Dialogue-based data analysis method, device and electronic equipment based on multi-language model

Abstract

The present disclosure relates to the technical field of computers, and discloses a dialog data analysis method and device based on a multilingual model and an electronic device. The method comprises: displaying a data analysis page, the data analysis page being provided with an analysis control, the analysis control corresponding to at least one language model; in response to a selection instruction for the analysis control, determining a target language model; in response to a first question determined on the data analysis page, generating a first data analysis instruction, the first data analysis instruction being used to determine a first analysis result of the target language model for the first question; and displaying the first analysis result on the data analysis page. The present disclosure can solve the problem of low accuracy of data analysis.

Description

TECHNICAL FIELD

[0001] The present disclosure relates to the technical field of computers, in particular to a dialog-based data analysis method and device based on a multi-language model and an electronic device. BACKGROUND

[0002] At present, more and more natural language processing technologies are integrated into data consumption tools to provide more intelligent data analysis and prediction capabilities, and some data analysis tools begin to provide more advanced data analysis functions such as prediction modeling and natural language queries. In the related art, these data analysis tools mainly use dialog analysis. However, due to the complexity of the data analysis field, the user input question is a very similar set of expressions in the data table, and different data analysis tools will have different understandings for the same question, resulting in low accuracy of data analysis. SUMMARY

[0003] Therefore, the present disclosure provides a dialog-based data analysis method and device based on a multi-language model to solve the problem of low accuracy of data analysis.

[0004] In a first aspect, the present disclosure provides a dialog-based data analysis method based on a multi-language model, the method comprising:

[0005] displaying a data analysis page, the data analysis page being displayed with an analysis control, the analysis control corresponding to at least one language model;

[0006] in response to a selection instruction for the analysis control, determining a target language model;

[0007] in response to a first question determined on the data analysis page, generating a first data analysis instruction, the first data analysis instruction being used to determine a first analysis result of the target language model for the first question;

[0008] displaying the first analysis result on the data analysis page.

[0009] In a second aspect, the present disclosure provides a dialog-based data analysis device based on a multi-language model, the device comprising:

[0010] a first display module configured to display a data analysis page, the data analysis page being displayed with an analysis control, the analysis control corresponding to at least one language model;

[0011] a model selection module configured to determine a target language model in response to a selection instruction for the analysis control;

[0012] The first analysis module is configured to generate a first data analysis instruction in response to a first question determined on the data analysis page, the first data analysis instruction being used to determine a first analysis result of the target language model for the first question.

[0013] The second display module is configured to display the first analysis result on the data analysis page.

[0014] In a third aspect, the present disclosure provides an electronic device, comprising a memory and a processor, the memory and the processor being communicatively connected with each other, the memory storing computer instructions, and the processor executing the computer instructions to perform the multi-language model based conversational data analysis method of the first aspect.

[0015] In a fourth aspect, the present disclosure provides a computer readable storage medium, the computer readable storage medium storing computer instructions, and the computer instructions being used to make a computer execute the multi-language model based conversational data analysis method of the first aspect.

[0016] In a fifth aspect, the present disclosure provides a computer program product, the computer program product comprising computer instructions, and the computer instructions being used to make a computer execute the multi-language model based conversational data analysis method of the first aspect.

[0017] The multi-language model based conversational data analysis method provided by the embodiments of the present disclosure displays analysis controls corresponding to at least one language model on a data analysis page. Therefore, language models suitable for different data analysis scenarios can be provided through the analysis controls. Further, a selected language model corresponding to an analysis control is determined as a target language model. The first analysis result of the target language model for a first question is determined and displayed. Therefore, a target language model suitable for a current data analysis scenario can be selected to perform data analysis, so as to improve the accuracy of data analysis. BRIEF DESCRIPTION OF DRAWINGS

[0018] In order to more clearly illustrate the specific embodiments of the present disclosure or the technical solutions in the prior art, the drawings needed in the specific embodiments or prior art description will be briefly introduced. Obviously, the drawings in the following description are some embodiments of the present disclosure, and those skilled in the art can obtain other drawings according to these drawings without creative labor.

[0019] Figure 1 is a flow diagram of a multi-language model based conversational data analysis method according to an embodiment of the present disclosure;

[0020] Figure 2 is a schematic diagram of a first data analysis page according to an embodiment of the present disclosure;

[0021] Figure 3 FIG. 7 is a schematic diagram of a second data analysis page according to an embodiment of the present disclosure;

[0022] Figure 4 FIG. 8 is a schematic diagram of a third data analysis page according to an embodiment of the present disclosure;

[0023] Figure 5 FIG. 9 is a schematic diagram of a fourth data analysis page according to an embodiment of the present disclosure;

[0024] Figure 6 FIG. 10 is a schematic diagram of a fifth data analysis page according to an embodiment of the present disclosure;

[0025] Figure 7 FIG. 11 is a schematic diagram of a first problem optimization according to an embodiment of the present disclosure;

[0026] Figure 8 FIG. 12 is a schematic diagram of a second problem optimization according to an embodiment of the present disclosure;

[0027] Figure 9 FIG. 13 is a schematic diagram of a third problem optimization according to an embodiment of the present disclosure;

[0028] Figure 10 FIG. 14 is a schematic diagram of a sixth data analysis page according to an embodiment of the present disclosure;

[0029] Figure 11 FIG. 15 is a schematic diagram of a detail page according to an embodiment of the present disclosure;

[0030] Figure 12 FIG. 16 is a schematic diagram of another detail page according to an embodiment of the present disclosure;

[0031] Figure 13 FIG. 17 is a schematic diagram of a seventh data analysis page according to an embodiment of the present disclosure;

[0032] Figure 14 FIG. 18 is a schematic diagram of an eighth data analysis page according to an embodiment of the present disclosure;

[0033] Figure 15 FIG. 19 is a schematic diagram of a ninth data analysis page according to an embodiment of the present disclosure;

[0034] Figure 16 FIG. 20 is a schematic diagram of a task subscription page according to an embodiment of the present disclosure;

[0035] Figure 17 FIG. 21 is a schematic diagram of data push according to an embodiment of the present disclosure;

[0036] Figure 18 FIG. 22 is a schematic diagram of a tenth data analysis page according to an embodiment of the present disclosure;

[0037] Figure 19 is a schematic diagram of storing results according to an embodiment of the present disclosure;

[0038] Figure 20 is a structural block diagram of a dialog-based data analysis device based on a multi-language model according to an embodiment of the present disclosure;

[0039] Figure 21 is a structural block diagram of an electronic device according to an embodiment of the present disclosure. DETAILED DESCRIPTION

[0040] In order to make the objects, technical solutions and advantages of the embodiments of the present disclosure clearer, the technical solutions in the embodiments of the present disclosure will be described clearly and completely below with reference to the drawings in the embodiments of the present disclosure. Obviously, the described embodiments are only part of the embodiments of the present disclosure, rather than all the embodiments of the present disclosure. Based on the embodiments in the present disclosure, all other embodiments obtained by a person skilled in the art without creative work under the premise of the present disclosure, all belong to the scope of protection of the present disclosure.

[0041] It can be understood that, before using the technical solutions disclosed in the embodiments of the present disclosure, the type, use range, use scenario and the like of the personal information involved in the present disclosure should be informed to the user and the authorization of the user should be obtained through appropriate means according to relevant laws and regulations.

[0042] For example, in response to receiving the active request of the user, prompt information is sent to the user to explicitly prompt the user that the operation requested to be executed will need to obtain and use the personal information of the user. Thus, the user can voluntarily choose whether to provide the personal information to the software or hardware such as an electronic device, an application program, a server or a storage medium, etc. that executes the operation of the technical solutions of the present disclosure according to the prompt information.

[0043] As an optional but not limited implementation manner, in response to receiving the active request of the user, the manner of sending the prompt information to the user may, for example, be a pop-up window manner, and the prompt information may, for example, be presented in the form of text in the pop-up window. In addition, the pop-up window may, for example, also carry a selection control for the user to select “agree” or “disagree” to provide the personal information to the electronic device.

[0044] It can be understood that the above notification and obtaining of the authorization of the user are only illustrative, and do not limit the implementation manners of the present disclosure, and other manners meeting the relevant laws and regulations can also be applied to the implementation manners of the present disclosure.

[0045] At present, more and more natural language processing technologies are integrated into data analysis tools to provide more intelligent data analysis and prediction capabilities, and some data analysis tools begin to provide more advanced data analysis functions such as prediction modeling and natural language query. In the related art, these data analysis tools mainly adopt dialog analysis. However, due to the complexity of the data analysis field, the user input question is a very similar set of expressions in the data table, and different data analysis tools have different understandings of the same question, resulting in low accuracy of data analysis.

[0046] Therefore, according to an embodiment of the present disclosure, a dialog data analysis method based on a multi-language model is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although the logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0047] In this embodiment, a dialog data analysis method based on a multi-language model is provided, which can be used in a data analysis platform, Figure 1 is a flowchart of a dialog data analysis method based on a multi-language model according to an embodiment of the present disclosure, as Figure 1 shown, the flow includes the following steps:

[0048] Step S101, display a data analysis page, the data analysis page displays an analysis control, and the analysis control corresponds to at least one language model.

[0049] Specifically, the data analysis page displays analysis controls corresponding to different language models, and the language model suitable for the current data analysis scene can be selected by selecting the analysis control to perform data analysis.

[0050] Step S102, in response to a selection instruction for the analysis control, determine the target language model.

[0051] Specifically, in response to the selection instruction for the analysis control, the language model corresponding to the selected analysis control is determined as the target language model.

[0052] Specifically, the user can select language models with different functions in the language model warehouse, and display the analysis controls corresponding to each language model in the data analysis page. For example, Figure 2As shown, the data analysis page displays analysis control 1, analysis control 2, and analysis control 3. In addition, the model information of the language model corresponding to the analysis control is displayed in the area corresponding to the analysis control, so as to select the analysis control used in the current data analysis from the displayed analysis controls, and determine the language model used in the current data analysis. During the data analysis, the user can switch or add the analysis control corresponding to the different language model at any time.

[0053] In step S103, a first data analysis instruction is generated in response to the first question determined in the data analysis page, and the first data analysis instruction is used to determine the first analysis result of the target language model for the first question.

[0054] Specifically, the target data source for data analysis is displayed on the data analysis page, and the first analysis result is obtained by data analysis based on the target language model, the target data source, and the first question.

[0055] In step S104, the first analysis result is displayed on the data analysis page.

[0056] Specifically, the first analysis result is displayed in the form of a dialogue in the data analysis page. If the selected analysis control corresponds to multiple language models, a dialogue group is constructed based on the multiple language models, and the first question is answered by the multiple language models at the same time.

[0057] As shown in the example, Figure 3 Assuming that the analysis control 2 corresponds to the language model 1 and the language model 2, the first question is answered by the language model 1 and the language model 2 at the same time, and the first analysis result of the language model 1 and the language model 2 for the first question is displayed on the data analysis page.

[0058] The dialogue type data analysis method based on multiple language models provided in the embodiment displays the analysis control corresponding to at least one language model on the data analysis page. Therefore, the language model suitable for different data analysis scenarios can be provided through the analysis control. Further, the language model corresponding to the selected analysis control is determined as the target language model. The first analysis result of the target language model for the first question is determined and displayed. Therefore, the target language model suitable for the current data analysis scenario can be selected to perform data analysis, so as to improve the accuracy of data analysis.

[0059] In some optional embodiments, the data analysis page further displays the target data source for data analysis and the recommended question corresponding to the target data source. Then, the step S103 of generating the first data analysis instruction in response to the first question determined in the data analysis page includes:

[0060] In step a1, the first question is determined in response to the selection instruction for the recommended question.

[0061] Specifically, the selected recommended question is determined as the first question.

[0062] Specifically, attractive questions are provided for the target data source, through which the user is guided to carry out data analysis, master the use skills of the data analysis platform configured with the multi-language model-based conversational data analysis method of the present disclosure, and improve the quality of data analysis services.

[0063] Further, on the one hand, a public data set can be selected as the target data source, and attractive questions are provided for the public data set to guide the user to carry out data analysis. On the other hand, for different user roles or access permissions, business-related permission data sets are prepared in advance, and attractive questions are recommended to guide the user to carry out data analysis. For example, for operation users, business core indicators, operation frequently asked questions, and the like can be provided as recommended questions to guide operation users to carry out data analysis.

[0064] Step a2, in response to the determined first question, generating a first data analysis instruction.

[0065] The multi-language model-based conversational data analysis method provided in the embodiment can display a target data source and corresponding recommended questions on a data analysis page, and determine a first question by selecting a recommended question. Therefore, guidance for data analysis can be provided, the difficulty of carrying out data analysis is reduced, the accuracy of input questions is improved, and the accuracy of data analysis is improved.

[0066] Exemplarily, as shown in Figure 4 the target data source is a public data set, recommended question 1, recommended question 2, and recommended question 3 can be provided based on the public data set. The selected recommended question is determined as the first question, and a first data analysis instruction is generated based on the first question.

[0067] In some optional embodiments, the multi-language model-based conversational data analysis method of the present disclosure further includes:

[0068] Step b1, in response to an interactive operation on the target data source, displaying an optional data set, the optional data set including at least one of a public data set and other permission data sets having access permissions.

[0069] Optionally, the optional data set can further include other data sets without access permissions, and the data analysis page displays a permission application control in a region corresponding to the other data sets, the permission application control being used to apply for access permissions of the corresponding other data sets. Alternatively, in response to an interactive operation on the other data sets, an application instruction is generated, the application instruction being used to apply for access permissions of the other data sets.

[0070] Step b2, in response to the selection instruction for the optional data set, updating the target data source with the selected optional data set.

[0071] Specifically, in response to the selection operation for the optional data set, a selection instruction is generated, the selection instruction being used to determine the selected optional data set. The target data source is updated to the selected optional data set.

[0072] It is worth noting that in order to protect data security, the data of the data analysis platform needs to be opened for use, thus leading to the fact that some users are difficult to quickly carry out data analysis work due to unfamiliarity with the data analysis process. Therefore, providing a bottomed public data set can guide users to quickly carry out data analysis and improve data analysis efficiency.

[0073] The method for dialog data analysis based on a multi-language model provided in the embodiment displays at least one of the optional data set of the public data set and other permission data sets with access permission. The target data source is updated with the selected optional data set. Therefore, on the one hand, the public data set can be provided as the bottomed data for data analysis, avoiding the situation that the analysis result cannot be generated, and improving the data analysis effect. On the other hand, the optional data set is provided for switching the target data source, which can explore data from different sources, adapt to changes in data analysis requirements, and improve the generalization ability of data analysis.

[0074] Exemplarily, as shown in Figure 5 , it is assumed that the optional data set includes a public data set, a permission data set 1, and other data sets without access permission. The other data sets without access permission and the accessible public data set and permission data set can be distinguished by the prompt words "the following are the data sets that may be of interest but have no access permission" and "according to your role, the system has preset the following data sets for you". As shown in Figure 5 , it is assumed that the selected optional data set is the public data set, and the public data set is displayed in the area corresponding to the target data source, prompting the user that the current data set used is the public data set.

[0075] Exemplarily, as shown in Figure 6 , in the case where the number of selected optional data sets is greater than 1, the data analysis page displays the number of selected data sets in the area corresponding to the target data source, for example, the public data set and the permission data set 1 are selected as the target data source, and 2 data sets are displayed. At the same time, the target data source switching data set will be explicitly prompted in the dialog flow. Specifically, a prompt word for indicating that the target data source has been successfully updated is displayed in the data analysis page, for example, "the target data source has been switched to the specified 2 data sets".

[0076] In some optional embodiments, the data analysis page further displays a new theme control (see Figure 5 andFigure 6 As shown in the data analysis theme), a new theme control is used to add a data analysis theme.

[0077] In some optional embodiments, the first data analysis instruction is generated in response to a first question determined in the data analysis page in the step S103, including:

[0078] Step c1, in response to the input question in the input control of the data analysis page, determining an initial question.

[0079] For example, the input question in the input control of the data analysis page is “How much is the sales?”, and the initial question is determined as “How much is the sales?”.

[0080] Step c2, in response to the optimization instruction for the initial question, determining the first question.

[0081] In some optional embodiments, as shown in the data analysis page, the area where the input control is located displays an optimization control, and the optimization control is used to optimize the initial question to determine the first question. Figure 7

[0082] Optionally, the step c2 includes: in response to an interaction operation on the optimization control, generating an optimization instruction to determine the first question.

[0083] For example, as shown in the data analysis page, the initial question is optimized to “How much is the total sales today?”, so as to improve the accuracy of the first question. Figure 7

[0084] In some optional embodiments, as shown in the data analysis page, the area where the input control is located displays optional data corresponding to the target data in the initial question, and the initial question can be adjusted according to the optional data to determine the first question. Figure 8

[0085] Optionally, the step c2 includes: in response to the input question in the input control of the data analysis page, determining the target data in the initial question; displaying the optional data corresponding to the target data in the area corresponding to the target data in the data analysis page; and in response to a selection instruction for the optional data, updating the target data based on the selected optional data.

[0086] Specifically, the target data is data satisfying a preset condition, for example, data having a time attribute.

[0087] For example, as shown in the data analysis page, the initial question is optimized to “How much is the total sales today?”, so as to improve the accuracy of the first question. Figure 8 ​​​As shown, assuming that the initial question is "Number of stores opened in the last 1 week, by opening time", the data analysis page provides selectable data "last 2 weeks", "last 30 days", and "today" for "last 1 week". If "last 30 days" is selected, the first question is determined as "Number of stores opened in the last 30 days, by opening time".

[0088] In some optional embodiments, as shown, the data analysis page displays recommended questions, such as recommended question 1 and recommended question 2, in the area corresponding to the initial question. The user can select a recommended question to update the initial question to determine the first question. Figure 9

[0089] Optionally, the step c2 includes: displaying the recommended question on the data analysis page; and generating the optimization instruction to determine the first question in response to a selection instruction for the recommended question.

[0090] Step c3, generating the first data analysis instruction in response to the determined first question.

[0091] The method for conversational data analysis based on a multi-language model provided in the embodiments can determine the first question in response to the optimization instruction for the input initial question. Then, the first data analysis instruction is generated based on the first question. Therefore, the accuracy of the first question can be improved, and thus the accuracy of the data analysis can be further improved.

[0092] It is worth noting that in the method for conversational data analysis based on a multi-language model, the predicted recommended question can be displayed for determining the first question. Meanwhile, the keywords (i.e., selectable data) are predicted and recommended in combination with the operation data, etc. in response to the input operation of the user in the input control of the data analysis page, to determine the first question. In addition, the input question can be optimized through the optimization control to determine the first question. Therefore, the user can be guided to describe the question in natural language, and the accuracy of the data analysis can be improved.

[0093] In some optional embodiments, in the process of determining the first analysis result, the data analysis process of the target language model for the first question is displayed on the data analysis page.

[0094] Specifically, in the process of determining the first analysis result of the target language model for the first question, the data analysis process of the target language model for the first question can be displayed. The data analysis process includes at least one level of data analysis elements. The at least one level of data analysis elements includes at least one of searching for a specified data set, searching for a specified dimension, searching for a specified indicator, and searching for a filtering item. Therefore, the visualization of the data analysis process can be achieved.

[0095] Exemplarily, as shown, Figure 10 ​As shown, the first question is "Target product 3-month sales data, statistical trend, and performance in line chart", the target language model is the language model corresponding to the analysis control 1, and the data analysis process of the target language model for the first question is displayed in the process of determining the first analysis result of the target language model for the first question. For example, the search specified dataset is "real-time closed-loop order wide table", the search specified dimension is "target product", the search specified index is "year-on-year growth rate", and the search filter is "sales time from March 1 to March 31". The state of each data analysis element corresponding to the data analysis link is displayed. At the same time, the data analysis page corresponding to the input control area is displayed with a pause control. The data analysis of the target language model for the first question can be paused in response to the interactive operation of the pause control. Moreover, the modified data analysis element can be determined in response to the modification operation of the data analysis element. Then, the data analysis is performed again according to the modified data analysis element to obtain the corresponding analysis result.

[0096] The method for conversational data analysis based on multiple language models provided in the embodiment displays the data analysis process of the target language model for the first question on the data analysis page. Therefore, the visualization of the data analysis process can be realized, and the experience of the data analysis service can be improved. At the same time, the data analysis element with a problem in the data analysis process can be easily modified to further improve the accuracy of the data analysis result.

[0097] It should be noted that after the user selects or inputs the question to determine the first question, the data analysis process of the target language model for the first question is entered. The data analysis process of the target language model is displayed through a dynamic effect in the data display page. Then, the data analysis process is presented in the form of information visualization. In the data analysis process, the user can also interrupt the data analysis process to modify the data analysis element to re-analyze the first question.

[0098] In some optional embodiments, the data analysis page corresponding to the area of the first analysis result also displays a view control, and the view control is used to view the data analysis process of the first analysis result, and the data analysis process includes at least one level of data analysis element.

[0099] It should be noted that the data analysis process of the first analysis result is consistent with the data analysis process of the first question.

[0100] The multi-language model-based conversational data analysis method provided in the embodiment can display the viewing control for viewing the data analysis process of the first analysis result in the region of the data analysis page corresponding to the first analysis result. The data analysis process includes at least one level of data analysis elements. Therefore, the data analysis details of the first analysis result can be understood, and the data analysis element with an abnormality in the data analysis process can be modified to adjust the first analysis result and improve the accuracy of data analysis.

[0101] In some optional embodiments, the multi-language model-based conversational data analysis method provided in the disclosure further includes:

[0102] Step d1, in response to the interactive operation on the viewing control, displaying at least one level of data analysis elements on the detail page.

[0103] Specifically, the at least one level of data analysis elements includes at least one of the proposed question, the search of the specified data set, the search of the specified dimension, the search of the specified index, the execution of the filtering condition, and the written structured query language database (see Figure 11 SQL written in the figure).

[0104] Step d2, in response to the modification instruction for the at least one level of data analysis elements, determining the modified data analysis element.

[0105] Specifically, the detail page displays the modification control in the region corresponding to the data analysis element. In response to the interactive operation on the modification control, the modification page is displayed. In response to the modification operation in the modification page, the modification instruction corresponding to the data analysis element is generated, and the modification instruction is used to determine the modified data analysis element.

[0106] Step d3, displaying the analysis result corresponding to the modified data analysis element on the data analysis page.

[0107] Specifically, in response to the modified data analysis element, the analysis result corresponding to the modified data analysis element is determined, and the analysis result is displayed on the data analysis page.

[0108] Specifically, the detail page can be displayed on one side of the data analysis page in the form of a drawer, a pop-up window, or the like.

[0109] The multi-language model-based conversational data analysis method provided in the embodiment displays at least one level of data analysis elements in the data analysis process on the detail page. Therefore, the visualization of the data analysis process can be realized, and the experience of data analysis service can be improved. Meanwhile, the data analysis element with a problem in the data analysis process can be modified to further improve the accuracy of the data analysis result.

[0110] For example, such as Figure 11 As shown, the first analysis result includes Answer 1 and Answer 2, and prompts "Different answers were found in multiple specified data tables. Please select the answer you think is correct." Answer 1 is the product details data for Group A, showing "Total sales volume in the past three months is 70 million. Product 1's sales volume in January, February, and March is 2367. Product 2's sales volume in January, February, and March is 325." Answer 2 is the operating details data for the main products, showing "Total sales volume in the past three months is 68 million. Product 1's sales volume in January is 4567, and its sales volume in February and March is 2367. Product 2's sales volume in January, February, and March is 325." Figure 11 (The data shown is for illustrative purposes only.) It's clear that the two answers show different sales volumes for Product 1 in January. The area on the data analysis page corresponding to the first analysis result displays viewing controls for reviewing the data analysis process (see [link]). Figure 11 (As shown, view the complete thought process). In response to interactive actions on this view control, a view command is generated to display the details page. The details page displays the data analysis process for both answers. Users can view the data analysis process for both answers, or modify the data analysis elements in the data analysis process to adjust the first analysis result for the first question.

[0111] In some alternative implementations, the details page also displays at least one first display control, which is used to determine how at least one level of data analysis elements are displayed on the details page.

[0112] Specifically, the first display control corresponds to a display mode, and the display mode of the data analysis elements can be switched by selecting the first display control. For example, the data analysis elements can be displayed using a form or a tree diagram display mode.

[0113] The conversational data analysis method based on a multilingual model provided in this embodiment features a first display control on the details page to adjust the display method of data analysis elements. Therefore, it allows for flexible adjustment of the display method of data analysis elements, better showcasing the analytical thinking represented by the data analysis process.

[0114] For example, such as Figure 12As shown, the data analysis elements can be displayed in a tree diagram display mode through the first display control. For example, from the node of "start" to the node of "understand the problem intention", and then, the node of "strategy 1" is executed. The node of "strategy 1" branches out the node of "subquery 1" and the node of "subquery 2". The node of "subquery 1" is provided with the node of "search for a specified data set". The node of "search for a specified data set" is provided with the node of "search for a specified dimension". The node of "search for a specified dimension" is provided with the node of "search for a specified index". The node of "search for a specified index" is provided with the node of "execute a filtering condition". Further, the node of "first analysis result" is reached. The node of "subquery 2" directly reaches the node of "first analysis result". The edges between the nodes of the various data analysis elements display the data analysis time between the data analysis elements. For example, the data analysis time from the node of "search for a specified data set" to the node of "search for a specified dimension" is 1.3s.

[0115] Understandably, for the question that has been answered, the user can click the view control to display the data analysis process, view the overall idea of data analysis, or make secondary modifications to the data analysis elements.

[0116] In some optional embodiments, the multi-language model-based conversational data analysis method of the present disclosure further comprises:

[0117] Step e1, in response to a selection instruction for the first data in the first analysis result and / or the second data in the historical analysis result, determining a target analysis object.

[0118] Specifically, if the first data in the first analysis result is selected, the first data is determined as the target analysis object. If the second data in the historical analysis result is selected, the second data is determined as the target analysis object.

[0119] Specifically, for the data in the first analysis result and / or the historical analysis result, the data corresponding to the columns, rows, and cells can be flexibly selected as the target analysis object. For example, Figure 13 As shown, the data of product 1 and product 2 from January to March in the local merged data is 2367. According to actual needs, the data of March in the 4th column in the local merged data can be selected as the target analysis object.

[0120] Specifically, the historical analysis records can be viewed by scrolling up or clicking the historical record control on the data analysis page to select the target analysis object from the historical analysis records. For example, Figure 14 As shown, the data of product 1 from 3.10 to 3.24 in the sales data of March in the historical record 1 is selected as the target analysis object.

[0121] Further, the target analysis object is displayed in the data analysis page, for example, the target analysis object is displayed in the input control of the data analysis page, so as to facilitate the secondary query and analysis for the target analysis object.

[0122] Step e2, in response to the second question input in the input control of the data analysis page, a second data analysis instruction is generated, and the second data analysis instruction is used to determine a second analysis result of the target language model for the second question and the target analysis object.

[0123] Specifically, the second analysis result is obtained by data analysis of the target analysis object based on the target language model and the second question.

[0124] Step e3, the second analysis result is displayed on the data analysis page.

[0125] The method for conversational data analysis based on a multi-language model provided in the embodiment selects the first data in the first analysis result and / or the second data in the historical analysis result as the target analysis object. Then, a second data analysis instruction is generated based on the second question input in the input control of the data analysis page, so as to perform data analysis on the target analysis object and obtain a second analysis result. Therefore, data analysis can be performed across data, and the flexibility of data analysis is improved.

[0126] In some optional embodiments, an upload control is displayed on the data analysis page, and the upload control is used to upload target content. The method for conversational data analysis based on a multi-language model provided in the disclosure further includes:

[0127] Step f1, in response to an interactive operation on the upload control, the target content is determined.

[0128] Specifically, in response to the interactive operation of the upload control, the uploaded content is determined as the target content.

[0129] Exemplarily, as shown in Figure 15 The local data 1 and the local data 2 can be uploaded through the upload control of the data analysis page. Therefore, data analysis can be performed based on the local data 1 and the local data 2, so as to obtain the data analysis result.

[0130] Step f2, in response to a third question input in the input control of the data analysis page, a third data analysis instruction is generated, and the third data analysis instruction is used to determine a third analysis result of the target language model for the third question and the target content.

[0131] Specifically, the third analysis result is obtained by data analysis of the target content based on the target language model and the third question.

[0132] Step f3, the third analysis result is displayed on the data analysis page.

[0133] The method for conversational data analysis based on a multi-language model provided in this embodiment has an upload control displayed on a data analysis page. The upload control is used to upload target content. Therefore, local data upload can be supported by using the upload control, so that data analysis can be performed using local data, thereby improving the flexibility of data analysis.

[0134] It can be understood that the user can perform cross-data merging query and processing on the analysis result given by the language model, and local file upload is also supported. Therefore, the flexibility of data analysis can be improved.

[0135] In some optional embodiments, the data analysis page also displays a save control in the area corresponding to the first analysis result. The save control is used to store the first question and / or the first analysis result.

[0136] Specifically, the user can store the first question alone or store the first question and the first analysis result at the same time through the save control. In addition, the first analysis result can also be stored alone according to actual needs.

[0137] The method for conversational data analysis based on a multi-language model provided in this embodiment displays a save control to store the first question and / or the first analysis result, so that subsequent query of data analysis results and further data analysis can be facilitated.

[0138] In some optional embodiments, the method for conversational data analysis based on a multi-language model of the present disclosure further includes:

[0139] Step g1, in response to an interactive operation on the save control, a task subscription page is displayed. The task subscription page is used to store the first question to obtain a target subscription task of the first question.

[0140] Specifically, as shown in Figure 16 The task subscription page can be displayed through a display mode such as a floating layer or a pop-up window. The task subscription page displays at least one configuration item of a subscription question, an update rule, and a deadline of the subscription task. The subscription question can be configured to store the first question. In response to a configuration operation on the configuration item in the task subscription page, a target subscription task corresponding to the first question is determined. The user can save the target subscription task or skip to directly save the first question.

[0141] Step g2, in response to a fourth analysis result fed back for the target subscription task, push content is generated.

[0142] Specifically, assuming that the update rule of the target subscription task is monthly update and the deadline is never ending, a second analysis result of the target subscription task is obtained every month, and push content is generated based on the second analysis result.

[0143] Step g3, pushing the push content to the target push object.

[0144] Specifically, the corresponding push content can be pushed to the target push object in the form of an email, a document, a message card, etc.

[0145] Exemplarily, as shown in Figure 17 , the corresponding push content can be pushed to the target push object in the form of a message card.

[0146] Understandably, in the multi-language model-based conversational data analysis method of the present disclosure, the configuration items required for the subscription task corresponding to the first question and the first analysis result can be pre-configured, and the first question and the first analysis result can be subscribed according to the configuration items.

[0147] The multi-language model-based conversational data analysis method provided in the embodiment can display a task subscription page for storing the first question and obtaining the target subscription task corresponding to the first question through the interactive operation of the save control. Then, the push content corresponding to the target subscription task is pushed to the target push object. Therefore, the temporary query question can be converted into the need of routine query, and the data query efficiency of the historical query question is improved. At the same time, the first question and the first analysis result can be stored for a long time for the next data analysis.

[0148] In some optional embodiments, the data analysis page includes a first display area for displaying the first question and the first analysis result, and a second display area for displaying the storage result corresponding to the save control.

[0149] Specifically, the first question and the first analysis result are displayed in the form of a conversation in the first display area. At the same time, the first display area also displays the data analysis theme, the target data source, the data analysis process, etc. of the data analysis page. The stored question, the analysis result, etc. of the storage result are displayed in the second display area. Please refer to the “My Space” part in the drawing. At the same time, the second display area also displays the analysis control.

[0150] The multi-language model-based conversational data analysis method provided in the embodiment divides the data analysis page into a first display area for displaying the first question and the first analysis result, and a second display area for displaying the storage result corresponding to the save control. Therefore, the storage result can be used for data analysis, and the flexibility of data analysis is improved.

[0151] In some optional embodiments, the data method of the present disclosure further includes: in response to a selection instruction for the storage result, displaying summary content of the selected storage result in the first display area.

[0152] Specifically, as shown in Figure 18 The user can select the stored result displayed in the second display area, display the summary content of the selected stored result in the first display area, and realize data analysis of the collected query question and analysis result, so as to improve the query efficiency of the collected query question and analysis result.

[0153] The method for dialog data analysis based on a multi-language model provided in this embodiment displays the summary content of the selected stored result in the first display area in the case of selecting the stored result. Therefore, the query efficiency of the collected query question and analysis result can be improved.

[0154] In some optional embodiments, the second display area displays at least one second display control, and the second display control is used to determine the display mode of the stored result in the second display area.

[0155] Specifically, each display control corresponds to a display mode, such as a form, a slide, a video, and the like. The user can switch the display mode of the stored result through the display control. For example, Figure 19 As shown in the figure, the display mode of the slide can be used to display each stored result.

[0156] It should be noted that the display mode of the video can generate a video of the stored result in the form of automatically playing the slide, so as to automatically play the stored result.

[0157] The method for dialog data analysis based on a multi-language model provided in this embodiment displays at least one second display control in the second display area, so the display mode of the stored result can be flexibly adjusted, and the display effect of the data analysis result can be improved.

[0158] In some optional embodiments, the second display area corresponds to the area of the stored result, and further displays a download control to download the stored result.

[0159] In some optional embodiments, the second display area corresponds to the area of the stored result, and further displays a sharing control to share the stored result.

[0160] In some optional embodiments, the data analysis page corresponds to the area of the target analysis result, and further displays a download control to download the first question and the target analysis result.

[0161] In some optional embodiments, the data analysis page corresponds to the area of the target analysis result, and further displays a sharing control to share the first question and the target analysis result.

[0162] As a specific application example, a target program is installed on the data analysis platform, which is used to perform data analysis by using the multi-language model-based conversational data analysis method of the present disclosure. A user can select a language model suitable for the current data analysis scene through the data analysis platform to perform data analysis, so as to improve the accuracy of the data analysis result.

[0163] It is worth noting that in the multi-language model-based conversational data analysis method of the present disclosure, language models suitable for different data analysis scenes are provided through the analysis control for users to select. Therefore, a suitable language model can be selected for data analysis according to the data analysis scene requirements, so as to be compatible with multiple data analysis scenes. In addition, multiple language models can also be used to provide natural language conversational analysis in one data analysis scene. The multi-language model-based conversational data analysis method of the present disclosure can be performed by multiple language models of subdivided scenes individually or in combination, so as to realize flexible switching of data analysis scenes. At the same time, through continuous questioning and collection in the multi-language model-based conversational data analysis method of the present disclosure, a personalized data analysis platform is created, which can realize the sedimentation, analysis, collaboration and reporting of analysis assets.

[0164] In some optional embodiments, the data analysis page displays an exploration control in a region corresponding to the first analysis result. The multi-language model-based conversational data analysis method of the present disclosure further includes: in response to an interactive operation on the exploration control, displaying a visualization canvas of the first analysis result on an exploration analysis page to obtain an exploration analysis result of the first analysis result, the visualization canvas being used to provide a data processing environment of the first analysis result; and in response to an interactive operation on a return control in the exploration analysis page, displaying the exploration analysis result on the data analysis page.

[0165] In some optional embodiments, the exploration analysis page further displays an upload control, and the upload control of the exploration analysis page is used to upload analysis content. The multi-language model-based conversational data analysis method of the present disclosure further includes: in response to an interactive operation on the upload control, displaying the analysis content on the visualization canvas; and in response to an association operation on the first analysis result and the analysis content, associating the first analysis result and the analysis content to obtain the exploration analysis result.

[0166] In some optional embodiments, the first analysis result includes first analysis data of at least one hierarchical dimension. The multi-language model-based conversational data analysis method of the present disclosure further includes: in response to a disassembly instruction on the first analysis data in the visualization canvas, obtaining second analysis data disassembled from the first analysis data to obtain the exploration analysis result, the exploration analysis result being obtained based on the first analysis data and the second analysis data.

[0167] In some optional implementations, the exploration analysis page also displays analysis controls used to invoke a dialog page corresponding to the language model linked to the visualization canvas. The multilingual model-based conversational data analysis method disclosed herein further includes: displaying a dialog page in response to an interactive operation on the second analysis control; and generating an operation instruction in response to a fourth question entered on the dialog page. The operation instruction is generated based on the language model linked to the visualization canvas and the fourth question, and is used to perform a target operation on the first analysis result in the visualization canvas.

[0168] Specifically, the target operations include at least one of the following: adding a data node for the first analysis result, explaining the first analysis result or the first analysis data, breaking down the first analysis data, or summarizing the content in the visualization canvas. In addition, other operations on the first analysis result, target content, first analysis data, or second analysis data in the visualization canvas are also included, without limitation. Therefore, it is possible to achieve linkage between the language model and the visualization canvas.

[0169] It should be noted that the analysis controls in the Explore Analysis page can be the same as those in the Data Analysis page. That is, the language model used in conjunction with the Explore Analysis page can be the same as that used in conjunction with the Data Analysis page, or it can be another language model; there are no restrictions here.

[0170] It's worth noting that the data analysis page displays exploration controls in the area corresponding to the first analysis result. Interacting with these controls displays a visual canvas of the first analysis result on the exploration analysis page, providing an exploratory analysis of that result. This visual canvas offers a data processing environment for the first analysis result. Therefore, if the first analysis result doesn't meet expectations, the exploration controls and visual canvas allow for further exploration and verification, leading to a new exploratory analysis result. Then, responding to interactions with the return control on the exploration analysis page, the exploratory analysis result is displayed on the data analysis page, synchronizing the data analysis results between the conversational interaction model and the exploratory analysis mode. This effectively improves the accuracy of the data analysis results.

[0171] This embodiment also provides a conversational data analysis device based on a multilingual model, which is used to implement the above embodiments and preferred embodiments; details already described will not be repeated. As used below, the term "module" can be a combination of software and / or hardware that implements a predetermined function. Although the device described in the following embodiments is preferably implemented in software, hardware implementation, or a combination of software and hardware, is also possible and contemplated.

[0172] This embodiment provides a conversational data analysis device based on a multilingual model, such as...Figure 20 As shown, comprising:

[0173] The first display module 2001 is configured to display a data analysis page, and the data analysis page displays an analysis control, and the analysis control corresponds to at least one language model;

[0174] The model selection module 2002 is configured to determine a target language model in response to a selection instruction for the analysis control;

[0175] The first analysis module 2003 is configured to generate a first data analysis instruction in response to a first question determined on the data analysis page, and the first data analysis instruction is used to determine a first analysis result of the target language model for the first question;

[0176] The second display module 2004 is configured to display the first analysis result on the data analysis page.

[0177] In some optional embodiments, the data analysis page further displays a target data source for data analysis and a recommended question corresponding to the target data source. Then, the first analysis module 2003 comprises:

[0178] The question selection unit is configured to determine the first question in response to a selection instruction for the recommended question;

[0179] The first analysis unit is configured to generate the first data analysis instruction in response to the determined first question.

[0180] In some optional embodiments, the multi-language model-based conversational data analysis apparatus of the present disclosure further comprises:

[0181] The third display module is configured to display selectable data sets in response to an interactive operation for the target data source, and the selectable data sets comprise at least one of a public data set and other permission data sets with access authority;

[0182] The data source selection module is configured to update the target data source by using a selected selectable data set in response to a selection instruction for the selectable data set.

[0183] In some other optional embodiments, the first analysis module 2003 comprises:

[0184] The question determination unit is configured to determine an initial question in response to an input question in an input control of the data analysis page;

[0185] The question optimization unit is configured to determine the first question in response to an optimization instruction for the initial question;

[0186] The second analysis unit is configured to generate the first data analysis instruction in response to the determined first question.

[0187] In some optional embodiments, in the process of determining the first analysis result, the data analysis page displays the data analysis process of the target language model for the first question.

[0188] In some optional embodiments, the region of the data analysis page corresponding to the first analysis result further displays a viewing control for viewing the data analysis process of the first analysis result, the data analysis process including at least one level of data analysis elements.

[0189] In some optional embodiments, the multi-language model-based conversational data analysis apparatus of the present disclosure further includes:

[0190] A fourth display module for displaying the at least one level of data analysis elements on the detail page in response to an interactive operation on the viewing control;

[0191] A data modification module for determining modified data analysis elements in response to a modification instruction on the at least one level of data analysis elements;

[0192] A fifth display module for displaying an analysis result corresponding to the modified data analysis elements on the data analysis page.

[0193] In some optional embodiments, the detail page further displays at least one first display control for determining a display manner of the at least one level of data analysis elements on the detail page.

[0194] In some optional embodiments, the multi-language model-based conversational data analysis apparatus of the present disclosure further includes:

[0195] An analysis data selection module for determining a target analysis object in response to a selection instruction on first data in the first analysis result and / or second data in the historical analysis result;

[0196] A second analysis module for generating a second data analysis instruction in response to a second question input in the input control of the data analysis page, the second data analysis instruction being used to determine a second analysis result of the target language model for the second question and the target analysis object;

[0197] A sixth display module for displaying the second analysis result on the data analysis page.

[0198] In some optional embodiments, the data analysis page displays an upload control for uploading target content. Then, the multi-language model-based conversational data analysis apparatus of the present disclosure further includes:

[0199] A data upload module for determining the target content in response to an interactive operation on the upload control;

[0200] The third analysis module is configured to generate a third data analysis instruction in response to a third question input in the input control of the data analysis page, and the third data analysis instruction is used to determine a third analysis result of the target language model for the third question and the target content.

[0201] The seventh display module is configured to display the third analysis result on the data analysis page.

[0202] In some optional embodiments, the data analysis page corresponding to the region of the first analysis result further displays a save control, and the save control is used to store the first question and / or the first analysis result.

[0203] In some optional embodiments, the multi-language model-based conversational data analysis apparatus of the present disclosure further comprises:

[0204] The task subscription module is configured to display a task subscription page in response to an interaction operation on the save control, and the task subscription page is used to store the first question to obtain a target subscription task of the first question.

[0205] The push generation module is configured to generate a push content in response to a fourth analysis result fed back for the target subscription task.

[0206] The data push module is configured to push the push content to a target push object.

[0207] In some optional embodiments, the data analysis page comprises a first display region used to display the first question and the first analysis result, and a second display region used to display a storage result corresponding to the save control.

[0208] In some optional embodiments, the multi-language model-based conversational data analysis apparatus of the present disclosure further comprises:

[0209] The data summary module is configured to display a summary content of the selected storage result in the first display region in response to a selection instruction for the storage result.

[0210] In some optional embodiments, the second display region displays at least one second display control, and the second display control is used to determine a display manner of the storage result in the second display region.

[0211] The further function descriptions of the above-mentioned various modules and units are the same as those of the above-mentioned corresponding embodiments, and will not be described here.

[0212] In this embodiment, the conversational data analysis device based on a multilingual model is presented in the form of functional units. Here, a unit refers to an ASIC (Application Specific Integrated Circuit) circuit, a processor and memory that execute one or more software or fixed programs, and / or other devices that can provide the above functions.

[0213] This disclosure also provides an electronic device having the above-described features. Figure 20 The diagram shows a conversational data analysis device based on a multilingual model.

[0214] Please see Figure 21 , Figure 21 This is a structural block diagram of an electronic device provided in an optional embodiment of this disclosure, such as... Figure 21 As shown, the electronic device includes one or more processors 2101, a memory 2102, and interfaces for connecting the components, including high-speed interfaces and low-speed interfaces. The components communicate with each other via different buses and can be mounted on a common motherboard or otherwise as required. The processors can process instructions executed within the electronic device, including instructions stored in or on memory to display graphical information of a GUI on external input / output devices (such as display devices coupled to the interfaces). In some alternative implementations, multiple processors and / or multiple buses can be used with multiple memories and multiple memory modules, if desired. Similarly, multiple electronic devices can be connected, each providing some of the necessary operations (e.g., as a server array, a group of blade servers, or a multiprocessor system). Figure 21 Take a processor 2101 as an example.

[0215] Processor 2101 may be a central processing unit, a network processor, or a combination thereof. Processor 2101 may further include a hardware chip. The hardware chip may be an application-specific integrated circuit (ASIC), a programmable logic device (PLD), or a combination thereof. The programmable logic device may be a complex programmable logic device (CAMP), a field-programmable gate array (FPGA), a general-purpose array logic (GPA), or any combination thereof.

[0216] The memory 2102 stores instructions executable by at least one processor 2101 to cause at least one processor 2101 to perform the method shown in the above embodiments.

[0217] The memory 2102 may include a program storage area and a data storage area. The program storage area may store the operating system and applications required for at least one function; the data storage area may store data created based on the use of the electronic device. Furthermore, the memory 2102 may include high-speed random access memory and may also include non-transitory memory, such as at least one disk storage device, flash memory device, or other non-transitory solid-state storage device. In some alternative embodiments, the memory 2102 may optionally include memory remotely located relative to the processor 2101, and these remote memories can be connected to the electronic device via a network. Examples of such networks include, but are not limited to, the Internet, intranets, local area networks, mobile communication networks, and combinations thereof.

[0218] The memory 2102 may include volatile memory, such as random access memory; the memory may also include non-volatile memory, such as flash memory, hard disk or solid-state drive; the memory 2102 may also include a combination of the above types of memory.

[0219] The electronic device also includes an input device 2103 and an output device 2104. The processor 2101, memory 2102, input device 2103, and output device 2104 can be connected via a bus or other means. Figure 21 Taking the example of a connection between China and Israel via a bus.

[0220] Input device 2103 can receive input numerical or character information, and generate key signal inputs related to user settings and function control of the electronic device, such as a touch screen, keypad, mouse, trackpad, touchpad, joystick, one or more mouse buttons, trackball, joystick, etc. Output device 2104 may include display devices, auxiliary lighting devices (e.g., LEDs), and haptic feedback devices (e.g., vibration motors). The aforementioned display devices include, but are not limited to, liquid crystal displays, light-emitting diodes, displays, and plasma displays. In some alternative embodiments, the display device may be a touch screen.

[0221] The embodiments of the present disclosure further provide a computer readable storage medium, and the method according to the embodiments of the present disclosure can be implemented in hardware, firmware, or recorded in a storage medium, or be implemented as computer code originally stored in a remote storage medium or a non-transitory machine readable storage medium and downloaded through a network and stored in a local storage medium, so that the method described herein can be processed by such software on a storage medium using a general purpose computer, a special purpose processor, or programmable or special purpose hardware. Among them, the storage medium can be a magnetic disk, an optical disk, a read-only memory, a random access memory, a flash memory, a hard disk or a solid state disk, etc.; further, the storage medium can also include a combination of the above types of memories. It can be understood that the computer, processor, microprocessor controller or programmable hardware includes a storage component that can store or receive software or computer code, when the software or computer code is accessed and executed by the computer, processor or hardware, the method shown in the above embodiments is implemented.

[0222] Part of the present disclosure can be applied as a computer program product, for example, computer program instructions, when executed by a computer, through the operation of the computer, the method and / or technical solutions according to the present disclosure can be called or provided. Those skilled in the art should understand that the form of computer program instructions in computer readable medium includes but is not limited to source file, executable file, installation package file, etc., accordingly, the way of computer program instructions executed by computer includes but is not limited to: the computer directly executes the instructions, or the computer compiles the instructions and then executes the corresponding compiled program, or the computer reads and executes the instructions, or the computer reads and installs the instructions and then executes the corresponding installed program. Here, the computer readable medium can be any available computer readable storage medium or communication medium accessible to the computer.

[0223] Although the embodiments of the present disclosure are described in conjunction with the accompanying drawings, various modifications and changes can be made by those skilled in the art without departing from the spirit and scope of the present disclosure, and such modifications and changes fall within the scope defined by the appended claims.

Claims

1. A multi-language model based conversational data analysis method, characterized in that, The method comprises: displaying a data analysis page, the data analysis page being provided with an analysis control corresponding to at least one language model, and the analysis control being provided with model information of the language model corresponding to the analysis control in a region corresponding to the analysis control; in response to a selection instruction for the analysis control, determining a target language model; in response to a first question determined on the data analysis page, generating a first data analysis instruction for determining a first analysis result of the target language model for the first question, if the selected analysis control corresponds to multiple language models, the multiple language models are taken as target language models, and the first question is answered by the multiple target language models simultaneously to obtain multiple first analysis results; displaying the first analysis result on the data analysis page, the region of the data analysis page corresponding to the first analysis result further being provided with a save control for storing the first question and the first analysis result, the data analysis page comprising a first display region for displaying the first question and the first analysis result, and a second display region for displaying a storage result corresponding to the save control, and in response to a selection instruction for the storage result, displaying summary content of the selected storage result in the first display region; wherein, in response to the first question determined on the data analysis page, the first data analysis instruction is generated, comprising: in response to an input question in an input control of the data analysis page, determining an initial question; in response to an optimization instruction for the initial question, determining the first question; in response to the optimization instruction for the initial question, the first question is determined, comprising: determining target data in the initial question, and displaying selectable data corresponding to the target data in a region of the data analysis page corresponding to the target data; in response to a selection instruction for the selectable data, updating the target data based on the selected selectable data; in response to the determined first question, the first data analysis instruction is generated.

2. The method of claim 1, wherein, The data analysis page is further provided with a target data source for data analysis and a recommended question corresponding to the target data source; in response to the first question determined on the data analysis page, the first data analysis instruction is generated, comprising: in response to a selection instruction for the recommended question, determining the first question; in response to the determined first question, the first data analysis instruction is generated.

3. The method of claim 2, wherein, The method further comprises: in response to an interactive operation for the target data source, displaying a selectable data set, the selectable data set comprising at least one of a public data set and other permission data sets having access authority; in response to a selection instruction for the selectable data set, updating the target data source by using the selected selectable data set.

4. The method of claim 1, wherein, In the process of determining the first analysis result, the data analysis page is provided with a data analysis process of the target language model for the first question.

5. The method of claim 1, wherein, The data analysis page corresponding to the region of the first analysis result also displays a viewing control for viewing a data analysis process of the first analysis result, the data analysis process including at least one hierarchical data analysis element.

6. The method of claim 5, wherein, The method further includes: in response to an interactive operation on the viewing control, displaying the at least one hierarchical data analysis element on a details page; in response to a modification instruction for the at least one hierarchical data analysis element, determining a modified data analysis element; displaying an analysis result corresponding to the modified data analysis element on the data analysis page.

7. The method of claim 6, wherein, The details page also displays at least one first display control for determining a display manner of the at least one hierarchical data analysis element on the details page.

8. The method of claim 1, wherein, The method further includes: in response to a selection instruction for first data in the first analysis result and / or second data in a historical analysis result, determining a target analysis object; in response to a second question input in an input control of the data analysis page, generating a second data analysis instruction for determining a second analysis result of the target language model for the second question and the target analysis object; displaying the second analysis result on the data analysis page.

9. The method of claim 1, wherein, The data analysis page displays an upload control for uploading target content; the method further includes: in response to an interactive operation on the upload control, determining the target content; in response to a third question input in an input control of the data analysis page, generating a third data analysis instruction for determining a third analysis result of the target language model for the third question and the target content; displaying the third analysis result on the data analysis page.

10. The method of claim 1, wherein, The method further includes: in response to an interactive operation on the save control, displaying a task subscription page for storing the first question to obtain a target subscription task of the first question; in response to a fourth analysis result fed back for the target subscription task, generating push content; pushing the push content to a target push object.

11. The method of claim 1, wherein, The second display region displays at least one second display control for determining a display manner of the storage result in the second display region. 12.A multi-language model based conversational data analysis apparatus, characterized by comprising: The apparatus includes: a first display module for displaying a data analysis page, the data analysis page displaying an analysis control, the analysis control corresponding to at least one language model, a region corresponding to the analysis control displaying model information of a language model corresponding to the analysis control; a model selection module for determining a target language model in response to a selection instruction for the analysis control; The first analysis module is configured to generate a first data analysis instruction in response to a first question determined on the data analysis page, the first data analysis instruction being used to determine a first analysis result of the target language model for the first question, and if the selected analysis control corresponds to multiple language models, the multiple language models are taken as target language models, and the first question is answered by the multiple target language models simultaneously to obtain multiple first analysis results. The second display module is configured to display the first analysis result on the data analysis page, and a save control is further displayed in a region of the data analysis page corresponding to the first analysis result, the save control being used to store the first question and the first analysis result, the data analysis page including a first display region used to display the first question and the first analysis result, and a second display region used to display a storage result corresponding to the save control, and in response to a selection instruction for the storage result, a summary content of the selected storage result is displayed on the first display region. The first analysis module includes: A question determination unit is configured to determine an initial question in response to an input question in an input control of the data analysis page. A question optimization unit is configured to determine the first question in response to an optimization instruction for the initial question, determine target data in the initial question, and display selectable data corresponding to the target data in a region of the data analysis page corresponding to the target data; and update the target data based on the selected selectable data in response to a selection instruction for the selectable data. A second analysis unit is configured to generate the first data analysis instruction in response to the determined first question.

13. An electronic device, comprising: The memory and the processor are communicatively connected, the memory stores computer instructions, and the processor executes the computer instructions to perform the dialog-based data analysis method based on multiple language models according to any one of claims 1 to 11. The computer readable storage medium stores computer instructions, and the computer instructions are used to make a computer execute the dialog-based data analysis method based on multiple language models according to any one of claims 1 to 11.

14. A computer-readable storage medium, characterized in that, The computer instructions are used to make a computer execute the dialog-based data analysis method based on multiple language models according to any one of claims 1 to 11.

15. A computer program product, characterised in that, ​

Images