirement text in the previous embodiment, providing the user with other possible inference rules, thereby helping the user adjust the target requirement to obtain a more optimized reply message. However, the difference is that the extended text is obtained by processing the reply message obtained in the previous step. Therefore, the extended text can implement an inference rule that represents inference rules for tabular data based on target advanced statistical information. For example, if the reply message includes the target advanced statistical information "weekly average fluctuation value of reading volume," the extended text then represents a corresponding inference rule based on the target advanced statistical information "weekly average fluctuation value of reading volume," such as "monthly average fluctuation value of reading volume" or "change in weekly average fluctuation value of reading volume relative to the previous week." Subsequently, in response to a selection instruction for the extended text, the selected extended text is used as the updated target requirement, and the above data processing steps are repeated to obtain an updated reply message, which is then displayed in the dialog component window. The specific process will not be repeated here.

[0078] FIG6 is a schematic diagram of a process for displaying extended text according to an embodiment of the present disclosure. As shown in FIG6 , the conclusion content is displayed in the dialog component window in the interactive interface. The conclusion content includes the following text content:

[0079] "The three books with the largest weekly average fluctuations in reading volume are: 1. 'The Autobiography of XX'; 2. 'The Adventures of XX'; 3. 'XX Sports'." Then, automatically or by triggering a corresponding control, extended text is displayed in the second window. For example, as shown in the figure, the following extended text is displayed in the second window:

[0080] You could also say:

[0081] Question 1: Which three books have the largest average monthly fluctuation in reading volume?

[0082] Question 2: Which three books have the largest weekly average fluctuation in reading volume compared to the previous week?

[0083] Furthermore, the user enters the text "What are the three books with the largest monthly average fluctuation in reading volume?" in the input component (shown as edit_text in the figure), which is "Question 1" in the above extended text. The terminal device then uses the above extended text "What are the three books with the largest monthly average fluctuation in reading volume?" as the target requirement and repeats the above step S103, that is, regenerating the conclusion content based on the target requirement, thereby updating the reply message in the dialogue component window. For example, as shown in the figure, the updated conclusion content is:

[0084] The three books with the largest weekly average fluctuations in reading volume are: 1. "The Adventures of XXX"; 2. "XX Anime"; and 3. "XX Game."

[0085] Furthermore, in a possible implementation, after step S103, the following steps are further included:

[0086] Step S106: Displaying first prompt information, where the first prompt information is used to indicate form data and / or question data in the reply message.

[0087] Exemplarily, the problem data includes at least one of the following: a data object whose object identifier is missing or exceeds a threshold; and advanced statistical information whose statistical value is missing or exceeds a threshold.

[0088] Exemplarily, the first prompt information is an error prompt information, which is used to prompt the problem data in the table data displayed in the main window, and / or the problem in the reply message. For example, the data object is used to represent the name of the material data. More specifically, in the scenario of book data management, the data object represents the name of the book. After the reply message is generated and displayed, if the target data object or non-target data object involved in the reply message is missing or abnormal, for example, the name of the book is empty, there are unrecognizable garbled characters, etc., then a prompt is given through the first prompt information, or, in the process of generating the reply message, if it is found that a certain basic statistical information is empty, there are unrecognizable garbled characters, or exceeds the threshold, it is marked as problem data and displayed through the first prompt information, thereby realizing automatic detection of the original table data and improving the efficiency of data correction.

[0089] Furthermore, after step S103, the method further includes:

[0090] Step S107: Generate chart information corresponding to the reply message through the language model, and display the chart information in the dialogue component window. The chart information is used to visualize the reply message.

[0091] Furthermore, after receiving the reply message, the terminal device may further generate chart information based on the reply message, and the chart information is used to visually display the reply message. Optionally, a corresponding data report may be further generated based on the chart information. In one possible implementation, the chart information is used to characterize the proportional relationship between the advanced statistical information corresponding to at least two target data objects. Specifically, after receiving the reply message, the terminal device may extract one or more statistical information from the corresponding advanced statistical information or multiple basic statistical information based on the target data object represented by the reply message, generate chart information directly or after calculation, and generate a corresponding picture for display. Furthermore, with respect to the display process of the chart information, after generating the reply message, the terminal device may pop up a third window in the main window or the first window displaying the table data, and display the chart information corresponding to the reply message in the third window, or display the corresponding chart information in the second window corresponding to the dialog component window.

[0092] Figure 7 is a schematic diagram of displaying conclusion content and chart information provided by an embodiment of the present disclosure. As shown in Figure 7, the table data is displayed in the first window of the terminal device, and after responding to the trigger instruction (triggered by clicking the interactive component #1), the table data is processed based on the language model to obtain the conclusion content and chart information, and displayed in the second window. Specifically, with reference to Figure 7, the text content corresponding to the conclusion content includes: "1. The top three books with the highest average sales in the past week are "XX Autobiography", "XXX Adventures", and "X Story"." Accordingly, based on the conclusion content, the terminal device generates chart information. Specifically, the terminal device obtains the average sales of the target objects: "XX Autobiography", "XXX Adventures", and "X Story" in the past week, and the sum of the average sales of all data objects in the past week (that is, the total sales of all books), and obtains the proportion of the average sales of "XX Autobiography", "XXX Adventures", and "X Story" in the past week, that is, the chart information, and then displays the chart information in the form of a picture. For example, as shown in the figure, the proportion of the average sales of "XX Autobiography" in the past week is 20%; the proportion of the average sales of "XXX Adventures" in the past week is 12%; the proportion of the average sales of "X Story" in the past week is 9%, and the sum of the average sales of other data objects in the past week accounts for 59%.

[0093] In this embodiment, by further generating and displaying chart information corresponding to the conclusion content after generating the conclusion content, the information display dimension for the table data is improved, and the information display efficiency and display effect after data analysis are improved.

[0094] It should be noted that the above-mentioned steps S104-S107 executed after step S103, and step S100 (including S100A and S100B) executed before step S102, can be executed separately as needed, for example, step S100 and steps S101-S103 of the above-mentioned embodiment are executed separately, or, steps S101-S103 and step S106 are executed separately. Thereby, the corresponding technical objectives and the corresponding technical effects are achieved. Of course, the above-mentioned multiple steps can also be executed simultaneously, for example, step S100, steps S101-S103, steps S104-S105 are executed in sequence, or, steps S101-S107 are executed, etc., so as to achieve the corresponding technical objectives. When the above-mentioned multiple steps are executed simultaneously, the execution order of steps S104-S107 can be adjusted as needed, and no further examples will be given here.

[0095] Referring to FIG8 , FIG8 is a second flow chart of a data processing method according to an embodiment of the present disclosure. Based on the embodiment shown in FIG2 , this embodiment further refines step S103 , and the data processing method includes:

[0096] Step S201: displaying table data, where the table data includes at least one data object and at least one basic statistical information corresponding to the data object.

[0097] Step S202: In response to a triggering operation on a preset control, a dialog component window is presented on the page of the table data.

[0098] Step S203: In response to receiving a first message containing a target requirement through the dialog component window, a target prompt word is obtained, where the target prompt word is used to guide the language model to generate a reply message based on an inference rule.

[0099] For example, after receiving a target requirement input from the user, the terminal device executes a preset function program or accesses a preset service to obtain a corresponding prompt template, thereby obtaining a target prompt. The target prompt is used to guide the language model to generate a reply message based on inference rules. The target prompt can be recognized by the language model used subsequently and guides the output content of the language model. The conversion process between the target requirement and the target prompt, as well as the specific content and form of the target prompt, are determined by the prompt template. The specific implementation method for obtaining the target prompt based on the prompt template is not specifically limited here.

[0100] Step S204: Process the target prompt word through the preset table data service to generate a data inference request. The table data service is used to obtain table data and generate a data inference request based on the table data and the target prompt word.

[0101] Exemplarily, afterward, by accessing a preset table data service and processing the target prompt word, the table data service is used to respond to the above-mentioned intention of generating a reply message, identify and convert the table data required to generate the reply message, and combine it with the target prompt word to generate a data inference request that can be processed by the language model.

[0102] For example, as shown in FIG9 , the specific implementation of step S204 includes:

[0103] Step S2041: identifying the table data structure of the table data through the table data service;

[0104] Step S2042: Based on the table data structure, load the table data and convert the table data into parsed data that can be used to input the execution plug-in;

[0105] Step S2043: Generate a data inference request based on the parsed data and the target prompt word.

[0106] Exemplarily, the terminal device sends the target prompt word to the table data service in an intention distribution manner. The table data service can be deployed as a server-side of an online spreadsheet function or deployed locally on the terminal device. After receiving the target prompt word, the table data service identifies the table data structure of the table data, the number of rows and columns of the table data, and the position of the data labels in the table data (for example, the first column is the data label), etc., through the table data service, so as to load the table data into the memory and convert it into data with a specific data format that can be processed by the subsequent execution plug-in, that is, parsed data. Afterwards, the parsed data and prompt word information are packaged into a data inference request. Among them, the prompt word information is generated based on the target prompt word, which can be the target prompt word itself, or it can be text, logo or other information generated after processing the target prompt word.

[0107] Step S205: Based on the data inference request, call the execution plug-in implemented based on the language model to obtain a reply message corresponding to the table data. The execution plug-in is used to call the external knowledge base to process the table data to generate a reply message.

[0108] Exemplarily, after a data inference request is generated through a table data service, an execution plug-in is called to process the inference request and generate a reply message. In one possible implementation, the execution plug-in is a functional plug-in implemented based on a language model, i.e., the execution plug-in includes a language model or can implement the language model's functionality by accessing or calling an external model. After the data inference request is input into the execution plug-in, the execution plug-in processes the prompt word information and parsed data included or indicated in the data inference request, leveraging the capabilities of the language model to generate a reply message.

[0109] FIG10 is a schematic diagram of a system structure for implementing the method of the present embodiment. Referring to FIG10 , first, after receiving a target prompt word or intent information containing the target prompt word, the table data service processes it, obtains and parses the table data, and generates a data inference request. The table data service then sends the data inference request to an execution plug-in. The execution plug-in, which includes a large language model or is capable of implementing a language model, processes the data inference request and generates a response message, shown as inference information Info_1, which is returned to the table data service. Specifically, the execution plug-in includes a first agent implemented based on the large language model. The first agent invokes an external knowledge base, obtains a response message output by the external knowledge base, and sends it to the table data service. The table data service then formats the response message returned by the execution plug-in, generating a response message in a specific target data format that can be displayed on the terminal device's interactive interface, shown as inference information Info_2. The response message is then returned to the terminal device, which displays it on the interactive interface.

[0110] In the steps of this embodiment, by decoupling the table data service and the execution plug-in, the process of processing the table data to generate a reply message is decomposed into two independent steps: data processing and language model processing, thereby achieving the decoupling of the business data processing flow. Subsequently, the table data service and the execution plug-in can be easily adjusted independently to improve the scalability and maintainability of the program.

[0111] For example, a first agent based on a reason-action (ReAct) model is deployed in the execution plug-in, as shown in FIG11 , and a specific implementation of step S205 includes:

[0112] Step S2051: Process the data reasoning request through the first agent to obtain a first result, wherein the first agent is used to call the first external interface of the external knowledge base to obtain the first result returned by the external knowledge base.

[0113] First, a brief introduction to the reason-action model: The ReAct (reason, act) model integrates reasoning and action capabilities and is a language model that uses natural language reasoning to solve complex tasks. The ReAct model is designed to be used for tasks that allow language models to perform certain operations. For example, in the MRKL system, the language model can interact with an external knowledge base by calling an external interface to retrieve information. When a question is asked, the language model can choose to perform operations to retrieve information, and then answer the question based on the retrieved information. The intelligent agent (Agent, also called an agent) based on the ReAct model can receive data reasoning requests and retrieve the content of the external knowledge base by calling the functional interface of the external knowledge base, or use the capabilities provided by the external knowledge base to obtain the first result returned by the external knowledge base.

[0114] For example, as shown in FIG12 , a possible implementation of step S2051 includes:

[0115] Step S2051A: Process the data reasoning request through the first agent and generate the reasoning target.

[0116] Step S2051B: Generate a first action instruction based on the inference target.

[0117] Step S2051C: By executing the first action instruction, the first module interface of the code processing module is called to generate a first program code corresponding to the first action instruction and applied to the external knowledge base.

[0118] Step S2051D: By executing the first program code, the first external interface of the external knowledge base is called to obtain the first result returned by the external knowledge base.

[0119] Exemplarily, Figure 13 is a structural diagram of a first intelligent agent provided by an embodiment of the present disclosure. The above steps are described in detail below in conjunction with Figure 13. As shown in the figure, a language model (shown as LLM in the figure) is included in the first intelligent agent. First, after the data reasoning request is input into the first intelligent agent, a system prompt word (target prompt word) is obtained based on the data reasoning request, and the language model (shown as LLM+system prompt word in the figure) is input. The language model generates an inference target. Then, a first action instruction is generated based on the inference target, and by executing the first action instruction, the first module interface (i.e., action) of the code processing module is called to generate a first program code, wherein the code processing module is a functional module for generating a program code matching an external knowledge base. By calling the first module interface of the code processing module, the program code corresponding to the first action instruction, i.e., the first program code, can be obtained. Access to the external knowledge base can be achieved through the first program code; then, by executing the first program code, the first external interface of the external knowledge base is called to obtain the first result output by the external knowledge base, and the first result is returned to the language model as an observation result (observation). Optionally, the code processing module may further include a second code module, and the specific function of the second code module will be introduced in the subsequent embodiment steps.

[0120] Furthermore, in one possible implementation, in conjunction with the above example, after obtaining the first result, the language model can directly use the first result as a reply message. In another possible implementation, the reply message can also be generated by repeating the above reason-action cycle multiple times. That is, based on the obtained first result, a new first action instruction is generated, and the above steps are repeated. After multiple cycles of reasoning, a final, multiple-updated first result is obtained, and then a reply message is generated based on the first result. Specifically, after step S2051D, the following is also included:

[0121] Step S2051E: If the preset conditions are not met, the first result is processed by the first agent to generate an updated reasoning target, and return to step S2051B; if the preset conditions are met, the first result is output.

[0122] Specifically, after returning to step S2051B, the execution steps include:

[0123] Based on the updated reasoning target, a second action instruction is generated; by executing the second action instruction, the first module interface of the code processing module is called to generate a second program code corresponding to the second action instruction and applied to the external knowledge base; by executing the second program code, the second external interface of the external knowledge base is called to obtain the second result returned by the external knowledge base (i.e., the updated first result). Afterwards, a reply message can be generated based on the second result, or, based on the second result, an updated reasoning target is obtained, and the above steps are further repeated based on the updated reasoning target again until the reasoning target or the obtained reasoning result (such as the third result, the fourth result, etc.) meets the requirements. Afterwards, a reply message is generated based on the reasoning result (such as the second result) to realize chain processing of multiple steps of the table data.

[0124] Step S2052: Generate a reply message based on the first result.

[0125] For example, as shown in FIG14 , the specific implementation of step S2052 includes:

[0126] Step S2052A: Send the first result to the table data service.

[0127] Step S2052B: Process the first result through the form data service and generate a reply message in the target data format.

[0128] For example, the first result can be the first result obtained in the first reason-action cycle in the above-described embodiment, or it can be an updated first result obtained after multiple reason-action cycles. The first result generated by the first agent is sent to the table data service, which processes the first result and generates a reply message in the target data format, so that the reply message can be displayed in the interactive interface.

[0129] Optionally, before step S2051, the method further includes:

[0130] Step S2050: Acquire historical request information received before the trigger instruction.

[0131] Correspondingly, the specific implementation method of step S2051 includes: processing the data inference request and historical request information through the first intelligent agent to obtain the first result.

[0132] Exemplarily, on the other hand, the terminal device can record historical request information input by the user, thereby enabling data analysis based on multiple conversations. Specifically, before the user inputs the trigger command, the user also inputs other commands into the terminal device. For example, the user enters the text command "Count the book sales volume in the past week" into the terminal device through the intelligent assistant. The terminal device will save this text command as historical request information. In one possible implementation, the execution plug-in is provided with a new memory module for caching these historical requests. Thereafter, before the terminal device processes the data reasoning request through the first agent (i.e., before executing step S2051), the terminal device loads at least one historical request information through this memory module. The language model then processes the historical request information and the data reasoning request (i.e., the historical request information and the data reasoning request are simultaneously input to the first agent), executing the reason-action loop described in the above embodiment to obtain the first result. In this embodiment, by combining the historical request information, the language model can be combined with contextual information, thereby obtaining more accurate data analysis results (reply messages) that meet user needs in deep questioning scenarios.

[0133] Optionally, after step S2051, the method further includes:

[0134] Step S2053: Through the first intelligent agent, call the second module interface of the code processing module to generate a correction code corresponding to the first program code, and by executing the correction code, call the first external interface of the external knowledge base to regain the first result returned by the external knowledge base.

[0135] Exemplarily, since the first result is generated by executing the first program code, when the first program code has an exception or error, it will cause an error in calling the external knowledge base, which in turn causes an error in the generated first result. Therefore, in this embodiment, the first result can be detected, and when the detection result is abnormal, the first agent calls the second module interface of the code processing module to repair the generated first program code and obtain the corresponding correction code. Afterwards, the first external interface of the external knowledge base is called by executing the correction code to re-obtain the first result returned by the external knowledge base. Optionally, after re-obtaining the first result returned by the external knowledge base, the first result can be further detected. If the detection result of the first result is normal, the subsequent steps are executed, that is, a reply message is generated based on the first result; and if the first result is still abnormal, the above-mentioned code correction steps can be repeated to try to correct the program code (correction code) again, and the first result returned by the external knowledge base is re-obtained again until the detection result of the first result is normal or a preset condition is met, such as reaching a target number of cycles.

[0136] In this embodiment, the first intelligent agent calls the second module interface of the code processing module to generate the modification code corresponding to the first program code, thereby realizing automatic detection and repair of the external call code created in the process of generating the reply message without manual intervention, thereby improving the stability of program operation and the efficiency of data analysis.

[0137] Step S206: Display the reply message in the dialog component window.

[0138] In this embodiment, the implementation of steps S201, S202, and S206 is the same as the implementation of steps S101, S102, and some of the solutions in step S103 in the embodiment shown in FIG2 of the present disclosure, and will not be described in detail here. It is understood that in this embodiment, it is possible to further combine steps S100 and any one or more steps of S104-S107 in the embodiment shown in FIG2 to achieve the corresponding technical effects and solve the corresponding technical problems. The configuration can be made as needed, and will not be described in detail here.

[0139] Corresponding to the data processing method of the above embodiment, FIG15 is a block diagram of the structure of the data processing device provided by the embodiment of the present disclosure. For ease of explanation, only the parts related to the embodiment of the present disclosure are shown. Referring to FIG15, the data processing device 3 includes:

[0140] A display unit 31 is configured to display table data, where the table data includes at least one data object and at least one basic statistical information corresponding to the data object;

[0141] The receiving unit 32 is configured to present a dialog component window on the page of the table data in response to a triggering operation on a preset control;

[0142] The processing unit 33 is used to display a reply message in the dialogue component window in response to receiving a first message containing a target requirement through the dialogue component window, wherein the reply message includes conclusion content and / or a recommended requirement template for the target requirement; wherein the conclusion content includes at least one of text, table and chart.

[0143] In one embodiment of the present disclosure, after the dialog component window is presented on the page of tabular data, the processing unit 33 is further used to: display at least one alternative requirement text in the dialog component window, the alternative requirement text representing a requirement template for the target requirement; and / or, display capability description information of the dialog component window in the dialog component window.

[0144] In one embodiment of the present disclosure, the first message includes an indicator and a first requirement text, wherein the first requirement text is used to indicate target data in the table data: the processing unit 33 is specifically used to: in response to the indicator received through the dialogue component window, obtain the target data indicated by the first requirement text, and display a reply message in the dialogue component window based on the target data.

[0145] In one embodiment of the present disclosure, after the dialog component window displays the reply message, the processing unit 33 is further configured to: in response to a second message received through the dialog component window, insert the conclusion content into the table data to generate modified table data.

[0146] In one embodiment of the present disclosure, the reply message includes target advanced statistical information. After the reply message is generated, the display unit 31 is further used to: display at least one extended text, which is generated based on the reply message, and the extended text is used to represent the inference rules for inferring tabular data based on the target advanced statistical information; the processing unit 33 is further used to: obtain updated target requirements in response to a selection instruction for the extended text, and generate an updated reply message based on the updated target requirements; the display unit 31 is further used to: display the updated reply message in the dialog component window.

[0147] In one embodiment of the present disclosure, the display unit 31 is further used to: display at least one implementation text corresponding to the target requirement, and the implementation text is used to characterize the implementation method of the inference rule corresponding to the target requirement; the receiving unit 32 is further used to: obtain the target requirement in response to the selection instruction for the implementation text.

[0148] In one embodiment of the present disclosure, after generating a reply message, the display unit 31 is further used to: display a first prompt information, the first prompt information is used to indicate the problem data in the table data and / or the reply message, wherein the problem data includes at least one of the following: a data object whose object identifier is missing or exceeds a threshold; advanced statistical information whose statistical value is missing or exceeds a threshold.

[0149] In one embodiment of the present disclosure, the reply message includes target advanced statistical information; after the reply message is displayed in the dialog component window, the display unit 31 is further used to: display second prompt information corresponding to the target advanced statistical information, and the second prompt information is used to represent the generation rules of the target advanced statistical information.

[0150] In one embodiment of the present disclosure, the processing unit 33 is further used to: generate chart information corresponding to the reply message through the language model, and the chart information is used to visually display the reply message; the display unit 31 is further used to: display the chart information in the dialogue component window.

[0151] In one embodiment of the present disclosure, the target requirement is used to characterize the inference rules for inferring tabular data based on advanced statistical information, and the advanced statistical information is generated based on basic statistical information. The processing unit 33 is also used to: generate a reply message according to the target requirement, wherein the reply message is generated based on the language model, and the reply message includes the target data object obtained based on the inference rule.

[0152] In one embodiment of the present disclosure, when generating a reply message according to a target requirement, the processing unit 33 is specifically used to: obtain a target prompt word according to the target requirement, and the target prompt word is used to guide the language model to generate a reply message based on an inference rule; process the target prompt word through a preset table data service to generate a data inference request, and the table data service is used to obtain table data, and generate a data inference request based on the table data and the target prompt word; based on the data inference request, call an execution plug-in implemented based on the language model to obtain a reply message corresponding to the table data, and the execution plug-in is used to call an external knowledge base to process the table data to generate a reply message.

[0153] In one embodiment of the present disclosure, when the processing unit 33 processes the target prompt word through a preset table data service and generates a data inference request, it is specifically used to: identify the table data structure of the table data through the table data service; load the table data based on the table data structure, and convert the table data into parsed data that can be used to input the execution plug-in; generate a data inference request based on the parsed data and the target prompt word.

[0154] In one embodiment of the present disclosure, a first agent based on a reason-action model is deployed in the execution plug-in. When the processing unit 33 calls the execution plug-in implemented based on the language model based on the data reasoning request and obtains a reply message corresponding to the table data, it is specifically used to: process the data reasoning request through the first agent to obtain a first result, wherein the first agent is used to call the first external interface of the external knowledge base to obtain the first result returned by the external knowledge base; and generate a reply message based on the first result.

[0155] In one embodiment of the present disclosure, when the processing unit 33 processes a data reasoning request through a first intelligent agent and obtains a first result, it is specifically used to: process the data reasoning request through the first intelligent agent to generate an reasoning target; generate a first action instruction based on the reasoning target; call the first module interface of the code processing module by executing the first action instruction to generate a first program code corresponding to the first action instruction and applied to an external knowledge base; call the first external interface of the external knowledge base by executing the first program code to obtain the first result returned by the external knowledge base.

[0156] In one embodiment of the present disclosure, after obtaining the first result returned by the external knowledge base, the processing unit 33 is also used to: process the first result through the first intelligent agent to generate an updated reasoning target; when the processing unit 33 generates a reply message based on the first result, it is specifically used to: generate a second action instruction based on the updated reasoning target; by executing the second action instruction, call the first module interface of the code processing module to generate a second program code corresponding to the second action instruction applied to the external knowledge base; by executing the second program code, call the second external interface of the external knowledge base to obtain the second result returned by the external knowledge base; and generate a reply message based on the second result.

[0157] In one embodiment of the present disclosure, after obtaining the first result returned by the external knowledge base, the processing unit 33 is also used to: call the second module interface of the code processing module through the first intelligent agent to generate a correction code corresponding to the first program code; and call the first external interface of the external knowledge base by executing the correction code to regain the first result returned by the external knowledge base.

[0158] In one embodiment of the present disclosure, when generating a reply message based on the first result, the processing unit 33 is specifically configured to: send the first result to a table data service; process the first result through the table data service to generate a reply message having a target data format.

[0159] In one embodiment of the present disclosure, the processing unit 33 is also used to: obtain historical target requirements received before the target requirements; when the processing unit 33 processes the data reasoning request through the first intelligent agent to obtain the first result, it is specifically used to: process the data reasoning request and the historical target requirements through the first intelligent agent to obtain the first result.

[0160] The display unit 31, receiving unit 32 and processing unit 33 are connected in sequence. The data processing device 3 provided in this embodiment can execute the technical solution of the above method embodiment, and its implementation principle and technical effect are similar, so this embodiment will not be repeated here.

[0161] FIG16 is a schematic diagram of the structure of an electronic device provided by an embodiment of the present disclosure. As shown in FIG16 , the electronic device 4 includes:

[0162] A processor 41, and a memory 42 communicatively connected to the processor 41;

[0163] Memory 42 stores computer-executable instructions;

[0164] The processor 41 executes the computer-executable instructions stored in the memory 42 to implement the data processing method in the embodiments shown in Figures 2 to 14.

[0165] Optionally, the processor 41 and the memory 42 are connected via a bus 43 .

[0166] The relevant explanations can be understood by referring to the relevant descriptions and effects corresponding to the steps in the embodiments corresponding to Figures 2 to 14, and no further details will be given here.

[0167] An embodiment of the present disclosure provides a computer-readable storage medium, which stores computer-executable instructions. When the computer-executable instructions are executed by a processor, they are used to implement the data processing method provided in any of the embodiments corresponding to Figures 2 to 14 of the present disclosure.

[0168] In order to implement the above embodiment, the present disclosure further provides an electronic device.

[0169] Referring to FIG17 , a schematic diagram of the structure of an electronic device 900 suitable for implementing an embodiment of the present disclosure is shown. The electronic device 900 may be a terminal device or a server. The terminal device may include, but is not limited to, mobile terminals such as mobile phones, laptop computers, digital broadcast receivers, personal digital assistants (PDAs), tablet computers (Portable Android Devices, PADs), portable multimedia players (PMPs), vehicle-mounted terminals (e.g., vehicle-mounted navigation terminals), and fixed terminals such as digital TVs and desktop computers. The electronic device shown in FIG17 is merely an example and should not limit the functionality and scope of use of the embodiments of the present disclosure.

[0170] As shown in FIG17 , the electronic device 900 may include a processing device (e.g., a central processing unit, a graphics processing unit, etc.) 901, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 902 or a program loaded from a storage device 908 into a random access memory (RAM) 903. Various programs and data required for the operation of the electronic device 900 are also stored in the RAM 903. The processing device 901, the ROM 902, and the RAM 903 are connected to each other via a bus 904. An input / output (I / O) interface 905 is also connected to the bus 904.

[0171] Typically, the following devices can be connected to the I / O interface 905: input devices 906 including, for example, a touch screen, a touchpad, a keyboard, a mouse, a camera, a microphone, an accelerometer, a gyroscope, etc.; output devices 907 including, for example, a liquid crystal display (LCD), a speaker, a vibrator, etc.; storage devices 908 including, for example, a magnetic tape, a hard disk, etc.; and communication devices 909. The communication device 909 can allow the electronic device 900 to communicate with other devices wirelessly or by wire to exchange data. Although FIG17 shows an electronic device 900 with various devices, it should be understood that it is not required to implement or have all of the devices shown. More or fewer devices may be implemented or have alternatively.

[0172] According to an embodiment of the present disclosure, the process described above with reference to the flowchart can be implemented as a computer software program. For example, an embodiment of the present disclosure includes a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program includes program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from the network via the communication device 909, or installed from the storage device 908, or installed from the ROM 902. When the computer program is executed by the processing device 901, the above-mentioned functions defined in the method of the embodiment of the present disclosure are performed.

[0173] It should be noted that the computer-readable medium mentioned above in the present disclosure may be a computer-readable signal medium or a computer-readable storage medium, or any combination of the two. A computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or component, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present disclosure, a computer-readable storage medium may be any tangible medium that contains or stores a program that can be used by or in conjunction with an instruction execution system, device, or component. In the present disclosure, a computer-readable signal medium may include a data signal propagated in baseband or as part of a carrier wave, which carries computer-readable program code. Such a propagated data signal may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. A computer-readable signal medium may also be any computer-readable medium other than a computer-readable storage medium that can transmit, propagate, or transport a program for use by or in conjunction with an instruction execution system, apparatus, or device. The program code contained on the computer-readable medium may be transmitted using any suitable medium, including but not limited to wires, optical cables, RF (radio frequency), etc., or any suitable combination thereof.

[0174] The computer-readable medium may be included in the electronic device, or may exist independently without being incorporated into the electronic device.

[0175] The computer-readable medium carries one or more programs. When the one or more programs are executed by the electronic device, the electronic device executes the method shown in the above embodiment.

[0176] Computer program code for performing the operations of the present disclosure may be written in one or more programming languages, or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, C++, and conventional procedural programming languages ​​such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a stand-alone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving a remote computer, the remote computer may be connected to the user's computer through any type of network, including a Local Area Network (LAN) or a Wide Area Network (WAN), or may be connected to an external computer (e.g., via the Internet using an Internet Service Provider).

[0177] The flowcharts and block diagrams in the accompanying drawings illustrate the possible implementation architecture, functions and operations of the systems, methods and computer program products according to various embodiments of the present disclosure. In this regard, each box in the flowchart or block diagram can represent a module, program segment, or a part of code, and the module, program segment, or a part of code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in a different order than that marked in the accompanying drawings. For example, two boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram and / or flowchart, and the combination of the boxes in the block diagram and / or flowchart, can be implemented with a dedicated hardware-based system that performs the specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0178] The units involved in the embodiments described in this disclosure may be implemented in software or hardware. In some cases, the name of a unit does not limit the unit itself. For example, the first acquisition unit may also be described as a "unit for acquiring at least two Internet Protocol addresses."

[0179] The functions described above in the present disclosure may be performed at least in part by one or more hardware logic components. For example, and without limitation, exemplary types of hardware logic components that may be used include: field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), complex programmable logic devices (CPLDs), and the like.

[0180] In the context of the present disclosure, a machine-readable medium can be a tangible medium that can contain or store a program for use by or in conjunction with an instruction execution system, device or equipment. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium can include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or equipment, or any suitable combination of the foregoing. A more specific example of a machine-readable storage medium can include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0181] In a first aspect, according to one or more embodiments of the present disclosure, a data processing method is provided, comprising:

[0182] Displaying tabular data, wherein the tabular data includes at least one data object and at least one basic statistical information corresponding to the data object; presenting a dialog component window on the page of the tabular data in response to a triggering operation on a preset control; displaying a reply message in the dialog component window in response to receiving a first message containing a target requirement through the dialog component window, wherein the reply message includes conclusion content and / or a recommended requirement template for the target requirement; wherein the conclusion content includes at least one of text, a table and a chart.

[0183] According to one or more embodiments of the present disclosure, after the dialog component window is presented on the page of the tabular data, it also includes: displaying at least one alternative requirement text within the dialog component window, wherein the alternative requirement text represents a requirement template for the target requirement; and / or displaying capability description information of the dialog component window within the dialog component window.

[0184] According to one or more embodiments of the present disclosure, the first message includes an indicator and a first requirement text, wherein the first requirement text is used to indicate the target data in the table data: in response to receiving the first message containing the target requirement through the dialogue component window, displaying a reply message in the dialogue component window includes: in response to the indicator received through the dialogue component window, obtaining the target data indicated by the first requirement text, and displaying the reply message in the dialogue component window based on the target data.

[0185] According to one or more embodiments of the present disclosure, after the dialog component window displays the reply message, the method further includes: in response to a second message received through the dialog component window, inserting the conclusion content into the table data to generate modified table data.

[0186] According to one or more embodiments of the present disclosure, the reply message includes target high-level statistical information, and after the reply message is displayed in the dialog component window, it also includes: displaying at least one extended text, wherein the extended text is generated based on the reply message, and the extended text is used to represent the inference rules for inferring the tabular data based on the target high-level statistical information; in response to a selection instruction for the extended text, obtaining an updated target requirement, and generating an updated reply message based on the updated target requirement; and displaying the updated reply message in the dialog component window.

[0187] According to one or more embodiments of the present disclosure, after the reply message is displayed in the dialogue component window, it also includes: displaying at least one implementation text corresponding to the target requirement, the implementation text is used to characterize the implementation method of the inference rule corresponding to the target requirement; and obtaining the target requirement in response to a selection instruction for the implementation text.

[0188] According to one or more embodiments of the present disclosure, after the reply message is displayed in the dialog component window, it also includes: displaying a first prompt information, wherein the first prompt information is used to indicate problem data in the table data and / or the reply message, wherein the problem data includes at least one of the following: a data object whose object identification is missing or exceeds a threshold; advanced statistical information whose statistical value is missing or exceeds a threshold.

[0189] According to one or more embodiments of the present disclosure, the reply message includes target advanced statistical information; after the reply message is displayed in the dialog component window, it also includes: displaying second prompt information corresponding to the target advanced statistical information, and the second prompt information is used to represent the generation rule of the target advanced statistical information.

[0190] According to one or more embodiments of the present disclosure, the method further includes: generating chart information corresponding to the reply message through the language model, the chart information being used to visually display the reply message; and displaying the chart information in the dialog component window.

[0191] According to one or more embodiments of the present disclosure, the target requirement is used to characterize the inference rules for inferring the tabular data based on advanced statistical information, and the advanced statistical information is generated based on the basic statistical information; the method also includes: generating the reply message according to the target requirement, wherein the reply message is generated based on a language model, and the reply message includes a target data object obtained based on the inference rule.

[0192] According to one or more embodiments of the present disclosure, generating the reply message according to the target requirement includes: obtaining a target prompt word according to the target requirement, the target prompt word being used to guide the language model to generate a reply message based on the inference rule; processing the target prompt word through a preset table data service to generate a data inference request, the table data service being used to obtain the table data, and generating the data inference request based on the table data and the target prompt word; based on the data inference request, calling an execution plug-in implemented based on the language model to obtain a reply message corresponding to the table data, the execution plug-in being used to call an external knowledge base to process the table data to generate the reply message.

[0193] According to one or more embodiments of the present disclosure, the target prompt word is processed through a preset table data service to generate a data inference request, including: identifying the table data structure of the table data through the table data service; loading the table data based on the table data structure, and converting the table data into parsed data that can be used to input the execution plug-in; and generating the data inference request based on the parsed data and the target prompt word.

[0194] According to one or more embodiments of the present disclosure, a first agent based on a reason-action model is deployed in the execution plug-in, and based on the data reasoning request, the execution plug-in implemented based on the language model is called to obtain a reply message corresponding to the tabular data, including: processing the data reasoning request through the first agent to obtain a first result, wherein the first agent is used to call the first external interface of the external knowledge base to obtain the first result returned by the external knowledge base; and generating the reply message based on the first result.

[0195] According to one or more embodiments of the present disclosure, the data reasoning request is processed by the first intelligent agent to obtain a first result, including: processing the data reasoning request by the first intelligent agent to generate an inference target; generating a first action instruction based on the inference target; calling the first module interface of the code processing module by executing the first action instruction to generate a first program code corresponding to the first action instruction and applied to the external knowledge base; calling the first external interface of the external knowledge base by executing the first program code to obtain the first result returned by the external knowledge base.

[0196] According to one or more embodiments of the present disclosure, after obtaining the first result returned by the external knowledge base, it also includes: processing the first result through the first intelligent agent to generate an updated reasoning target; generating the reply message based on the first result includes: generating a second action instruction based on the updated reasoning target; calling the first module interface of the code processing module by executing the second action instruction to generate a second program code corresponding to the second action instruction applied to the external knowledge base; calling the second external interface of the external knowledge base by executing the second program code to obtain the second result returned by the external knowledge base; generating the reply message according to the second result.

[0197] According to one or more embodiments of the present disclosure, after obtaining the first result returned by the external knowledge base, it also includes: calling the second module interface of the code processing module through the first intelligent agent to generate a correction code corresponding to the first program code; calling the first external interface of the external knowledge base by executing the correction code to re-obtain the first result returned by the external knowledge base.

[0198] According to one or more embodiments of the present disclosure, generating the reply message based on the first result includes: sending the first result to the form data service; processing the first result through the form data service to generate a reply message with a target data format.

[0199] According to one or more embodiments of the present disclosure, the method further includes: obtaining historical target requirements received before the target requirement; processing the data reasoning request through the first intelligent agent to obtain a first result, including: processing the data reasoning request and the historical target requirements through the first intelligent agent to obtain the first result.

[0200] In a second aspect, according to one or more embodiments of the present disclosure, there is provided a data processing apparatus, comprising:

[0201] a display unit, configured to display tabular data, wherein the tabular data includes at least one data object and at least one basic statistical information corresponding to the data object;

[0202] A receiving unit, configured to present a dialog component window on the page of the table data in response to a triggering operation on a preset control;

[0203] A processing unit is used to display a reply message in the dialogue component window in response to receiving a first message containing a target requirement through the dialogue component window, wherein the reply message includes conclusion content and / or a recommended requirement template for the target requirement; wherein the conclusion content includes at least one of text, table and chart.

[0204] In one embodiment of the present disclosure, after the dialog component window is presented on the page of the table data, the processing unit is further used to: display at least one alternative requirement text in the dialog component window, wherein the alternative requirement text represents a requirement template for the target requirement; and / or display capability description information of the dialog component window in the dialog component window.

[0205] In one embodiment of the present disclosure, the first message includes an indicator and a first requirement text, wherein the first requirement text is used to indicate the target data in the table data: the processing unit 33 is specifically used to: in response to the indicator received through the dialogue component window, obtain the target data indicated by the first requirement text, and display a reply message in the dialogue component window based on the target data.

[0206] In one embodiment of the present disclosure, after the dialogue component window displays a reply message, the processing unit is further used to: in response to a second message received through the dialogue component window, insert the conclusion content into the table data to generate modified table data.

[0207] According to one or more embodiments of the present disclosure, the reply message includes target advanced statistical information. After the reply message is generated, the display unit 31 is further used to: display at least one extended text, which is generated based on the reply message, and the extended text is used to represent the inference rules for inferring tabular data based on the target advanced statistical information; the processing unit is further used to: obtain updated target requirements in response to a selection instruction for the extended text, and generate an updated reply message based on the updated target requirements; the display unit is further used to: display the updated reply message in the dialog component window.

[0208] According to one or more embodiments of the present disclosure, the display unit is further used to: display at least one implementation text corresponding to the target requirement, and the implementation text is used to characterize the implementation method of the inference rule corresponding to the target requirement; the receiving unit is further used to: obtain the target requirement in response to the selection instruction for the implementation text.

[0209] According to one or more embodiments of the present disclosure, after generating a reply message, the display unit is further used to: display a first prompt information, the first prompt information is used to indicate the problem data in the table data and / or the reply message, wherein the problem data includes at least one of the following: a data object whose object identification is missing or exceeds a threshold; advanced statistical information whose statistical value is missing or exceeds a threshold.

[0210] According to one or more embodiments of the present disclosure, the reply message includes target advanced statistical information; after the reply message is displayed in the dialog component window, the display unit is further used to: display second prompt information corresponding to the target advanced statistical information, and the second prompt information is used to characterize the generation rules of the target advanced statistical information.

[0211] In one embodiment of the present disclosure, the processing unit is further used to: generate chart information corresponding to the reply message through the language model, and the chart information is used to visually display the reply message; the display unit is further used to: display the chart information in the dialogue component window.

[0212] According to one or more embodiments of the present disclosure, the target requirement is used to characterize the inference rules for inferring the tabular data based on advanced statistical information, and the advanced statistical information is generated based on the basic statistical information. The processing unit 33 is also used to: generate the reply message according to the target requirement, wherein the reply message is generated based on a language model, and the reply message includes a target data object obtained based on the inference rule.

[0213] According to one or more embodiments of the present disclosure, when the processing unit generates the reply message according to the target requirement, it is specifically used to: obtain the target prompt word according to the target requirement, and the target prompt word is used to guide the language model to generate a reply message based on the inference rule; process the target prompt word through a preset table data service to generate a data inference request, and the table data service is used to obtain table data and generate a data inference request based on the table data and the target prompt word; based on the data inference request, call the execution plug-in implemented based on the language model to obtain the reply message corresponding to the table data, and the execution plug-in is used to call the external knowledge base to process the table data to generate a reply message.

[0214] According to one or more embodiments of the present disclosure, when the processing unit processes the target prompt word through a preset table data service and generates a data inference request, it is specifically used to: identify the table data structure of the table data through the table data service; load the table data based on the table data structure, and convert the table data into parsed data that can be used to input the execution plug-in; and generate the data inference request based on the parsed data and the target prompt word.

[0215] According to one or more embodiments of the present disclosure, a first agent based on a reason-action model is deployed in the execution plug-in. When the processing unit calls the execution plug-in implemented based on the language model based on the data reasoning request and obtains a reply message corresponding to the table data, it is specifically used to: process the data reasoning request through the first agent to obtain a first result, wherein the first agent is used to call the first external interface of the external knowledge base to obtain the first result returned by the external knowledge base; and generate the reply message based on the first result.

[0216] According to one or more embodiments of the present disclosure, when the processing unit processes the data reasoning request through the first intelligent agent and obtains the first result, it is specifically used to: process the data reasoning request through the first intelligent agent to generate an inference target; generate a first action instruction based on the inference target; call the first module interface of the code processing module by executing the first action instruction to generate a first program code corresponding to the first action instruction and applied to the external knowledge base; call the first external interface of the external knowledge base by executing the first program code to obtain the first result returned by the external knowledge base.

[0217] According to one or more embodiments of the present disclosure, after obtaining the first result returned by the external knowledge base, the processing unit is further used to: process the first result through the first agent to generate an updated reasoning target; when the processing unit generates the reply message based on the first result, it is specifically used to: generate a second action instruction based on the updated reasoning target; by executing the second action instruction, call the first module interface of the code processing module to generate a second program code corresponding to the second action instruction applied to the external knowledge base; by executing the second program code, call the second external interface of the external knowledge base to obtain the second result returned by the external knowledge base; and generate the reply message based on the second result.

[0218] According to one or more embodiments of the present disclosure, after obtaining the first result returned by the external knowledge base, the processing unit is also used to: call the second module interface of the code processing module through the first intelligent agent to generate a correction code corresponding to the first program code; and call the first external interface of the external knowledge base by executing the correction code to regain the first result returned by the external knowledge base.

[0219] According to one or more embodiments of the present disclosure, when the processing unit generates the reply message based on the first result, it is specifically used to: send the first result to the table data service; process the first result through the table data service to generate a reply message with a target data format.

[0220] According to one or more embodiments of the present disclosure, the processing unit is further used to: obtain historical target requirements received before the target requirement; when the processing unit processes the data reasoning request through the first intelligent agent to obtain the first result, it is specifically used to: process the data reasoning request and the historical target requirements through the first intelligent agent to obtain the first result.

[0221] In a third aspect, according to one or more embodiments of the present disclosure, there is provided an electronic device, comprising: at least one processor and a memory;

[0222] The memory stores computer-executable instructions;

[0223] The at least one processor executes the computer-executable instructions stored in the memory, so that the at least one processor performs the data processing method described in the first aspect and various possible designs of the first aspect.

[0224] In a fourth aspect, according to one or more embodiments of the present disclosure, a computer-readable storage medium is provided, in which computer-executable instructions are stored. When a processor executes the computer-executable instructions, the data processing method described in the first aspect and various possible designs of the first aspect is implemented.

[0225] In a fifth aspect, according to one or more embodiments of the present disclosure, a computer program product is provided, comprising a computer program, which, when executed by a processor, implements the data processing method described in the first aspect and various possible designs of the first aspect.

[0226] The above description is merely an illustration of the technical principles employed in the embodiments of the present disclosure. Those skilled in the art should understand that the scope of disclosure involved in the present disclosure is not limited to the technical solutions formed by a specific combination of the above-mentioned technical features, but should also encompass other technical solutions formed by any combination of the above-mentioned technical features or their equivalents without departing from the above-mentioned disclosed concepts. For example, a technical solution formed by mutually replacing the above-mentioned features with (but not limited to) technical features with similar functions disclosed in this disclosure.

[0227] In addition, although each operation is described in a specific order, this should not be understood as requiring these operations to be performed in the specific order shown or in a sequential order. Under certain circumstances, multitasking and parallel processing may be advantageous. Similarly, although some specific implementation details have been included in the above discussion, these should not be interpreted as limiting the scope of the present disclosure. Some features described in the context of a separate embodiment can also be implemented in a single embodiment in combination. On the contrary, the various features described in the context of a single embodiment can also be implemented in multiple embodiments individually or in any suitable sub-combination mode.

[0228] Although the present disclosure has been described in language specific to structural features and / or methodological logical acts, it should be understood that the subject matter defined in the appended claims is not necessarily limited to the specific features or acts described above. Rather, the specific features and acts described above are merely example forms of implementing the claims.

Claims

1. A data processing method, comprising: Displaying table data, wherein the table data includes at least one data object and at least one basic statistical information corresponding to the data object; In response to a triggering operation on a preset control, presenting a dialog component window on the page of the table data; In response to receiving a first message containing a target requirement through the dialogue component window, a reply message is displayed in the dialogue component window, the reply message including conclusion content and / or a recommended requirement template for the target requirement; wherein the conclusion content includes at least one of text, table and chart.

2. The method according to claim 1, wherein: After the page of the table data presents the dialog component window, the method further comprises: In the dialog component window, displaying at least one candidate requirement text, wherein the candidate requirement text represents a requirement template for the target requirement; and / or, In the dialog component window, capability description information of the dialog component window is displayed.

3. The method according to claim 1 or 2, wherein: The first message includes an indication mark and a first requirement text, and the first requirement text is used to indicate the target data in the table data: In response to receiving a first message containing a target requirement through the dialogue component window, displaying a reply message in the dialogue component window includes: in response to an indication identifier received through the dialogue component window, obtaining target data indicated by the first requirement text, and displaying the reply message in the dialogue component window based on the target data.

4. The method according to any one of claims 1 to 3, wherein: After the dialog component window displays the reply message, the method further includes: In response to a second message received through the dialog component window, the conclusion content is inserted into the table data to generate modified table data.

5. The method according to any one of claims 1 to 4, wherein: The reply message includes target advanced statistical information. After the reply message is displayed in the dialog component window, the method further includes: displaying at least one extended text, the extended text being generated based on the reply message, the extended text being used to represent an inference rule for inferring the table data based on the target high-level statistical information; In response to the selection instruction for the extended text, an updated target requirement is obtained, and an updated reply message is generated according to the updated target requirement; The updated reply message is displayed in the dialog component window.

6. The method according to any one of claims 1 to 5, wherein: After the dialog component window displays the reply message, the method further includes: Displaying at least one implementation text corresponding to the target requirement, where the implementation text is used to represent the implementation method of the inference rule corresponding to the target requirement; In response to a selection instruction for the implementation text, the target requirement is obtained.

7. The method according to any one of claims 1 to 6, wherein: After the dialog component window displays the reply message, the method further includes: Displaying first prompt information, where the first prompt information is used to indicate the form data and / or question data in the reply message, wherein the question data includes at least one of the following: Object identification of data objects that are missing or exceed the threshold; Advanced statistics where the statistic value is missing or exceeds the threshold.

8. The method according to any one of claims 1 to 7, wherein: The reply message includes target advanced statistical information; After the dialog component window displays the reply message, the method further includes: Second prompt information corresponding to the target advanced statistical information is displayed, where the second prompt information is used to represent a generation rule of the target advanced statistical information.

9. The method according to any one of claims 1 to 8, further comprising: Generate chart information corresponding to the reply message through the language model, where the chart information is used to visually display the reply message; The chart information is displayed in the dialog component window.

10. The method according to any one of claims 1 to 9, wherein: The target requirement is used to characterize an inference rule for inferring the table data based on advanced statistical information, where the advanced statistical information is generated based on the basic statistical information; the method further includes: The reply message is generated according to the target requirement, wherein the reply message is generated based on a language model and includes a target data object obtained based on the inference rule.

11. The method according to claim 10, wherein: The step of generating the reply message according to the target requirement includes: According to the target requirement, a target prompt word is obtained, wherein the target prompt word is used to guide the language model to generate a reply message based on the inference rule; Processing the target prompt word through a preset table data service to generate a data inference request, wherein the table data service is used to obtain the table data and generate the data inference request based on the table data and the target prompt word; Based on the data inference request, an execution plug-in implemented based on the language model is called to obtain a reply message corresponding to the table data. The execution plug-in is used to call an external knowledge base to process the table data to generate the reply message.

12. The method according to claim 11, wherein: The process of processing the target prompt word through a preset table data service to generate a data inference request includes: identifying, by the table data service, a table data structure of the table data; Based on the table data structure, loading the table data and converting the table data into parsed data that can be used to input the execution plug-in; The data inference request is generated based on the parsed data and the target prompt word.

13. The method according to claim 11, wherein: The execution plug-in has a first agent based on a reason-action model deployed therein, and based on the data reasoning request, the execution plug-in implemented based on the language model is called to obtain a reply message corresponding to the table data, including: Processing the data reasoning request through the first agent to obtain a first result, wherein the first agent is used to call a first external interface of the external knowledge base to obtain the first result returned by the external knowledge base; The reply message is generated based on the first result.

14. The method according to claim 13, wherein: The step of processing the data reasoning request through the first agent to obtain a first result includes: Processing the data reasoning request through the first agent to generate a reasoning target; Based on the inference target, generating a first action instruction; By executing the first action instruction, a first module interface of a code processing module is called to generate a first program code corresponding to the first action instruction and applied to the external knowledge base; By executing the first program code, the first external interface of the external knowledge base is called to obtain the first result returned by the external knowledge base.

15. The method according to claim 14, wherein: After obtaining the first result returned by the external knowledge base, the method further includes: Processing the first result through the first agent to generate an updated reasoning target; The generating the reply message based on the first result includes: Based on the updated reasoning target, generating a second action instruction; By executing the second action instruction, the first module interface of the code processing module is called to generate a second program code corresponding to the second action instruction and applied to the external knowledge base; By executing the second program code, a second external interface of the external knowledge base is called to obtain a second result returned by the external knowledge base; The reply message is generated according to the second result.

16. The method according to claim 14, wherein: After obtaining the first result returned by the external knowledge base, the method further includes: Calling, through the first agent, a second module interface of a code processing module to generate a correction code corresponding to the first program code; By executing the correction code, the first external interface of the external knowledge base is called to retrieve the first result returned by the external knowledge base.

17. The method according to any one of claims 13 to 16, wherein: The generating the reply message based on the first result includes: sending the first result to the table data service; The first result is processed through the form data service to generate a reply message having a target data format.

18. The method according to any one of claims 13 to 17, further comprising: Acquire historical target demands received before the target demand; The step of processing the data reasoning request through the first agent to obtain a first result includes: The first intelligent agent processes the data reasoning request and the historical target requirement to obtain the first result.

19. A data processing device, comprising: A display unit configured to display table data, wherein the table data includes at least one data object and at least one basic statistical information corresponding to the data object; A receiving unit, configured to present a dialog component window on the page of the table data in response to a triggering operation on a preset control; The processing unit is configured to display a reply message in the dialogue component window in response to receiving a first message containing a target requirement through the dialogue component window, wherein the reply message includes conclusion content and / or a recommended requirement template for the target requirement; wherein the conclusion content includes at least one of text, table and chart.

20. An electronic device comprising a processor and a memory, wherein: The memory is configured to store computer-executable instructions; The processor is configured to execute the computer-executable instructions stored in the memory, so that the processor performs the data processing method according to any one of claims 1 to 18.

21. A computer-readable storage medium storing computer-executable instructions, wherein: When the processor executes the computer-executable instructions, the data processing method according to any one of claims 1 to 18 is implemented.

22. A computer program product comprising a computer program, wherein: When the computer program is executed by a processor, the data processing method according to any one of claims 1 to 18 is implemented.

Images