Data board generation method and device, computer equipment, readable storage medium and program product

Abstract

The invention relates to a data board generation method and device, computer equipment, a computer readable storage medium and a computer program product. The method comprises the following steps: acquiring a billboard theme and user information corresponding to a current user; based on the billboard theme and the user information, determining a theme association field corresponding to the billboard theme and a field key element corresponding to the theme association field; based on the domain key elements, determining a plurality of candidate billboard indexes corresponding to the current user from a billboard index library; determining a target Kanban index from the plurality of candidate Kanban indexes based on the Kanban theme, the theme associated field and the field key elements; and obtaining index data corresponding to the target Kanban index, and generating a target data Kanban corresponding to the current user based on the target Kanban index and the index data. By adopting the method, the generation accuracy of the data board can be improved.

Description

Technical Field

[0001] The present application relates to the field of big data technology, and in particular to a data dashboard generation method, apparatus, computer equipment, computer-readable storage medium, and computer program product. Background Art

[0002] In the traditional data dashboard generation method, selectable cards are usually configured for users. Users select the required card type, enter the card title, configure the data source, configure the card display data and other parameters, generate a data card, and then display the data card in the dashboard. However, this method requires all parameters to be manually configured, and since there are many parameters to be selected, users cannot accurately select the required parameters, resulting in the problem of low accuracy in data dashboard generation. Summary of the invention

[0003] Based on this, it is necessary to provide a data dashboard generation method, apparatus, computer equipment, computer-readable storage medium and computer program product that can improve the accuracy of data dashboard generation in response to the above technical problems.

[0004] In a first aspect, the present application provides a data dashboard generation method, comprising:

[0005] Get the board topic and user information corresponding to the current user;

[0006] Based on the kanban theme and user information, determine the theme-related field corresponding to the kanban theme and the field key elements corresponding to the theme-related field;

[0007] Based on the key elements of the domain, multiple candidate dashboard indicators corresponding to the current user are determined from the dashboard indicator library;

[0008] Determine the target dashboard indicator from multiple candidate dashboard indicators based on the dashboard theme, thematic related fields and key elements of the field;

[0009] Obtain indicator data corresponding to the target dashboard indicator, and generate a target data dashboard corresponding to the current user based on the target dashboard indicator and the indicator data.

[0010] In a second aspect, the present application also provides a data dashboard generating device, comprising:

[0011] The acquisition module is used to obtain the board topic and user information corresponding to the current user;

[0012] An element determination module, used to determine the theme-related field corresponding to the theme of the board and the field key elements corresponding to the theme-related field based on the theme of the board and user information;

[0013] A candidate indicator determination module is used to determine multiple candidate kanban indicators corresponding to the current user from the kanban indicator library based on key elements of the domain;

[0014] A target indicator determination module, used for determining a target dashboard indicator from multiple candidate dashboard indicators based on the dashboard theme, the theme-related fields and the field key elements;

[0015] The dashboard generation module is used to obtain the indicator data corresponding to the target dashboard indicator, and generate the target data dashboard corresponding to the current user based on the target dashboard indicator and the indicator data.

[0016] In a third aspect, the present application further provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the following steps are implemented:

[0017] Get the board topic and user information corresponding to the current user;

[0018] Based on the kanban theme and user information, determine the theme-related field corresponding to the kanban theme and the field key elements corresponding to the theme-related field;

[0019] Based on the key elements of the domain, multiple candidate dashboard indicators corresponding to the current user are determined from the dashboard indicator library;

[0020] Determine the target dashboard indicator from multiple candidate dashboard indicators based on the dashboard theme, thematic related fields and key elements of the field;

[0021] Obtain indicator data corresponding to the target dashboard indicator, and generate a target data dashboard corresponding to the current user based on the target dashboard indicator and the indicator data.

[0022] In a fourth aspect, the present application further provides a computer-readable storage medium having a computer program stored thereon, wherein when the computer program is executed by a processor, the following steps are implemented:

[0023] Get the board topic and user information corresponding to the current user;

[0024] Based on the kanban theme and user information, determine the theme-related field corresponding to the kanban theme and the field key elements corresponding to the theme-related field;

[0025] Based on the key elements of the domain, multiple candidate dashboard indicators corresponding to the current user are determined from the dashboard indicator library;

[0026] Determine the target dashboard indicator from multiple candidate dashboard indicators based on the dashboard theme, thematic related fields and key elements of the field;

[0027] Obtain indicator data corresponding to the target dashboard indicator, and generate a target data dashboard corresponding to the current user based on the target dashboard indicator and the indicator data.

[0028] In a fifth aspect, the present application further provides a computer program product, including a computer program, which implements the following steps when executed by a processor:

[0029] Get the board topic and user information corresponding to the current user;

[0030] Based on the kanban theme and user information, determine the theme-related field corresponding to the kanban theme and the field key elements corresponding to the theme-related field;

[0031] Based on the key elements of the domain, multiple candidate dashboard indicators corresponding to the current user are determined from the dashboard indicator library;

[0032] Determine the target dashboard indicator from multiple candidate dashboard indicators based on the dashboard theme, thematic related fields and key elements of the field;

[0033] Obtain indicator data corresponding to the target dashboard indicator, and generate a target data dashboard corresponding to the current user based on the target dashboard indicator and the indicator data.

[0034] The above-mentioned data dashboard generation method, device, computer equipment, computer-readable storage medium and computer program product determine the theme-related fields corresponding to the dashboard theme and the field key elements corresponding to the theme-related fields according to the dashboard theme and user information of the current user, and determine multiple candidate dashboard indicators corresponding to the current user from the dashboard indicator library according to the field key elements, and realize the layer-by-layer refinement of the dashboard theme, theme-related fields and field key elements through multi-layer analysis, thereby improving the accuracy of the candidate dashboard indicators; determine the target dashboard indicator from multiple candidate dashboard indicators according to the dashboard theme, theme-related fields and field key elements, and involve the dashboard theme, theme-related fields and field key elements in multiple dimensions to determine the target dashboard indicator, thereby improving the accuracy of the target dashboard indicator; then generate the target data dashboard corresponding to the current user according to the target dashboard theme and the corresponding indicator data, thereby improving the generation accuracy of the data dashboard. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] In order to more clearly illustrate the technical solutions in the embodiments of the present application or related technologies, the drawings required for use in the embodiments of the present application or related technical descriptions will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying creative work.

[0036] Figure 1 An application environment diagram of a data dashboard generation method in one embodiment;

[0037] Figure 2A schematic diagram of a flow chart of a method for generating a data dashboard in one embodiment;

[0038] Figure 3 A schematic diagram of a recommended dashboard indicator in an embodiment;

[0039] Figure 4 A schematic diagram of a process flow for demand modification in an embodiment;

[0040] Figure 5 A schematic diagram of a process for optimizing indicator recommendation in one embodiment;

[0041] Figure 6 A schematic diagram of the process of data dashboard generation and post-processing in one embodiment

[0042] Figure 7 It is a structural block diagram of a data dashboard generating device in one embodiment;

[0043] Figure 8 is an internal structure diagram of a computer device in one embodiment;

[0044] Fig. 9 FIG. 4 is a diagram showing the internal structure of a computer device in one embodiment. DETAILED DESCRIPTION

[0045] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0046] The data dashboard generation method provided in the embodiment of the present application can be applied to Figure 1In the application environment shown. Among them, the terminal 102 communicates with the server 104 through the network. The data storage system can store the data that the server 104 needs to process. The data storage system can be integrated on the server 104, or it can be placed on the cloud or other network servers. The terminal 102 obtains the kanban theme and user information corresponding to the current user; the terminal 102 determines the theme-related field corresponding to the kanban theme and the field key elements corresponding to the theme-related field based on the kanban theme and user information; the terminal 102 determines multiple candidate kanban indicators corresponding to the current user from the kanban indicator library based on the field key elements; the terminal 102 determines the target kanban indicator from multiple candidate kanban indicators based on the kanban theme, the theme-related field and the field key elements; the terminal 102 obtains the indicator data corresponding to the target kanban indicator, generates the target data kanban corresponding to the current user based on the target kanban indicator and the indicator data, and the terminal 102 can store the target data kanban corresponding to the current user in the server 104. Among them, the terminal 102 can be, but is not limited to, various personal computers, laptops, smart phones, tablet computers, Internet of Things devices and portable wearable devices. The Internet of Things devices can be smart speakers, smart TVs, smart air conditioners, smart car devices, projection devices, etc. Portable wearable devices can be smart watches, smart bracelets, head-mounted devices, etc. Head-mounted devices can be virtual reality (VR) devices, augmented reality (AR) devices, smart glasses, etc. The server 104 can be an independent physical server, or a server cluster or distributed system composed of multiple physical servers, or a cloud server that provides cloud computing services.

[0047] In an exemplary embodiment, Figure 2 As shown, a data dashboard generation method is provided, which is applied to Figure 1 The terminal in the example is used for explanation. It can be understood that the method can also be applied to a server, and can also be applied to a system including a terminal and a server, and is implemented through the interaction between the terminal and the server. In this embodiment, the method includes the following steps:

[0048] Step 202, obtaining the board theme and user information corresponding to the current user.

[0049] The current user refers to the user who currently needs to generate a data dashboard. The data dashboard is a visualization page used to display specified data. The dashboard topic refers to the business subject corresponding to the business problem that the user needs to solve, which can be the business subject reflected by the question statement entered by the current user. For example, if the question statement is "how to increase the sales of the sales department", the dashboard topic can be "sales business subject". User information is information that represents the identity and role of the current user.

[0050] Exemplarily, the terminal responds to a data dashboard generation instruction and displays an information input interface. Through the information input interface, the question statement and user information input by the current user are obtained and the question statement is used as the dashboard theme. The business subject corresponding to the question statement can also be identified and the identified business subject is used as the dashboard theme.

[0051] Step 204 , based on the kanban theme and the user information, determine the theme-related field corresponding to the kanban theme and the field key elements corresponding to the theme-related field.

[0052] Among them, the subject-related field refers to the business field objectively involved in the business subject corresponding to the Kanban theme, such as the customer business field (such as "improving customer satisfaction", etc.), supplier business field (such as "supplier management", etc.), product field (such as "improving product quality", etc.) involved in the sales business subject, which can be used as business guidance information corresponding to solving the business problems raised by users through the Kanban theme. The key elements of the field refer to the solution elements of the program that can be implemented in the subject-related field to solve the business problems raised by users through the Kanban theme, such as "maintaining customer relationships" corresponding to "improving customer satisfaction", "selecting reliable suppliers" corresponding to "supplier management", "improving product quality" corresponding to "improving quality inspection requirements" and other solution execution information.

[0053] Exemplarily, the terminal obtains the kanban theme and user information input by the current user, and generates the theme-related fields according to the field keywords involved in the kanban theme. For example, the field keywords involved in "how to increase the sales of the sales department" are "sales department", "customer", "supplier", and the generated theme-related fields can be "increase the sales of each organization", "improve customer satisfaction", "supplier management" and other information. Then, the element search scope is determined according to the user role in the user information. For example, the element search scope corresponding to the manager role includes department sales, sales of each employee, etc. Then, within the element search scope, according to the key elements involved in the theme-related fields, the field key elements are generated. For example, the field key element corresponding to the theme-related field "increase the number of customers" is "maintain customer relationships", etc.

[0054] Step 206 , based on the domain key elements, determine multiple candidate dashboard indicators corresponding to the current user from the dashboard indicator library.

[0055] Among them, the candidate dashboard indicators are data attributes used to be displayed on the dashboard and represent key elements of the field. For example, the candidate dashboard indicator corresponding to the subject-related field "increasing the number of customers" is "the number of customers".

[0056] Exemplarily, the terminal matches the domain key elements corresponding to the key domain of the subject with various existing dashboard indicators in the dashboard indicator library, and uses the successfully matched dashboard indicators as multiple candidate dashboard indicators corresponding to the current user.

[0057] In an exemplary embodiment, each existing kanban indicator in the kanban indicator library is vector data after vectorization processing. After obtaining the domain key elements, the terminal vectorizes the domain key elements and determines multiple candidate kanban indicators corresponding to the current user based on the vector similarity between the processed domain key elements and each kanban indicator in the kanban indicator library.

[0058] Step 208 , based on the kanban theme, theme-related fields, and field key elements, determine a target kanban indicator from a plurality of candidate kanban indicators.

[0059] The target dashboard indicator refers to the dashboard indicator currently recommended to the user for display.

[0060] Exemplarily, an indicator recommendation model is provided in the terminal. The terminal obtains an indicator recommendation example for indicator recommendation, and the indicator recommendation example refers to example information with specific recommendation of selected dashboard indicators. For example, for the domain key element B of the subject-related domain A, the recommended indicators should be selected as indicators a, b, c, etc., which can be used to instruct the indicator recommendation model to refer to the recommended indicator recommendation example for indicator selection and recommendation. The indicator recommendation example includes reference key elements and reference recommended indicators as example explanations. The terminal merges the dashboard theme, subject-related domain, domain key elements and each candidate dashboard indicator corresponding to the current user to obtain a dashboard indicator set corresponding to the current user.

[0061] The terminal calls the indicator recommendation model, inputs the dashboard indicator set and indicator recommendation examples corresponding to the current user into the indicator recommendation model, and determines the target dashboard indicator corresponding to the current user from multiple candidate dashboard indicators in the dashboard indicator set by referring to the mapping relationship between the reference key elements and the reference recommended indicators in the indicator recommendation examples through the indicator recommendation model. After determining the target dashboard indicator, the indicator recommendation model can obtain various data display types used by the target dashboard indicator in historical time, such as graphic type, text type, etc., and determine the target display image type that can be used for the target dashboard indicator in various display graphic types. For example, the target dashboard indicator A can be displayed using the line chart type in the graphic type.

[0062] Step 210, obtaining indicator data corresponding to the target dashboard indicator, and generating a target data dashboard corresponding to the current user based on the target dashboard indicator and the indicator data.

[0063] Among them, indicator data refers to the business data that needs to be obtained according to the dashboard indicators. The target data dashboard refers to the dashboard data that the current user can view.

[0064] Exemplarily, the terminal obtains the indicator data corresponding to the target dashboard indicator, which may be to obtain the indicator data corresponding to each target dashboard indicator in the same time period, and determine the data display type of each target dashboard indicator according to the indicator type of each target dashboard indicator, such as a graphic type, a text type, etc., and convert the indicator data of each target dashboard indicator into display data according to the data display type, and generate a target data dashboard corresponding to the current user according to each target dashboard indicator and the corresponding display data according to a preset layout, and display the target data dashboard.

[0065] In the above-mentioned data dashboard generation method, the theme-related fields corresponding to the kanban theme and the field key elements corresponding to the theme-related fields are determined according to the kanban theme and user information of the current user, and multiple candidate kanban indicators corresponding to the current user are determined from the kanban indicator library according to the field key elements. Through multi-layer analysis, the kanban theme, theme-related fields and field key elements are refined layer by layer, thereby improving the accuracy of the candidate kanban indicators; the target kanban indicator is determined from multiple candidate kanban indicators according to the kanban theme, theme-related fields and field key elements, and the kanban theme, theme-related fields and field key elements are multi-dimensionally involved in determining the target kanban indicator, thereby improving the accuracy of the target kanban indicator; then, the target data kanban corresponding to the current user is generated according to the target kanban theme and the corresponding indicator data, thereby improving the generation accuracy of the data kanban.

[0066] In an exemplary embodiment, step 204, based on the kanban theme and user information, determining the theme-related field corresponding to the kanban theme and the field key elements corresponding to the theme-related field, includes:

[0067] Input the kanban theme and user information into the topic analysis model, and use the topic analysis model to search for the kanban theme based on the built-in world knowledge graph to generate the topic-related domain corresponding to the kanban theme;

[0068] Through the topic analysis model, domain elements of topic-related fields are searched based on the world knowledge graph and user information, and key domain elements corresponding to the topic-related fields are generated.

[0069] Among them, the world knowledge graph refers to a knowledge base that contains various business data and business processing cases for various business problems (such as various solutions, processing measures and other information).

[0070] Exemplarily, the terminal has a built-in world knowledge graph in the topic analysis model, which can be used as a reference data source for the topic analysis model. The topic analysis model can be a large language model. The terminal inputs the kanban topic and user information into the topic analysis model. The kanban topic can be a question sentence input by the current user. The topic analysis model identifies the question intent of the question sentence, searches for similar kanban topics in the world knowledge graph based on the question intent, obtains business problem analysis cases corresponding to similar kanban topics, and generates the topic-related field corresponding to the kanban topic based on the context information and question intent in the business problem analysis case.

[0071] Then, the topic analysis model searches for similar fields in the world image based on the topic-related fields, obtains business problem implementation cases corresponding to the similar fields, and generates key elements of each candidate field corresponding to the topic-related fields based on the context information in the business problem implementation cases. Then, based on the authority scope corresponding to the user role, the field key elements corresponding to the current user are screened from each candidate field key element to obtain the field key elements corresponding to the topic-related fields.

[0072] In this embodiment, the topic analysis model is combined with the world knowledge image to generate the topic-related fields corresponding to the billboard theme and the field key elements corresponding to the topic-related fields, thereby ensuring the accuracy of the topic-related fields and the field key elements.

[0073] In an exemplary embodiment, step 208, based on the kanban theme, the theme-related fields and the field key elements, determining the target kanban indicator from a plurality of candidate kanban indicators includes:

[0074] Based on the hierarchical relationship between the kanban theme, the theme-related fields, the key elements of the field, and multiple candidate kanban indicators, a prompt word framework corresponding to the current user is generated;

[0075] Obtaining an indicator recommendation example, embedding the indicator recommendation example into a prompt word frame, and obtaining a target prompt word frame;

[0076] The target prompt word framework is input into the indicator recommendation model, and the indicator mapping features between the reference key elements and the reference recommended indicators in the indicator recommendation example are identified through the indicator recommendation model. Based on the indicator mapping features, the subject-related fields and the field key elements, the target dashboard indicator is determined from multiple candidate dashboard indicators.

[0077] The prompt word framework refers to a set of indicators obtained by combining the kanban theme, the theme-related fields, the field key elements and multiple candidate kanban indicators according to a specified data structure. The indicator mapping feature refers to a feature that can characterize the mapping relationship between the reference key elements and the reference recommended indicators.

[0078] Exemplarily, the terminal splices the kanban theme and the theme-related field as the root node and the field key elements as the leaf nodes according to the hierarchical relationship between the kanban theme, the theme-related field, the field key elements and the multiple candidate kanban indicators, constructs a tree structure, and then splices the multiple candidate kanban indicators to the corresponding field key elements in the tree structure to obtain the prompt word framework corresponding to the current user. The prompt word framework includes hierarchical data corresponding to the kanban theme, the theme-related field, the field key elements and the multiple candidate kanban indicators. The terminal obtains the indicator recommendation example, embeds the indicator recommendation example into the corresponding hierarchical data in the prompt word framework, and obtains the target prompt word framework corresponding to the current user. Generally, the indicator recommendation example includes indicator recommendation examples for the theme-related field and the field key elements.

[0079] The terminal calls the indicator recommendation model, inputs the target prompt word frame into the indicator recommendation model, and identifies the indicator mapping features between the reference key elements and the reference recommended indicators in the indicator recommendation example through the indicator recommendation model. Then, the mapping features between the subject-related fields, the field key elements, and multiple candidate dashboard indicators are calculated, including the first mapping features between the subject-related fields and multiple candidate dashboard indicators, and the second mapping features between the field key elements and multiple candidate dashboard indicators. The comprehensive mapping features of the first mapping features and the second mapping features are calculated to obtain the comprehensive mapping features corresponding to the multiple candidate dashboard indicators. The candidate dashboard indicator corresponding to the comprehensive mapping features that achieve the indicator mapping features is determined as the target dashboard indicator.

[0080] In an exemplary embodiment, the reference key elements and the reference recommendation indicators are vectorized, the reference vector similarity between the processed reference key elements and the reference recommendation indicators is calculated, and the reference vector similarity is used as the indicator mapping feature between the reference key elements and the reference recommendation indicators. Then, the first similarity between the subject-related field in the target prompt word framework and multiple candidate dashboard indicators is calculated, and the second similarity between the field key elements and the multiple candidate dashboard indicators is calculated to obtain the first similarity and the second similarity corresponding to the multiple candidate dashboard indicators. The average of the first similarity and the second similarity is calculated to obtain the comprehensive mapping features corresponding to the multiple candidate dashboard indicators.

[0081] In this embodiment, the target dashboard indicator is determined from multiple candidate dashboard indicators by using the indicator mapping characteristics between the reference key elements and the reference recommended indicators in the indicator recommendation example as a reference basis, thereby improving the accuracy of the target dashboard indicator.

[0082] In an exemplary embodiment, the indicator recommendation example further includes an indicator display graphic example; and the data dashboard generation method further includes:

[0083] Identify the graphic mapping features between the reference recommended indicators and the corresponding indicator display graphic types in the indicator display graphic example through the indicator recommendation model, and determine the target display graphic type corresponding to the target dashboard indicator based on the graphic mapping features;

[0084] Step 210, generating a target data dashboard corresponding to the current user based on the target dashboard indicator and indicator data, including:

[0085] According to the target display graphic type of the target dashboard indicator, the corresponding indicator data is converted into target graphic data;

[0086] A target data dashboard corresponding to the current user is generated based on the target dashboard indicators and target graphic data.

[0087] Among them, the indicator display graphic example refers to example information with a specific selection of graphic types to display indicator data, such as the indicator data of dashboard indicator A should be displayed in a bar chart, etc., which can be used to indicate that the indicator recommendation model refers to the indicator display graphic example to recommend the display graphic type of the corresponding indicator data of the target dashboard indicator. The graphic mapping feature can characterize the characteristics of the mapping relationship between the reference recommended indicator and the corresponding indicator display graphic type. The target display image type refers to the indicator data graphic type that displays the target dashboard indicator in a graphical form. The target graphic data refers to the data displayed in a graphical form for the indicator data of the target dashboard indicator in the dashboard.

[0088] Exemplarily, the indicator recommendation example also includes an indicator display graphic example. The terminal inputs the target prompt word framework into the indicator recommendation model. In the process of outputting the target dashboard indicator through the indicator recommendation model according to the target prompt word framework, the terminal also identifies the graphic mapping feature between the reference recommended indicator and the corresponding indicator display graphic type according to the indicator display graphic example in the indicator recommendation example. This can be to obtain the historical indicator display graphic set corresponding to the reference recommended indicator, calculate the proportion of each historical indicator display graphic type in the historical indicator display graphic set, extract the target proportion of the historical indicator display graphic type that is the same as the indicator display graphic type in the indicator display graphic example, and use the target proportion as the graphic mapping feature. For example, the historical indicator display graphic set of the reference recommended indicator A has the proportion of histograms, the proportion of bar graphs, and the proportion of line graphs, and the indicator display graphic type of the reference recommended indicator A in the indicator display graphic example is a line graph, then the proportion of line graphs in the historical indicator display graphic set is extracted. The historical indicator display graphic set can be obtained from the world knowledge graph or from the historical dashboard data set corresponding to the current user. After the indicator recommendation model recognizes the graphic mapping feature, it obtains the historical indicator display graphic set corresponding to the target dashboard indicator from the world knowledge graph, and determines the target display graphic type corresponding to the target dashboard indicator from the historical indicator display graphic set based on the graphic mapping feature.

[0089] The terminal obtains the target dashboard indicator and the target display graphic type output by the indicator recommendation model, obtains the indicator data according to the target dashboard indicator, converts the indicator data into graphic data according to the target display graphic type, and generates the target data dashboard corresponding to the current user according to the target dashboard indicator and the graphic data.

[0090] In an exemplary embodiment, Figure 3 As shown, a schematic diagram of a kanban indicator recommendation is provided. The kanban indicator recommendation process includes: generating a solution (theme-related fields and field key elements) for the kanban theme and its position (user information) input by the user through the indicator recommendation model, selecting the indicator (target kanban indicator) most relevant to the solution in the kanban indicator library through the indicator recommendation model, and outputting the most relevant indicator and the corresponding indicator data grouping dimension (such as the time dimension of the indicator data, the data source dimension, etc.) and display style (target display graphic type) through the indicator recommendation model.

[0091] Specifically, the framework structure of the prompt word framework is pre-defined in the terminal, such as a tree structure, etc., which can be defined through thought chain and Zero shot (zero sample learning) technology. The prompt word framework can be used to track and trace the process of generating solutions and recommending dashboard indicators. Get the kanban topic and position entered by the current user, and use the fine-tuned Embedding model (the Embedding model is the process of converting data (such as text, images, audio) into numerical vectors that can be processed by machine learning models. These vectors are usually low-dimensional and dense, and can capture the potential relationships and structures of the data) to vectorize the metadata of all existing indicators in the kanban indicator library (including indicator name, indicator definition, dimension list) into a multi-dimensional half-precision floating-point matrix (FP16) and store it in the vector library. Take a variety of enterprise operation optimization plans (topic-related fields and key elements of the fields) as the target, and also vectorize the enterprise operation optimization plans into a multi-dimensional half-precision floating-point matrix. Retrieve in the vector library through Euclidean distance (L2), or retrieve based on vector similarity, and recall the kanban indicators with the top 50 similarities in the kanban indicator library as candidate kanban indicators.

[0092] In an exemplary embodiment, the fine-tuning process for the Embedding model is as follows: no more than 50 correct question-answer pairs with correct mapping relationships are generated as training sets according to business scenarios, and the Embedding model is trained, wherein the core parameter learning_rate (learning rate) of the Embedding model can be set to 2e-5, the temperature (temperature parameter) can be set to 0.01, and the data transmission and calculation process can be made more efficient through fp16 half-floating point calculation, so that the Embedding model can identify what indicators can evaluate what indicator mapping features of the enterprise operation optimization plan according to the correct question-answer pairs, and generalize it to cover other business scenarios for indicator recommendation.

[0093] For example:

[0094] {query="Select and manage reliable suppliers",positive="Supplier on-time delivery rate"}

[0095] …

[0096] {query="Maintaining customer relationships",positive="Customer satisfaction"}.

[0097] Here, query represents the business optimization plan of the enterprise, and positive represents the correct evaluation indicator.

[0098] Then the recalled candidate dashboard indicators are assembled into the prompt word framework, and the indicator recommendation examples are embedded into the prompt word framework through Few shots (prompt word engineering) to obtain the target prompt word framework, which is input into the indicator recommendation model to output the target dashboard indicator. The graphics generation API can be called through the Text2API (allowing users to describe the operations they want to perform through natural language input, and the system will automatically convert these descriptions into corresponding API calls) to call the LLM (large language model) to return the image type and summary dimension (indicator data grouping dimension) according to the user information and indicator metadata, generate the corresponding graphics and form the dashboard.

[0099] In this embodiment, the display graphic type corresponding to the target dashboard indicator is determined by taking the graphic mapping characteristics between the reference recommended indicator and the corresponding indicator display graphic type in the indicator display graphic example as a reference basis, thereby improving the accuracy of the target dashboard indicator.

[0100] In an exemplary embodiment, the data dashboard generation method further includes:

[0101] Based on the user information, determine similar users corresponding to the current user from the user database;

[0102] The historical dashboard indicators corresponding to similar users are used as the first recommended dashboard indicators corresponding to the current user;

[0103] The first recommended dashboard indicator is added to the plurality of candidate dashboard indicators.

[0104] Among them, similar users refer to users with similar user characteristics. The first recommended dashboard indicator refers to the indicator that can be recommended to the current user in the historical dashboard indicators of similar users. The historical dashboard indicator refers to the dashboard indicator included in the data dashboard generated by similar users in historical time.

[0105] Exemplarily, after obtaining the user information of the kanban theme input by the current user, before generating the target data kanban of the current user, the terminal uses the Embedding model to vectorize the user information to obtain the user features of the current user. Then the user features of the current user are similarly calculated with the user features of other users in the vector library, and the user with the greatest similarity is regarded as a similar user. Obtain the various kanban data of similar users. Kanban data refers to the data used by similar users to generate data kanbans in historical time, including corresponding kanban themes and kanban indicators. The terminal searches for kanban data with the same kanban theme as the one input by the current user in the various kanban data of similar users, and obtains the historical kanban indicators in the kanban data. Display the historical dashboard indicators on the current page to recommend them to the current user, for example, display "Do you need to view the indicator A, indicator B, indicator B of the dashboard theme XXX?" on the current page. If the user chooses to accept the recommendation, the terminal responds to the current user's selection operation on at least one historical dashboard indicator, and uses the historical dashboard indicator as the first recommended dashboard indicator corresponding to the current user, and in the process of determining the target dashboard indicator from the multiple candidate dashboard indicators based on the dashboard theme, the theme-related field and the field key elements, the first recommended dashboard indicator is added to the multiple candidate dashboard indicators. If the user chooses not to accept the recommendation, the terminal responds to the current user's rejection operation on each historical dashboard indicator, and enters the step of determining the theme-related field corresponding to the dashboard theme and the field key elements corresponding to the theme-related field based on the dashboard theme and the user information.

[0106] In an exemplary embodiment, in order to understand the business background, a large language model can be used to introduce a world knowledge graph to perform pre-processing such as information supplementation or rewriting of user information of other users and the current user. For example, the user's position can be supplemented with a description of its specific responsibilities: User 1 "Position: Customer Relationship Maintenance Specialist; Job Responsibilities: Manage and develop customer relationships, improve customer satisfaction and loyalty", User 2: "Position: Customer Support Specialist; Job Responsibilities: Solve problems and difficulties encountered by customers in the process of using products or services".

[0107] The preprocessed user information is then vectorized using the Embedding model, sparse feature classifications are converted into dense latent semantic vector matrices, and the distance between similar features is shortened in the vector space. The user feature vector matrix processed by the fine-tuned Embedding model is used as a measurement standard to form a personalized representation of the user, and similar users can be accurately found based on user characteristics.

[0108] In an exemplary embodiment, the similarity of board users can be evaluated by calculating the Jaccard coefficient, and then the UserCF (a recommendation system algorithm that can generate personalized recommendations based on the similarity between users) is used to recommend indicators for the current user. Specifically, each user's interest set for indicators can be first constructed, such as obtaining the behavior data of each user on the corresponding historical data board to determine the indicators that the user is interested in, such as the indicators browsed, clicked, and manually maintained by the user. Assuming that the indicator interest sets of two users are A and B respectively, the specific calculation formula of the Jaccard coefficient is shown in formula (1).

[0109]

[0110] Among them, A∩B is the index that A and B are interested in together, and A∪B is the total number of indexes that user A or user B is interested in. For example, user A is interested in delivery timeliness, sales, and gross profit margin, and user B is interested in customer satisfaction, sales, and gross profit margin. The Jaccard coefficients of user A and user B are shown in formula (2).

[0111]

[0112] Indicates that user A and user B have 50% overlap in indicator interests.

[0113] In this embodiment, similar users are determined based on user characteristics, and the historical dashboard indicators of similar users are recommended to the current user, thereby improving the flexibility of recommending dashboard indicators. After the user accepts, the indicators are added to the candidate dashboard indicators for recommendation of target dashboard indicators, thereby improving the accuracy of the target dashboard indicators.

[0114] In an exemplary embodiment, in step 210, after generating a target data dashboard corresponding to the current user based on the target dashboard indicator and indicator data, the data dashboard generation method further includes:

[0115] Based on the target dashboard indicators, identify similar data dashboards from the data dashboard library;

[0116] Use the dashboard indicators other than the target dashboard indicators in the similar data dashboard as the second recommended dashboard indicators;

[0117] Showing the second recommended dashboard indicator.

[0118] Among them, the similar data dashboard refers to a data dashboard whose dashboard indicators are similar to the dashboard indicators of the target data dashboard of the current user. The second recommended dashboard indicator refers to a dashboard indicator that can be recommended to the current user in the similar data dashboard.

[0119] Exemplarily, after the terminal generates the target data dashboard corresponding to the current user, since the display data of each indicator is displayed in a display frame of the same size in the dashboard, the indicator features extracted from the display data of each indicator in the target data dashboard can be described as the card features corresponding to each indicator. Then, the large language model is used to introduce the world knowledge graph to perform preprocessing such as supplementing or rewriting the card features extracted from the target data dashboard of the current user, and then the preprocessed card features are vectorized using the Embedding model to obtain the card feature vectors corresponding to each target dashboard indicator. The card feature vectors corresponding to the current user are matched with the card feature vectors corresponding to the data dashboards of other users in the vector library, and the data dashboards of other users with a matching degree greater than a preset threshold are used as similar data dashboards. Generally, the dashboard data of users who completely match the card feature vectors of the current user are used as similar data. Then the terminal uses the dashboard indicators other than the target dashboard indicators in the similar data dashboard as the second recommended dashboard indicators. For example, the current user's target data dashboard has 10 card feature vectors of target dashboard indicators, and the searched similar data dashboard has 12 card feature vectors of dashboard indicators, among which the 10 card feature vectors corresponding to the similar users are exactly the same as the 10 card feature vectors of the current user, and the dashboard indicators of the remaining 2 card feature vectors of the similar users are used as the second recommended dashboard indicators.

[0120] The terminal displays the second recommended dashboard indicator on the current page of the target data dashboard to recommend it to the current user, such as "Do you need to add indicator C and indicator D?" If the user chooses to accept the recommendation, the terminal responds to the current user's selection operation of at least one second recommended dashboard indicator and adds the selected second recommended dashboard indicator to the prompt word frame corresponding to the current user. If the user chooses not to accept the recommendation, the second recommended dashboard indicator is hidden.

[0121] In this embodiment, by recommending indicators to the current user based on similar data dashboards, indicators that the current user is interested in or may have missed can be recommended to improve the flexibility of indicator recommendation.

[0122] In an exemplary embodiment, in step 210, after generating a target data dashboard corresponding to the current user based on the target dashboard indicator and indicator data, the data dashboard generation method further includes:

[0123] Obtain the modification requirement information of the current user for the target data dashboard, decompose the modification requirement information, and obtain at least one sub-requirement information;

[0124] Identify the modification intentions corresponding to at least one sub-demand information, and generate a kanban modification instruction based on the at least one sub-demand information and the corresponding modification intention;

[0125] Send the kanban modification instruction to the kanban modification server, and obtain the updated kanban data returned by the kanban modification server based on the kanban modification instruction;

[0126] The target data dashboard is updated based on the updated dashboard data to obtain an updated data dashboard.

[0127] Exemplarily, the terminal responds to the current user's kanban modification operation and obtains the modification requirement information input by the current user, which may be a modification of the display data of any indicator in the target data kanban, such as modification of the data content of the display data, modification of the graphic type of the display data, deletion of the display data of the indicator, etc. The terminal decomposes the modification requirement information to obtain at least one sub-requirement information, and identifies the intent of each sub-requirement information, determines whether the modification operation is allowed according to the identified intent, and displays the modification exception to the current user if it is not allowed; if it is allowed, generates a kanban modification instruction according to the sub-requirement information and the corresponding modification intent, calls an API to send the kanban modification instruction to the board modification server, obtains the updated kanban data returned by the kanban modification server based on the kanban modification instruction, and updates the target data kanban according to the updated kanban data to obtain an updated data kanban.

[0128] In an exemplary embodiment, Figure 4As shown, a flow diagram of demand modification is provided. An Agent node for executing the modification of the target data dashboard is set in the terminal (in a continuous integration / continuous deployment (CI / CD) system such as Jenkins, the Agent node is a slave server, which can communicate with the master server (Master) and execute the build task, and the master server (Master) can be a dashboard modification server). The terminal will first modify the demand information input problem disassembly Agent node, which is used to split the modification demand information entered by the user into N atomic problems; then the domain rejection Agent node prompts the user that the operation cannot be completed when the atomic problem is not related to the modification of the dashboard or indicator; then the intention recognition Agent node uses a large model to identify the atomic problem respectively, and transfers the intention of the atomic problem to one of the modification summary dimension node, the modification filter condition node, and the modification graphic type node. For example, when the user inputs the demand "help me summarize the sales through the sales department", the demand is routed to the modification summary dimension node; the Text2API Agent node is called, which includes modifying the summary dimension, modifying the filter condition, and modifying the graphic type, sending the dashboard modification instruction to the AIP, obtaining the returned result, and unifying the format of the returned result through the prompt word project. For example, specify that the large model must be output in JSON format, and the key-value pairs must be as follows:

[0129] {

[0130] "Aggregation Dimension":""

[0131] “Filter Condition”: [“”,…,””]

[0132] "Graphic Type":""

[0133] }.

[0134] In this embodiment, the intention recognition system is constructed through the Agent framework, so that users can freely modify the card summary dimensions, filtering conditions and graphic types by describing their own modification ideas and directions, which improves user participation and satisfaction. In addition, the automated process from user input to system response is realized through the Text2API technology, which reduces the need for manual operation and improves efficiency.

[0135] In an exemplary embodiment, Figure 5As shown, a flowchart of indicator recommendation optimization is provided. Specifically, it includes: pre-embedding in the terminal to obtain the changes of the current user after the dashboard is generated; feature engineering of the current user, extracting the corresponding sub-industry of the position and the company to which the user belongs on the user side (i.e. extracting user features), extracting indicator classification, graphic type, card position and the theme of the dashboard on the card side (i.e. extracting card features in the target data dashboard); performing feature vectorization on user features and card features, which can be passing user features and card features as input parameters of prompt words, introducing world knowledge through a large language model to supplement and rewrite user features and card features, and using an embedding model for vectorization, and evaluating the similarity between users based on the user feature vectors of the current user and other users.

[0136] After the target data dashboard is updated to the updated data dashboard according to the modification requirement information of the current user, the changed original card features are used as negative samples, and the changed card features are used as positive samples. The model is recommended based on the training indicators of the positive samples and the negative samples.

[0137] After finding the top 3 similar users based on the user feature vectors of the current user and other users, when the current user creates a dashboard, the indicator recommendation model is optimized based on the data dashboards and behavior data of similar users. For example, the first recommended indicator of similar users is involved in the indicator recommendation model to recommend indicators, or the second recommended indicator of similar data dashboards is used to optimize the output of the indicator recommendation model.

[0138] In an exemplary embodiment, Figure 6 As shown, a schematic diagram of the process of data dashboard generation and post-processing is provided. Specifically, it includes:

[0139] Get the board topic and position entered by the user: The user enters the board topic and position through the interface (such as Figure 6 -B) in the form provided in box 1) or by direct input, provide the theme of the board and the expected role (such as market analyst, project manager, etc.); if the user does not provide input, the system will automatically extract the user's personal information, including but not limited to the position, industry, company information, rank, etc.

[0140] Generate topic-related domains: Use a general large language model (LLM) to process the keywords input by users and generate multiple key elements in the fields for optimizing business operations. The specific operation is to use the topic-related domains generated by the large language model as the root node of the tree, and generate multiple key elements in the field based on each root node, and create a prompt word framework (such as Figure 6-B) After the dashboard is generated, a large language model can be used to generate business optimization suggestions based on key elements of the field. Then, after the user's industry and job role are converted into vectors, the similarity between users is calculated through vector matching.

[0141] Recall the TOP50 candidate dashboard indicators: With the key success area as the target, recall the TOP50 related indicators in the indicator vector library through vector similarity. The indicator vector library contains metadata such as the name, definition, and dimension of the indicator; insert the prompt word framework and generate recommendations; splice the recalled indicators according to the predetermined prompt word framework to form a complete set of dashboard indicators (target prompt word framework); then use the few shot prompt word project to guide the indicator recommendation model to recommend indicators and graphics through a small number of examples (indicator recommendation examples). For example, provide an example in the prompt word context. When the key success area is "increase sales of each organization", it is necessary to tell the indicator recommendation model that the result it hopes to return is a card with the indicator name "sales", the grouping dimension "business organization", and the graphic type "bar chart", and generate a description sentence such as "because the key success area mentions the sales indicator and the organizational dimension, and the best way to compare the gaps between organizations is a bar chart."

[0142] Complete the first dashboard generation: Automatically generate a user-specific dashboard based on the recommendation results and provide a preview (such as Figure 6 -b in box 3).

[0143] User input modification requirements: Users can submit modification requirements through the interface (such as Figure 6 -B, the system automatically parses the user input through the Text2API technology and converts it into the corresponding API call to adjust the indicator, dimension, card type and chart type of a card in the dashboard.

[0144] Feature engineering: Extract user features and card features, and perform feature engineering, including normalization, encoding conversion, etc. Mark the processed features with user features, card classification, and company attributes, and store them in the vector library. For example, workshop supervisors in the automotive manufacturing industry often use bar charts to analyze the yield rate of each workshop and record it as a user habit.

[0145] Calculate the similarity between users: Calculate the similarity between users through collaborative filtering (UserCF) technology. Use user feature vectors to find user groups with similar interests and behavior patterns. For example, two users are both in the seafood retail industry, and they are sales director and sales consultant respectively. They have both viewed the sales profit trend of seafood in the past. When the sales director has recently viewed the number of sales orders, the sales consultant may also be interested in it.

[0146] Find the top 3 similar users and optimize the generated results: Find the top 3 similar users for each user. When a user creates a dashboard, optimize the dashboard generation results based on the user preference model of similar users to provide more personalized recommendations.

[0147] Model and vector library optimization: Based on the user's modification requirements, the indicator recommendation algorithm is continuously optimized and iterated. The modification requirements include modifying indicators, dimensions, and graph types, and using them as negative samples for back propagation algorithm learning. The vector library and feature engineering process are regularly updated to ensure the accuracy and timeliness of dashboard generation.

[0148] It should be understood that, although the various steps in the flowcharts involved in the above-mentioned embodiments are displayed in sequence according to the indication of the arrows, these steps are not necessarily executed in sequence according to the order indicated by the arrows. Unless there is a clear explanation in this article, the execution of these steps does not have a strict order restriction, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above-mentioned embodiments can include multiple steps or multiple stages, and these steps or stages are not necessarily executed at the same time, but can be executed at different times, and the execution order of these steps or stages is not necessarily to be carried out in sequence, but can be executed in turn or alternately with other steps or at least a part of the steps or stages in other steps.

[0149] Based on the same inventive concept, the embodiment of the present application also provides a data dashboard generation device for implementing the data dashboard generation method involved above. The implementation scheme for solving the problem provided by the device is similar to the implementation scheme recorded in the above method, so the specific limitations in one or more data dashboard generation device embodiments provided below can refer to the limitations of the data dashboard generation method above, and will not be repeated here.

[0150] In an exemplary embodiment, Figure 7 As shown, a data dashboard generating device 700 is provided, comprising: an acquisition module 702, an element determination module 704, a candidate indicator determination module 706, a target indicator determination module 708 and a dashboard generating module 710, wherein:

[0151] The acquisition module 702 is used to acquire the board theme and user information corresponding to the current user;

[0152] An element determination module 704 is used to determine the theme-related field corresponding to the theme of the board and the field key elements corresponding to the theme-related field based on the theme of the board and the user information;

[0153] A candidate indicator determination module 706 is used to determine multiple candidate kanban indicators corresponding to the current user from the kanban indicator library based on the key elements of the domain;

[0154] A target indicator determination module 708 is used to determine a target kanban indicator from a plurality of candidate kanban indicators based on the kanban theme, the theme-related fields and the field key elements;

[0155] The dashboard generation module 710 is used to obtain indicator data corresponding to the target dashboard indicator, and generate a target data dashboard corresponding to the current user based on the target dashboard indicator and the indicator data.

[0156] In an exemplary embodiment, the element determination module 704 is also used to input the kanban theme and user information into the topic analysis model, perform a domain search on the kanban theme based on the built-in world knowledge graph through the topic analysis model, and generate a topic-related domain corresponding to the kanban theme; perform a domain element search on the topic-related domain based on the world knowledge graph and user information through the topic analysis model, and generate domain key elements corresponding to the topic-related domain.

[0157] In an exemplary embodiment, the target indicator determination module 708 is also used to generate a prompt word framework corresponding to the current user based on the hierarchical relationship between the board theme, the theme-related fields, the field key elements and multiple candidate board indicators; obtain the indicator recommendation example, embed the indicator recommendation example into the prompt word framework, and obtain the target prompt word framework; input the target prompt word framework into the indicator recommendation model, identify the indicator mapping features between the reference key elements in the indicator recommendation example and the reference recommended indicator through the indicator recommendation model, and determine the target board indicator from multiple candidate board indicators based on the indicator mapping features, the theme-related fields and the field key elements.

[0158] In an exemplary embodiment, the indicator recommendation example also includes an indicator display graphic example; the data dashboard generation device 700 is also used to identify the graphic mapping characteristics between the reference recommended indicator and the corresponding indicator display graphic type in the indicator display graphic example through the indicator recommendation model, and determine the target display graphic type corresponding to the target dashboard indicator based on the graphic mapping characteristics; the dashboard generation module 710 is also used to convert the corresponding indicator data into target graphic data according to the target display graphic type of the target dashboard indicator; and generate the target data dashboard corresponding to the current user based on the target dashboard indicator and the target graphic data.

[0159] In an exemplary embodiment, the data dashboard generating device 700 is also used to determine similar users corresponding to the current user from the user database based on user information; use the historical dashboard indicators corresponding to the similar users as the first recommended dashboard indicators corresponding to the current user; and add the first recommended dashboard indicators to multiple candidate dashboard indicators.

[0160] In an exemplary embodiment, the data dashboard generating device 700 is also used to determine similar data dashboards from the data dashboard library based on the target dashboard indicator; use the dashboard indicators other than the target dashboard indicators in the similar data dashboards as second recommended dashboard indicators; and display the second recommended dashboard indicators.

[0161] In an exemplary embodiment, the data kanban generating device 700 is also used to obtain the modification requirement information of the current user for the target data kanban, decompose the modification requirement information to obtain at least one sub-requirement information; identify the modification intention corresponding to at least one sub-requirement information, and generate a kanban modification instruction based on at least one sub-requirement information and the corresponding modification intention; send the kanban modification instruction to the kanban modification server, and obtain the updated kanban data returned by the kanban modification server based on the kanban modification instruction; update the target data kanban based on the updated kanban data to obtain an updated data kanban.

[0162] Each module in the above data dashboard generation device can be implemented in whole or in part by software, hardware, or a combination thereof. Each module can be embedded in or independent of a processor in a computer device in the form of hardware, or can be stored in a memory in a computer device in the form of software, so that the processor can call and execute operations corresponding to each module.

[0163] In an exemplary embodiment, a computer device is provided. The computer device may be a server, and its internal structure diagram may be as shown in FIG. Figure 8 As shown. The computer device includes a processor, a memory, an input / output interface (Input / Output, referred to as I / O) and a communication interface. Among them, the processor, the memory and the input / output interface are connected through a system bus, and the communication interface is connected to the system bus through the input / output interface. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store data such as a target data dashboard. The input / output interface of the computer device is used to exchange information between the processor and an external device. The communication interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a data dashboard generation method is implemented.

[0164] In an exemplary embodiment, a computer device is provided. The computer device may be a terminal, and its internal structure diagram may be as shown in FIG. Fig. 9As shown. The computer device includes a processor, a memory, an input / output interface, a communication interface, a display unit and an input device. Among them, the processor, the memory and the input / output interface are connected through a system bus, and the communication interface, the display unit and the input device are connected to the system bus through the input / output interface. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The input / output interface of the computer device is used to exchange information between the processor and the external device. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner can be implemented through WIFI, a mobile cellular network, near field communication (Near Field Communication, NFC) or other technologies. When the computer program is executed by the processor, a data dashboard generation method is implemented. The display unit of the computer device is used to form a visually visible picture, which can be a display screen, a projection device or a virtual reality imaging device. The display screen can be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device can be a touch layer covering the display screen, or a button, trackball or touchpad set on the computer device shell, or an external keyboard, touchpad or mouse.

[0165] Those skilled in the art will understand that Figure 8-Figure 9 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0166] In an exemplary embodiment, in one embodiment, a computer device is further provided, including a memory and a processor, wherein a computer program is stored in the memory, and the processor implements the steps in the above-mentioned method embodiments when executing the computer program.

[0167] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps in the above-mentioned method embodiments are implemented.

[0168] In one embodiment, a computer program product is provided, including a computer program, which implements the steps in the above method embodiments when executed by a processor.

[0169] 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 application 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 relevant regulations.

[0170] Those of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to the memory, database or other medium used in the embodiments provided in the present application can include at least one of non-volatile memory and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. Non-relational databases may include distributed databases based on blockchains, etc., but are not limited to this. The processor involved in each embodiment provided in this application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, an artificial intelligence (AI) processor, etc., but are not limited to this.

[0171] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this application.

[0172] The above-described embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the present application. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the attached claims.

Claims

1. A data dashboard generation method, characterized in that: The method comprises: Get the board topic and user information corresponding to the current user; Based on the kanban theme and the user information, determining a theme-related field corresponding to the kanban theme and a field key element corresponding to the theme-related field; Based on the key elements of the field, determining a plurality of candidate dashboard indicators corresponding to the current user from a dashboard indicator library; Determine a target dashboard indicator from the plurality of candidate dashboard indicators based on the dashboard theme, the theme-related fields and the field key elements; Obtain indicator data corresponding to the target dashboard indicator, and generate a target data dashboard corresponding to the current user based on the target dashboard indicator and the indicator data.

2. The method according to claim 1, characterized in that: The determining, based on the kanban theme and the user information, a theme-related field corresponding to the kanban theme and a field key element corresponding to the theme-related field includes: Inputting the kanban theme and the user information into a topic analysis model, and performing a domain search on the kanban theme based on a built-in world knowledge graph by the topic analysis model to generate a topic-related domain corresponding to the kanban theme; The topic analysis model searches for domain elements in the topic-related domain based on the world knowledge graph and the user information, and generates domain key elements corresponding to the topic-related domain.

3. The method according to claim 1, characterized in that The step of determining a target dashboard indicator from the plurality of candidate dashboard indicators based on the dashboard theme, the theme-related fields and the field key elements includes: Based on the hierarchical relationship between the kanban theme, the theme-related fields, the field key elements, and the plurality of candidate kanban indicators, a prompt word framework corresponding to the current user is generated; Obtaining an indicator recommendation example, embedding the indicator recommendation example into the prompt word framework, and obtaining a target prompt word framework; The target prompt word framework is input into an indicator recommendation model, and the indicator recommendation model is used to identify indicator mapping features between the reference key elements and the reference recommended indicators in the indicator recommendation example. Based on the indicator mapping features, the subject-related fields and the field key elements, the target dashboard indicator is determined from the multiple candidate dashboard indicators.

4. The method according to claim 3, characterized in that The indicator recommendation example also includes an indicator display graphic example; the method also includes: Identify, by the indicator recommendation model, a graphic mapping feature between the reference recommended indicator and the corresponding indicator display graphic type in the indicator display graphic example, and determine a target display graphic type corresponding to the target dashboard indicator based on the graphic mapping feature; The generating the target data dashboard corresponding to the current user based on the target dashboard indicator and the indicator data includes: According to the target display graphic type of the target dashboard indicator, the corresponding indicator data is converted into target graphic data; A target data dashboard corresponding to the current user is generated based on the target dashboard indicator and the target graphic data.

5. The method according to claim 1, characterized in that The method further comprises: Based on the user information, determining a similar user corresponding to the current user from a user database; Using the historical dashboard indicators corresponding to the similar users as the first recommended dashboard indicators corresponding to the current user; The first recommended dashboard indicator is added to the multiple candidate dashboard indicators.

6. The method according to claim 1, characterized in that After generating the target data dashboard corresponding to the current user based on the target dashboard indicator and the indicator data, the method further includes: Based on the target dashboard indicator, determine similar data dashboards from a data dashboard library; Using the dashboard indicators other than the target dashboard indicator in the similar data dashboard as the second recommended dashboard indicator; The second recommended dashboard indicator is displayed.

7. The method according to claim 1, characterized in that After generating the target data dashboard corresponding to the current user based on the target dashboard indicator and the indicator data, the method further includes: Acquire the modification requirement information of the current user for the target data dashboard, decompose the modification requirement information, and obtain at least one sub-requirement information; Identify the modification intentions respectively corresponding to the at least one sub-demand information, and generate a kanban modification instruction based on the at least one sub-demand information and the corresponding modification intention; Sending the kanban modification instruction to the kanban modification server, and obtaining updated kanban data returned by the kanban modification server based on the kanban modification instruction; The target data dashboard is updated based on the updated dashboard data to obtain an updated data dashboard.

8. A data dashboard generating device, characterized in that: The device comprises: The acquisition module is used to obtain the board topic and user information corresponding to the current user; An element determination module, used for determining, based on the kanban theme and the user information, a theme-related field corresponding to the kanban theme and a field key element corresponding to the theme-related field; A candidate indicator determination module, used to determine a plurality of candidate kanban indicators corresponding to the current user from a kanban indicator library based on the key elements of the field; A target indicator determination module, configured to determine a target kanban indicator from the plurality of candidate kanban indicators based on the kanban theme, the theme-related fields and the field key elements; The dashboard generation module is used to obtain the indicator data corresponding to the target dashboard indicator, and generate the target data dashboard corresponding to the current user based on the target dashboard indicator and the indicator data.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

11. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images