Data analysis method based on interactive smart card and related apparatus

Abstract

This application provides a data analysis method and related apparatus based on interactive smart cards. The method includes acquiring a data analysis command input by a user; understanding the intent of the data analysis command to obtain a command analysis status; if the command analysis status indicates ambiguity or missing parameters, generating and displaying an intent interaction card; responding to the user's interaction with the intent interaction card, confirming or correcting candidate data indicators to obtain target data indicators, and / or supplementing the target data parameters corresponding to the target data indicators; performing data extraction operations based on the target data indicators and target data parameters, and displaying the original analysis data in an analytical presentation format. Thus, by guiding the user to confirm the data intent and parameters through an interactive card mechanism, and presenting the extraction results based on an adapted analytical presentation format, the accuracy of data extraction can be ensured, the intuitiveness of the analysis results enhanced, and the data analysis experience optimized.

Description

Technical Field

[0001] This application belongs to the field of electronic digital data processing technology, specifically relating to a data analysis method and related apparatus based on interactive smart cards. Background Technology

[0002] With the rapid development of information technology, the scope and volume of data in the financial investment research field are extremely broad. Analysts typically rely on pre-designed analysis pages in traditional data terminals for data extraction, or use data indicator browsers to search for indicators one by one before analysis. Traditional data terminals are convenient for common analysis scenarios, but their pre-built modules have a fixed structure and are numerous, making it difficult to meet users' needs for quickly exploring new analytical approaches. Furthermore, the data browser mode requires users to have a high level of understanding of data systems.

[0003] Meanwhile, traditional terminals cannot help users clarify vague or incomplete analytical needs, leading to discrepancies between analytical results and users' actual needs, thus affecting analytical efficiency. Furthermore, they typically use fixed display templates when presenting analytical data, making it difficult to adapt the format to different data presentation requirements, resulting in less intuitive information delivery and making it difficult for users to quickly gain effective insights. Summary of the Invention

[0004] This application proposes a data analysis method and related apparatus based on interactive smart cards. By intelligently understanding user instructions and dynamically generating intent interaction cards, the complex data retrieval process is transformed into a lightweight interface confirmation interaction, significantly reducing the threshold for data analysis. At the same time, combined with automatic data extraction and visualization, the flexibility and intuitiveness of data analysis are effectively improved.

[0005] In a first aspect, embodiments of this application provide a data analysis method based on an interactive smart card, including:

[0006] Obtain data analysis commands input by the user;

[0007] The intent of the data analysis command is understood to obtain the command analysis status. The command analysis status is used to indicate whether there is intent ambiguity or missing parameters in the data analysis command, as well as the candidate data indicators identified from the data analysis command.

[0008] If the instruction analysis status indication has ambiguity or missing parameters, an intent interaction card is generated and displayed. The intent interaction card is a front-end interface interaction component used to guide the user to confirm the candidate data indicators and / or supplement the missing data parameters.

[0009] In response to user interaction, confirm or correct candidate data metrics to obtain target data metrics, and / or supplement target data parameters corresponding to the target data metrics;

[0010] Data extraction is performed based on the target data indicators and parameters to obtain raw analysis data, which is then displayed in an analytical presentation format. The analytical presentation format refers to the front-end interface format for visualizing and interactively analyzing the raw analysis data.

[0011] Secondly, embodiments of this application provide a data analysis device based on an interactive smart card, comprising:

[0012] The instruction acquisition unit is used to acquire data analysis instructions input by the user.

[0013] The status analysis unit is used to understand the intent of the data analysis instructions and obtain the instruction analysis status. The instruction analysis status is used to indicate whether there is any intent ambiguity or missing parameters in the data analysis instructions, as well as the candidate data indicators identified from the data analysis instructions.

[0014] The card generation unit is used to generate and display intent interaction cards if there is ambiguity in the intent or missing parameters in the instruction analysis status indication. Intent interaction cards are front-end interface interaction components used to guide users to confirm candidate data indicators and / or supplement missing data parameters.

[0015] An interactive response unit is used to respond to user interaction operations, confirm or correct candidate data indicators to obtain target data indicators, and / or supplement target data parameters corresponding to the target data indicators.

[0016] The data display unit is used to perform data extraction operations based on target data indicators and target data parameters to obtain raw analysis data, and to display the raw analysis data in an analysis display format. The analysis display format refers to the front-end interface format for visualizing and interactively analyzing the raw analysis data.

[0017] Thirdly, embodiments of this application provide an electronic device including a processor, a memory, and one or more programs, the one or more programs being stored in the memory and configured to be executed by the processor, the programs including instructions for performing the steps as described in the method of the first aspect of this application.

[0018] Fourthly, embodiments of this application provide a computer-readable storage medium having a computer program or instructions stored thereon, wherein the computer program or instructions, when executed by a processor, implement the steps of the method of the first aspect of this application.

[0019] As can be seen, this application's embodiment effectively solves the problems of rigid interaction, low flexibility, and high user expertise requirements in traditional data analysis tools by constructing a state-driven intelligent interaction closed loop. Specifically, this interaction process transforms the high cognitive load tasks of data indicator identification and query syntax construction, which traditionally require users to actively complete, into a collaborative process actively guided by a visual interface, where users only need to confirm with low cognitive load. When ambiguous or missing user commands are detected, the terminal interface immediately triggers state-adaptive interaction cards, guiding the user to confirm step by step, and finally automatically extracting data and presenting it in an intelligently matched visual form. This method effectively lowers the barrier to entry for professional data analysis tools and improves the efficiency and accuracy of task completion under vague or incomplete initial intentions, achieving an efficient and reliable conversion from vague needs to precise analysis results. Attached Figure Description

[0020] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0021] Figure 1 This is a structural block diagram of an intelligent interactive data analysis system provided in an embodiment of this application;

[0022] Figure 2 This is a structural block diagram of another intelligent interactive data analysis system provided in the embodiments of this application;

[0023] Figure 3 This is a flowchart illustrating a data analysis method based on an interactive smart card, as provided in an embodiment of this application.

[0024] Figure 4 This is a schematic diagram of a visual interface for ambiguous instructions provided in an embodiment of this application;

[0025] Figure 5 This is a schematic diagram of a visualization interface based on analysis and display provided in an embodiment of this application;

[0026] Figure 6 This is a schematic diagram of a visual interface for intelligent recommendation guidance provided in an embodiment of this application;

[0027] Figure 7 This is a functional unit block diagram of a data analysis device based on an interactive smart card provided in an embodiment of this application;

[0028] Figure 8This is a functional unit block diagram of another data analysis device based on an interactive smart card provided in this application embodiment;

[0029] Figure 9 This is a structural block diagram of an electronic device provided in an embodiment of this application. Detailed Implementation

[0030] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present application.

[0031] The terms "first," "second," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish different objects, not to describe a specific order. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or apparatus that includes a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to these processes, methods, products, or apparatuses.

[0032] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0033] Among related technologies, compared to general data analysis tools, intelligent data analysis systems based on natural language interaction are more suitable for solving data query and analysis needs in specific fields, especially in financial investment research. Due to the limitations of traditional financial data terminals in terms of pre-built analysis modules and data browsers, interactive analysis systems based on AI (Artificial Intelligence) intent understanding are gradually becoming the mainstream solution for complex financial data analysis scenarios, which can, to some extent, solve the problem of the high barrier to entry caused by complex database table structures and numerous indicator types. However, existing interactive analysis systems still use a uniform interaction method when guiding users to clarify their analysis intent, such as using plain text dialogue or fixed forms for clarification and confirmation. However, the types of ambiguity and missing parameters of user input intent vary significantly in different analysis scenarios. Especially in the data analysis scenario in the financial field, there are usually multiple situations that need to be confirmed, such as indicator ambiguity, direction specification, and missing parameters. The uniform interaction method cannot adapt to these differentiated confirmation needs, resulting in a still high burden of understanding and operation for users in the intent clarification stage, which affects analysis efficiency.

[0034] To address the aforementioned issues, embodiments of this application provide a data analysis method and related apparatus based on interactive smart cards.

[0035] The following describes an embodiment of an intelligent interactive data analysis system provided by this application.

[0036] Please see Figure 1 , Figure 1 This is a structural block diagram of an intelligent interactive data analysis system provided in an embodiment of this application. Figure 1 As shown, the intelligent interactive data analysis system 10 includes a client 110, a server 120 and a database 125. The server 120 is configured with an intent understanding module 121, a card generation module 122, a data extraction module 123 and a display matching module 124. The client 110 and the server 120 are connected in communication.

[0037] In this process, the user inputs a data analysis command through client 110. Client 110 then transmits the command to intent understanding module 121 via a communication connection with server 120. Intent understanding module 121 interprets the user's input command to obtain the command analysis status. This status indicates whether the command has ambiguous intent or missing parameters, and identifies candidate data indicators from the command. Card generation module 122 generates an intent interaction card when the command analysis status indicates ambiguity or missing parameters. This card guides the user to confirm candidate data indicators and / or supplement missing data parameters. Specifically, card generation module 122 displays the intent interaction card through interaction with client 110 to achieve visual interaction between client 110 and the user. Further, data extraction module 123 responds to the user's interaction by confirming or correcting candidate data indicators to obtain target data indicators, and / or supplementing the target data parameters. Based on the target data indicators and parameters, it performs data extraction from database 125 to obtain raw analysis data. Finally, the display matching module 124 is used to determine the target display format from multiple preset analysis display formats based on the data characteristics of the original analysis data, and return the original analysis data to the client 110 for display in the target display format (i.e., visualize the original analysis data). The analysis display format includes at least one of the following: data table format, trend chart format, comparison chart format, and comprehensive dashboard format.

[0038] The user inputs data analysis commands through client 110. Based on data interaction with server 120, client 110 executes the data analysis method based on interactive smart cards as described in the embodiments of this application. Client 110 refers to a smart device used by the user, such as a mobile phone, tablet, or personal computer. Server 120 can be a single server, a server cluster consisting of several servers, or a cloud computing service center. Database 125 can be a local database deployed on server 120 or a separately deployed remote database server. Figure 1 The database 125 shown in the image is a remotely deployed database server, and this is not intended to be the only one.

[0039] Please see Figure 2 , Figure 2 This is a structural block diagram of another intelligent interactive data analysis system provided in an embodiment of this application. For example... Figure 2 As shown, different from Figure 1The client-server collaborative architecture shown in this embodiment adopts a client-independent architecture. The intelligent interactive data analysis system 10 includes a client 110, which is configured with an intent understanding module 121, a card generation module 122, a data extraction module 123, a display matching module 124, and a local database 210.

[0040] The functions of the intent understanding module 121, card generation module 122, data extraction module 123, and display matching module 124 are as follows: Figure 1 The illustrated embodiment is the same, except that all the modules are deployed locally on the client 110. The data extraction module 123 performs data extraction operations from the local database 210 on the client 110. All processing steps are completed locally on the client 110, without the need for network communication with the server. The client 110 refers to a smart device used by the user, such as a mobile phone, tablet computer, or personal computer. The local database 210 can be a local database or local file system deployed on the client 110, and is not limited to a single type. Users directly obtain the visualized raw analysis data corresponding to the data analysis commands through visual interaction with the client 110.

[0041] The user inputs data analysis commands through client 110, and client 110 executes the data analysis method based on interactive smart cards as described in the embodiments of this application. Client 110 refers to a smart device used by the user, such as a mobile phone, tablet computer, or personal computer. Local database 210 can be a local database or local file system deployed on client 110, or it can be an independent storage device connected to client 110; no single limitation is made here.

[0042] The following describes a data analysis method based on interactive smart cards provided by an embodiment of this application.

[0043] Please see Figure 3 , Figure 3 This is a flowchart illustrating a data analysis method based on an interactive smart card, provided in an embodiment of this application, applicable to, for example... Figure 1 or Figure 2 The client shown is 110. (As shown) Figure 3 As shown, the method includes:

[0044] Step S301: Obtain the data analysis instructions input by the user.

[0045] In terms of input methods, users can input analysis commands in natural language text form through the text input box on client 110 via keyboard or touch, or they can acquire voice signals by calling the voice acquisition device, perform voice recognition processing on the voice signals, and use the recognized text content as data analysis commands. Regarding the format of the commands, data analysis commands can be natural language questions where the user directly specifies specific analysis indicators (e.g., "Query the net profit attributable to the parent company of Company A in the third quarter of 2024"), or natural language descriptions containing only the direction or intent of the analysis (e.g., "I want to see the profitability of the leading companies in the liquor industry" or "How is the recent performance of the chip sector?"). They can also be formatted analysis requests triggered by preset shortcut operations on the graphical interface of client 110 (such as clicking or dragging specific data tags).

[0046] Step S302: Perform intent understanding on the data analysis instructions to obtain the instruction analysis status.

[0047] The instruction analysis status is used to indicate whether there is any ambiguity in intent or missing parameters in the data analysis instruction, as well as the candidate data indicators identified from the data analysis instruction.

[0048] When performing intent understanding on data analysis instructions, the instruction text is first processed using natural language processing to extract semantic elements related to data metrics, analysis objects, time ranges, and analysis operations. Based on the extraction results, intent understanding identifies candidate data metrics in the following ways:

[0049] In the first scenario, the user-specified data metric name is successfully identified from the data analysis command. For example, the user input might include terms like "net profit," "price-to-earnings ratio," or "operating revenue," which can be directly mapped to database metric fields. In this case, the identified data metric is output as a candidate data metric. However, because the user's description may contain abbreviations, alternative names, or colloquial expressions, the candidate data metric may have a one-to-many or many-to-one mapping relationship with the official metric name in the database, leading to ambiguity. For instance, the user-input "price-to-earnings ratio (PE)" could correspond to any of the following: "static price-to-earnings ratio," "dynamic price-to-earnings ratio," or "rolling price-to-earnings ratio." In this case, the candidate data metric is marked as an ambiguous or similar metric.

[0050] The second scenario involves an empty set of candidate data indicators if no directly mappable data metric names are identified from the data analysis command. In this case, it's necessary to further determine whether the user has expressed a clear analytical direction or topic in their command. For example, if a user inputs "I want to see the profitability of the liquor industry," where "profitability" is an analytical direction rather than a specific indicator, the system needs to map the analytical direction to a set of potentially relevant candidate data indicators (such as "net profit attributable to parent company," "gross profit margin," and "net profit margin") based on a pre-set indicator recommendation model for the user to confirm later.

[0051] In addition, the instruction analysis status also comprehensively records whether there are any missing parameters in the data analysis instructions. Missing parameters include, but are not limited to: not specifying the code of the analysis object, not specifying the analysis time interval or reporting period, and not specifying the data unit setting. The instruction analysis status integrates the intent ambiguity type and parameter missing status into a unified status identifier, which is used by subsequent steps to determine whether an intent interaction card needs to be generated.

[0052] In one possible embodiment, intent ambiguity is categorized into first ambiguity, second ambiguity, or third ambiguity based on the type of ambiguity. First ambiguity indicates that there are ambiguous or similar indicators among the candidate data indicators. Second ambiguity indicates that the user specifies an analysis direction but does not specify a candidate data indicator. Third ambiguity indicates that the user does not specify either a candidate data indicator or an analysis direction. The intent understanding of the data analysis instruction to obtain the instruction analysis state includes: identifying indicators in the data analysis instruction and determining whether candidate data indicators are identified; if candidate data indicators are identified, and the candidate data indicators are ambiguous or similar, then first ambiguity is determined; if no candidate data indicators are identified, then it is determined whether the user specifies an analysis direction; if the user specifies an analysis direction, then based on a preset indicator recommendation model, candidate data indicators corresponding to the analysis direction are determined, and second ambiguity is determined; if no candidate data indicators are identified, and the user does not specify an analysis direction, then third ambiguity is determined; and the instruction analysis state is generated based on the existence of first, second, or third ambiguity, and whether any parameters are missing.

[0053] Among them, ambiguous indicators in the first ambiguity refer to indicators whose names entered by users have multiple matching formal indicator fields in the database with different meanings. For example, "PE" can correspond to multiple fields such as "static PE", "dynamic PE", and "rolling PE". Similar indicators refer to indicators whose names are semantically similar but have different calculation methods. For example, "profit" may correspond to "net profit attributable to parent company", "net profit excluding non-recurring items", and "operating profit".

[0054] For example, in this embodiment, the preset disambiguation confidence threshold is 85%. The confidence calculation model is built based on a bidirectional semantic matching network. The specific calculation rules are as follows: extract three types of feature vectors from the user data analysis instructions: core terms, industry of the analysis context, and historical question indicator labels; calculate the cosine similarity between each ambiguous indicator and the above three types of feature vectors to obtain the individual confidence; perform weighted summation with an industry matching weight of 0.5, a term matching weight of 0.3, and a historical preference weight of 0.2 to obtain the comprehensive confidence; if the highest comprehensive confidence is ≥85%, the candidate indicator is set to a pre-selected highlighted state, requiring only one confirmation from the user; if <85%, all candidate indicators are displayed in full for the user to choose from; for cold start users with no history, automatic disambiguation is turned off by default, all candidate indicators are directly displayed, and the user's selection is stored in the historical analysis record for subsequent confidence calculation.

[0055] Furthermore, the preset threshold can be determined based on the system's backend user interaction logs by statistically analyzing the selection probability of candidate indicators corresponding to different contexts. It can also be configured differently according to user type: professional financial users can have a lower threshold (e.g., 70%) to improve efficiency, while novice users can have a higher threshold (e.g., 95%) to ensure accuracy. The threshold will also be dynamically adjusted according to the semantic differences of ambiguous indicators; it will be automatically increased if the difference in indicator meaning is large, and appropriately decreased if only the time caliber differs.

[0056] Furthermore, the second ambiguity refers to a user expressing a desired research topic but not specifying concrete indicators, such as "profitability" or "valuation level." Based on a pre-defined indicator recommendation model, the analysis direction is mapped to a set of candidate data indicators. This model is built upon a financial analysis expertise graph or historical user query behavior. The third ambiguity refers to incomplete user input, such as "Analyze it for me." For the third ambiguity, due to the lack of sufficient semantic information in the user input for accurate matching, the indicator confirmation card employs a guided recommendation strategy based on popular analysis topics or user profiles. The system can obtain a list of frequently queried analysis topics on the platform (such as "Financial Analysis of Popular Stocks," "Industry Valuation Comparison," "Recent Price Increase Ranking," etc.) or generate personalized recommendation directions based on the user's historical query preferences. These directions are presented as guided options on the card, helping users gradually converge from incomplete expressions to specific analysis topics. For example, the card might display guided options such as "You might be interested in: Viewing the profitability of your holdings" or "Understanding the valuation levels of recently popular industries." Regarding parameter missing detection, when a missing parameter is detected, the system first automatically fills in the default value based on historical analysis records or industry standard data. For example, the default reporting period is the most recent complete financial reporting period, and the default unit setting is a commonly used unit in the industry or a unit frequently used by the user. After the automatic filling is completed, a parameter supplement card is generated, and the parameters are displayed in a pre-filled state. The user only needs to confirm or make minor adjustments.

[0057] This embodiment achieves accurate classification and processing of intent ambiguity through hierarchical progressive judgment and intelligent disambiguation and filling: First, the instruction is identified by indicator recognition. If a candidate indicator is identified, ambiguity is detected and automatic disambiguation is initiated. If the confidence level is insufficient, the first ambiguity is triggered. If no candidate indicator is identified, the analysis direction is further determined, and the corresponding second or third ambiguity is marked. After the ambiguity is determined, the completeness of the parameters is checked. Missing parameters are automatically filled with default values ​​according to historical records or industry standards. Finally, the ambiguity type and parameter status are merged to generate a complete instruction analysis status.

[0058] As can be seen in this example, by subdividing intent ambiguity into three types and establishing a hierarchical judgment mechanism, combined with automatic disambiguation and automatic parameter filling based on historical records and context, the frequency of user interaction can be significantly reduced while ensuring accuracy. The system only generates the corresponding interaction card when the confidence of automatic disambiguation is insufficient or when the automatic filling result requires user confirmation, thus achieving a balance between intelligent automation and manual confirmation. This reduces the burden on users to fill in each item one by one and avoids the deviation of analytical intent caused by excessive automation, effectively improving the accuracy and operational efficiency of the data analysis instruction understanding process.

[0059] In step S303, if the instruction analysis status indicates that there is intent ambiguity or missing parameters, then an intent interaction card is generated and displayed.

[0060] In this context, the intent interaction card refers to a front-end interface interaction component used to guide users in confirming candidate data indicators and / or supplementing missing data parameters. The intent interaction card is dynamically generated and presented on the front-end interface in the form of a structured card. The card may contain interface elements such as single-choice options, multiple-choice options, text input boxes, dropdown selectors, and default pre-filled values. In terms of interaction methods, in addition to visual card display and touch operations, when the system detects intent ambiguity, it can also use a client-side voice broadcasting device to ask the user a follow-up question, broadcasting the ambiguous option content to the user auditorily. The user can confirm or correct the candidate data indicator by answering with their voice. The voice recognition module converts the user's voice answer into a corresponding selection command, thus providing a voice-assisted interaction channel in addition to visual card interaction. The intent interaction card generation mechanism constructed in this application realizes "state-aware on-demand generation," meaning that card generation is not pre-placed in the interaction process but rather serves as a response to the preceding "instruction analysis state." By monitoring and analyzing the status in real time, the corresponding card generation module is invoked only when the status indicator indicates specific conditions such as "ambiguous intent" or "missing parameters," instantiating and rendering the corresponding intent interaction card in the front-end interface. This condition-triggered and on-demand generation mechanism based on instruction status ensures the accuracy and minimal necessity of intelligent human-computer interaction. For example, when the instruction analysis status simultaneously indicates both intention ambiguity and missing parameters, multiple cards can be generated sequentially to guide the confirmation process step by step.

[0061] In one possible embodiment, the intent interaction card includes an indicator confirmation card, a parameter supplementation card, and a data extraction confirmation card. The indicator confirmation card is a front-end interface component used to display candidate data indicators and receive user confirmation or correction operations for these indicators. The parameter supplementation card is a front-end interface component used to receive user input of missing data parameters. The data extraction confirmation card is a front-end interface component used to display data extraction rules composed of target data indicators and target data parameters and receive user confirmation of these rules. Generating and displaying the intent interaction card includes: generating and displaying an indicator confirmation card when the instruction analysis status indicates ambiguity; generating and displaying a parameter supplementation card when the instruction analysis status indicates missing parameters; and generating and displaying a data extraction confirmation card after confirming the target data indicators and target data parameters.

[0062] The intent interaction card generation mechanism constructed in this application realizes "adaptive intelligent filling of card content". That is, the specific content of each card is not a static template, but is dynamically and adaptively constructed according to the underlying reasons that trigger its generation (i.e., the specific type of ambiguity, the specific category of missing parameters) and the available context.

[0063] Specifically, the indicator confirmation card presents differentiated content based on the type of intent ambiguity (i.e., adaptive ambiguity resolution). For the first ambiguity, the card lists multiple alternative indicators and their explanations; users can confirm the target indicator by clicking to select or speaking the option number. For the second ambiguity, the card displays a set of relevant indicators recommended by the indicator recommendation model based on the analysis direction. For the third ambiguity, the card uses guided interaction to help users gradually clarify their analysis intent. Regarding voice assistance, when there are many ambiguous options or the user has difficulty seeing, the system can read aloud the numbers and names of each alternative indicator in sequence, allowing the user to confirm by speaking the selected indicator name or number. The parameter supplement card dynamically loads corresponding parameter setting components for different types of missing parameters. Each parameter component displays automatically filled default values ​​in a pre-filled state on the card, allowing users to directly confirm or fine-tune (i.e., intelligent parameter completion). The data extraction confirmation card is generated after both indicators and parameters have been confirmed, displaying the complete data extraction rules in a structured summary format. User confirmation directly triggers the data extraction operation.

[0064] The voice assistance can be triggered by displaying a card voice icon on the user interface of the client 110, and triggering the call when the user clicks the card voice icon. Alternatively, it can be automatically triggered when the client 110 detects that it is in vehicle mode, hands-free mode, or accessibility mode. The voice prompts are read out line by line, in the order of "serial number, indicator name, and brief explanation," with a preset interval between each line. Interruptions and re-reading are supported. If a voice command and a touch command are input simultaneously, the later command takes precedence. If the voice recognition confidence level is lower than a preset value, the system automatically returns to the card for reselection.

[0065] For example, please refer to Figure 4 , Figure 4 This is a schematic diagram of a visual interface for ambiguous instructions provided in an embodiment of this application. For example... Figure 4 As shown, this interface is the front-end interactive interface displayed by client 110, fully realizing the entire process of visual interaction, including clarifying intent ambiguity, supplementing missing parameters, and confirming data extraction rules. First is the instruction input area: the top of the interface has a "Data Analysis Instruction" input box, and the left display bar shows the data analysis instruction currently entered by the user. For example, if the user enters a fuzzy natural language data analysis instruction (such as "Help me check the recent profitability of leading companies in the liquor industry"), since this instruction does not specify specific data indicators and only specifies the analysis direction, the system, through intent understanding, determines it as a second ambiguity. Simultaneously, it detects missing parameters such as the reporting period and data unit, triggering the interactive card generation process.

[0066] Furthermore, in response to the second ambiguity, client 110 will display an indicator confirmation card, which pops up in the middle of the interface. This is a front-end interactive component generated by the system to address the second ambiguity. The card displays candidate data indicators (such as net profit attributable to the parent company and gross profit margin) recommended by a preset indicator recommendation model based on the "profitability" analysis direction. It is used to receive user confirmation / correction operations for candidate data indicators. Users can select the target indicator by touch, and it also supports auxiliary interaction methods such as voice broadcast of alternative indicators and voice response to selection to complete the confirmation of the target data indicator from the candidate data indicator. After the indicator is confirmed, the interface will then pop up a parameter supplement card, which is a front-end interactive component generated by the system to address missing parameters. The card dynamically loads parameter setting components such as data range, time range, data unit, and statistical method (the missing parameter displayed here is the "data range parameter"). The display box next to it is automatically filled with default values ​​(such as "the last four quarters", the specific content of which corresponds to the missing parameter) and pre-filled and displayed by the system. Users can directly confirm or fine-tune to complete the supplementation of the target data parameters.

[0067] Finally, after the target data metrics and parameters are confirmed, a data extraction confirmation card pops up on the interface. This is a front-end interactive component used by the system to display the data extraction rules. The card displays the complete data extraction rules, consisting of the target data metrics and target data parameters, in a structured format. Below the target data parameters, there is a drop-down selection control. When the user clicks this control, a drop-down selection control will appear (containing other optional data parameters corresponding to the target data metrics, such as...). Figure 4 (The data unit, time range, and statistical method are shown). When the user selects at least one of the optional controls in the drop-down selection control, the display bar corresponding to the target data parameter will be supplemented with the optional data parameter corresponding to the target optional control (that is, the user can also temporarily update the target data parameter by selecting the data extraction card). The user can directly trigger the background data extraction operation by performing a confirmation operation.

[0068] As can be seen, this example, through a state-driven card sequence generation mechanism, transforms the user's vague, proactive statements into system-guided, step-by-step confirmations, converting the burden of professional cognitive understanding into a low-cost interactive operation. Simultaneously, by supporting multimodal interaction adaptation with voice-guided follow-up questions and voice responses, it further enhances the flexibility and accessibility of the interaction. Compared to traditional human-computer interaction models, it efficiently and reliably completes the migration from vague intentions to precise query rules, improving data analysis efficiency.

[0069] Step S304: In response to the user's interactive operation, confirm or correct the candidate data indicators to obtain the target data indicators, and / or supplement the target data parameters corresponding to the target data indicators.

[0070] In one possible embodiment, in response to a user's interactive operation, confirming or correcting candidate data indicators to obtain a target data indicator, and / or supplementing the target data parameters corresponding to the target data indicator, includes: receiving a user's confirmation operation on the candidate data indicators displayed in the indicator confirmation card, and determining the candidate data indicators as target data indicators; or, receiving a user's correction operation on the candidate data indicators displayed in the indicator confirmation card, and determining the corrected candidate data indicators as target data indicators; receiving missing data parameters input by the user in the parameter supplementation card, and determining the missing data parameters as target data parameters; receiving a user's confirmation operation on the data extraction rules displayed in the data extraction confirmation card, and triggering a data extraction operation based on the confirmation operation.

[0071] The user interaction is divided into different processing paths based on the initial state of the candidate data indicators. When the candidate data indicator is not empty, the indicator confirmation card displays the candidate indicators and alternatives. The user can directly click to confirm the candidate data indicator as the target data indicator, or select another indicator from the alternatives as the corrected target data indicator. When the candidate data indicator is empty, the user's action of selecting the desired indicator from the recommended list is the correction, or the user can manually enter another indicator name to confirm the target data indicator. For the parameter supplement card, the user can directly confirm the automatically filled default parameter values, or fine-tune or re-enter the target data parameters. For the data extraction confirmation card, the user's confirmation action triggers the subsequent data extraction process.

[0072] As can be seen in this example, by incorporating the model recommendation selection or manual input operation when the candidate data indicators are empty into the "correction" category, the system can guide users from vague intentions to precise data extraction rules through a unified confirmation or correction interaction path, regardless of whether the user's initial input is an indicator name, analysis direction, or a completely vague expression. This reduces the risk of analysis interruption caused by incomplete initial input and improves the scenario coverage completeness and interaction robustness of the solution.

[0073] Step S305: Perform data extraction operation based on target data indicators and target data parameters to obtain raw analysis data, and display the raw analysis data in the form of analysis display.

[0074] The analysis and display format refers to the front-end interface format for visualizing and interactively analyzing raw analysis data. Data extraction refers to the automatic construction of database query statements based on confirmed target data indicators and parameters, retrieving matching data record sets from the database or data interface. Raw analysis data refers to the collection of raw data rows directly extracted from the database without front-end rendering. The analysis and display format is dynamically selected based on the multidimensional characteristics of the raw analysis data, rather than using a uniform fixed display template, ensuring that data with different characteristics is presented in the most suitable chart type.

[0075] In one possible implementation, when the raw analysis data returned by the data extraction operation is empty, the system performs special handling for the empty data situation. This handling includes displaying a prompt message in the analysis display area explaining to the user why no matching data was found.

[0076] For example, in this embodiment, the empty data processing adopts a closed-loop mechanism that combines cause classification and judgment with guided correction suggestions. First, the reasons for empty data are classified: if the reporting period is earlier than the listing date of the target, it is judged that the target is not listed; if there is no corresponding indicator record in the database, it is judged that the indicator is not disclosed; if the time interval is outside the valid range of data, it is judged that the time interval is invalid; if the applicable industry of the indicator does not match the current target industry, it is judged that the industry is mismatched. Then, 1-3 optimal correction suggestions are automatically generated according to different reasons, including recommending alternative targets, expanding the time range, changing the appropriate indicator, and adjusting the analysis industry. The correction suggestions are displayed in the form of clickable buttons. After the user clicks, the system automatically fills in the new parameters and re-initiates the data extraction operation to avoid interruption of the analysis process.

[0077] In one possible embodiment, displaying raw analytical data in an analytical presentation format includes: determining the data characteristics of the raw analytical data, the data characteristics including at least one of data type, data dimension, and data volume; and determining an analytical presentation format from multiple preset analytical presentation formats based on a combination of the data type, data dimension, and data volume of the raw analytical data.

[0078] The data types include those with object identifiers and those without object identifiers. The type with object identifiers is used to indicate that the original analysis data contains an analysis object identifier, while the type without object identifiers is used to indicate that the original analysis data is general statistical data. The data dimensions include the number of indicators, the number of target codes, and the number of analysis periods. Multiple preset analysis display formats include at least one of the following: data table format, trend chart format, comparison chart format, and comprehensive dashboard format.

[0079] In the data type classification, types with object identifiers refer to data rows containing explicit financial target codes or names, such as stock codes, bond codes, and fund codes. This type of data supports grouping and comparison or trend tracking by target. Types without object identifiers refer to macroeconomic statistical data or industry summary data, not associated with a single analytical object, such as industry prosperity. In terms of data dimensions, the indicator quantity dimension is divided into single indicator and multiple indicator; the target code quantity dimension is divided into single code and multiple code; and the analysis period quantity dimension is divided into single period (cross-sectional data) and multiple period (time series data). The cross-combination of various dimensions forms different data formats, typical formats include: cross-sectional data of a single indicator, single code, and single period, suitable for display in tabular form; time-series data of a single indicator, single code, and multiple periods, suitable for display in trend chart form; horizontal comparison data of a single indicator, multiple codes, and single period, suitable for display in bar chart form; panel data of a single indicator, multiple codes, and multiple periods, suitable for display in multi-series trend charts or comprehensive dashboard form; multi-indicator, single code, and multiple period data, suitable for display in multi-indicator comprehensive dashboard form; and multi-indicator, multi-code data, suitable for display in comprehensive dashboard form combined with filters. The display format matching rules can be implemented based on a preset mapping table, or automatically selected by a trained classification model based on data characteristics.

[0080] For example, please refer to Figure 5 , Figure 5 This is a schematic diagram of a visualization interface based on analysis and display provided in an embodiment of this application. For example... Figure 5As shown, the system determines the appropriate visualization scheme from preset analysis display formats based on the data characteristics (combination of data type, data dimension, and data volume) of the original analysis data. Specifically, these include: Data Table Format: Corresponding to the first sub-plot in the image, this format is suitable for cross-sectional data of a single indicator, single code, and single period; structured detailed data; or general statistical data without object identifiers. It clearly presents precise values ​​in a row-column table format, suitable for scenarios where users need to obtain explicit data results. Especially on small-screen terminals such as mobile devices, this format ensures data readability with low information density. Trend Chart Format: Corresponding to the second sub-plot in the image, this format is suitable for time-series data of a single indicator, single code, and multiple periods (such as multi-quarter net profit data of a target). It intuitively displays the data's trend over time in the form of a line chart or area chart, highlighting the continuity and evolution of the data, and helping users quickly identify characteristics such as growth and fluctuations. Comparison Chart Format: Corresponding to the third sub-chart in the figure, this format adapts to horizontal comparison data of single indicators, multiple codes, and single periods (such as comparing the single-period profits of multiple targets in the same industry). It presents the numerical differences between different analytical objects in bar chart format, allowing users to intuitively compare the relative performance and industry ranking of the targets. Comprehensive Dashboard Format: Corresponding to the fourth sub-chart in the figure, this format adapts to multi-indicator, multi-code, multi-indicator, or complex panel data. It integrates trend charts, comparison charts, key statistical values, and text summaries to achieve a multi-dimensional information aggregation display. This format is primarily used on large-screen desktop devices to fully utilize screen space to present a global analytical perspective and improve analytical efficiency in complex scenarios.

[0081] As can be seen in this example, the display format is determined by extracting the data's type, dimension, and magnitude characteristics. This allows different types of data to automatically adapt to the most suitable visualization charts, avoiding the limitations of traditional fixed templates in displaying diverse data. For example, time-series data is automatically presented as trend charts to intuitively reflect changing patterns, cross-sectional comparison data is automatically presented as comparison charts to highlight differences, and multi-dimensional complex data is automatically aggregated into comprehensive dashboards for global analysis. This intelligent matching of data characteristics and display formats reduces the burden on users to manually switch chart types, making the information delivery of analysis results more intuitive and efficient.

[0082] In one possible embodiment, after determining the analysis display format from multiple preset analysis display formats based on a combination of the data type, data dimension, and data volume of the original analysis data, the method further includes: generating a data summary text corresponding to the original analysis data; and displaying the data summary text in the analysis display format.

[0083] The data summary text is used to describe the core characteristics of the original analysis data in natural language. The system calculates key indicators for the original analysis data, including statistics such as maximum, minimum, average, growth rate, and ranking. Combining the time span of the data and information about the analysis object, it uses natural language generation templates to transform the statistical results into readable descriptive statements. For example, for the net profit data of a stock over multiple periods, it can generate a summary description such as, "The net profit attributable to the parent company of this stock has shown a quarterly upward trend over the past four quarters, with the latest period showing a year-on-year increase of 15.3%, reaching a new high in nearly two years." The data summary text is displayed in preset locations in the analysis presentation, such as at the bottom of a card or in the chart title area, allowing users to quickly obtain the core conclusions of the data without having to interpret the original charts themselves.

[0084] For example, please refer to Figure 6 , Figure 6 This is a schematic diagram of a visual interface for intelligent recommendation guidance provided in an embodiment of this application. For example... Figure 6 As shown, the interface adopts a three-column layout: the left side is a control button bar for displaying information, the middle part is the main visual area and the bottom text summary area, and the right side is the intelligent recommendation bar. Each area works together to realize the core functions of the two embodiments described in this application.

[0085] The central main visualization area (adaptive analysis display format) automatically matches and determines the analysis display format based on the combination of data type, data dimension, and data volume of the original analysis data. In this example, for time series data with a single indicator, single code, and multiple periods, the data's changing patterns over time are presented intuitively in the form of a trend chart. The display format control button bar on the left provides entry points for switching between preset display formats such as data tables, trend charts, comparison charts, and comprehensive dashboards. Users can also manually adjust the display mode, achieving a combination of "data feature-driven adaptive visualization" and "user-defined switching," avoiding the limitations of traditional fixed templates in displaying diverse data.

[0086] The bottom text summary area (data summary text generation example) is located below the main visualization area, corresponding to the data summary text generation example described in this application. The system calculates key indicators (including maximum, minimum, average, growth rate, year-on-year / month-on-month change rate, etc.) from the original analysis data, and automatically generates data summary text using a natural language generation template, combining the time span of the data and the information of the analysis object, and displays it in this area in the form of "AI summary".

[0087] The right-hand intelligent recommendation bar (intelligent recommendation guidance embodiment) is set on the right side of the interface. It displays recommended analysis models or data indicators related to the current analysis topic in the form of recommendation item control buttons, corresponding to the intelligent recommendation guidance embodiment described in this application. The recommended content is generated by the system based on three logics: 1. Data characteristics of the current original analysis data; 2. Semantic association weights between indicators in the financial analysis knowledge graph; 3. The user's historical analysis records (including historical questions and query results). After merging and deduplication, a recommendation list with the highest relevance is generated, which may include related data indicators, supplementary analysis models, industry comparative analysis, and other options. When the user triggers the recommendation item control button, the system automatically executes the context association logic: if the parameters required for the new analysis intent are compatible with the target code, time range, and other parameters in the current context, the context parameters are directly inherited, achieving referential resolution and parameter reuse without requiring repeated user input; if there are parameter conflicts or missing parameters, a lightweight parameter confirmation / supplementation card is generated to guide the user to correct them. Finally, based on the associated analysis intent, a new round of data extraction and analysis display is completed, achieving seamless connection from a single query to multiple rounds of in-depth analysis, effectively improving the coherence and depth of the analysis.

[0088] As can be seen in this example, by attaching automatically generated text summaries to the visual charts, users can quickly grasp the key trends and anomalies in the data, reducing the cognitive burden of manually interpreting the charts. This is especially effective in scenarios involving massive data analysis, significantly improving the efficiency of information extraction and the readability of analytical conclusions.

[0089] In one possible embodiment, after displaying the raw analysis data in an analytical presentation format, the method further includes: generating a recommended analysis model or recommended data indicator associated with the current analysis topic based on the data characteristics of the raw analysis data and the user's historical analysis records, and displaying it in a preset area of ​​the analytical presentation format; when a user triggers an operation on the recommended analysis model or recommended data indicator, associating the current analysis context with the new analysis intent corresponding to the trigger operation, and performing a new round of data extraction and analytical presentation based on the data indicators and data parameters corresponding to the associated analysis intent.

[0090] The historical analysis records include users' historical question data and historical query result data; the recommendation analysis model is a preset analysis framework related to the current analysis topic, including valuation analysis model and financial health scoring model; the recommended data indicators are indicators that are related to or complementary to the currently queried indicators, such as recommending gross profit margin and return on net assets after querying net profit. The recommended content is displayed in the preset area of ​​the analysis display interface in the form of tags, buttons or card lists.

[0091] Furthermore, when a user triggers recommended content, the system matches the current context's target code, time range, statistical units, and other parameters with the new analytical intent, executing a three-level parameter determination logic: If the parameters required by the new analytical intent are completely consistent with the current context parameters, all parameters are directly inherited, achieving referential resolution and context reuse; if the industry to which the new indicator belongs is incompatible with the current target's industry, or the statistical caliber is not supported, it is determined to be a parameter conflict, and a lightweight parameter confirmation card with pre-filled conflicting parameters pops up, allowing the user to correct and continue execution; if the new analytical intent has missing parameters that the context cannot inherit, such as report type or adjustment method, a parameter supplement card with pre-filled industry default values ​​pops up, requiring the user to supplement and confirm; the parameter execution priority is: user-manually modified parameters > context-inherited parameters > default-filled parameters.

[0092] For example, in this embodiment, the recommendation list adopts a fusion sorting rule: the knowledge graph semantic association score is calculated according to the complementary relationship weight of indicators (1.0), the industry association weight (0.7), and the general association weight (0.3); the user collaborative filtering score is the standard score obtained by normalizing the query frequency of subsequent indicators after completing the query of the current indicator for the same user group with the same industry attribute and the same analysis frequency; the fusion score calculation formula is: fusion score = knowledge graph score × 0.6 + collaborative filtering score × 0.4. Then, the fusion scores are sorted from high to low and duplicates are removed. The top seven recommended items are selected to generate the recommendation list. If the current terminal is a small-screen mobile device, it is automatically reduced to the top five recommended items for display.

[0093] As can be seen, in the process of this application embodiment, the complex operations that users need to complete manually in traditional data analysis, such as indicator retrieval, parameter filling, and chart selection, are transformed into a lightweight interaction guided by the system. When the user's intention is ambiguous, the interaction cards quickly converge to the precise extraction rules. After data extraction, the optimal visualization form is automatically matched and accompanied by text summaries to aid understanding. After the analysis is completed, related indicators and analysis models are proactively recommended based on context to support multiple rounds of in-depth exploration. The entire process does not require users to master database query languages ​​or data architecture, significantly reducing the usage threshold of financial data analysis and improving analysis efficiency and the intuitiveness of results.

[0094] The following are embodiments of the apparatus of this application. These embodiments of the apparatus and the embodiments of the method of this application belong to the same concept and are used to execute the methods described in the embodiments of this application. For ease of explanation, only the parts related to the apparatus embodiments of this application are shown in the embodiments of this application. For specific technical details not disclosed, please refer to the description of the embodiments of the method of this application, which will not be repeated here.

[0095] This application provides a data analysis device based on an interactive smart card. Specifically, the data analysis device based on an interactive smart card is used to execute the steps corresponding to the above-described data analysis method based on an interactive smart card. The data analysis device based on an interactive smart card provided in this application may include modules corresponding to the respective steps.

[0096] This application embodiment can divide the data analysis device based on interactive smart cards into functional modules according to the above method example. For example, each function can be divided into its own functional module, or two or more functions can be integrated into one processing module. The integrated module can be implemented in hardware or as a software functional module. The module division in this application embodiment is illustrative and only represents one logical functional division; other division methods may be used in actual implementation.

[0097] When dividing each function into modules according to its corresponding function. Figure 7 This is a functional unit block diagram of a data analysis device based on an interactive smart card, provided in an embodiment of this application; the data analysis device based on an interactive smart card is applied to... Figure 1 or Figure 2 The client 110 in the intelligent interactive data analysis system 10 shown. The data analysis device 70 based on interactive smart cards includes: an instruction acquisition unit 701, used to acquire data analysis instructions input by the user; a status analysis unit 702, used to understand the intent of the data analysis instructions and obtain the instruction analysis status, which indicates whether there is intent ambiguity or missing parameters in the data analysis instructions, as well as candidate data indicators identified from the data analysis instructions; a card generation unit 703, used to generate and display intent interaction cards if the instruction analysis status indicates intent ambiguity or missing parameters, the intent interaction cards being front-end interface interaction components used to guide the user to confirm candidate data indicators and / or supplement missing data parameters; an interaction response unit 704, used to respond to the user's interaction operation, confirm or correct candidate data indicators to obtain target data indicators, and / or supplement target data parameters corresponding to the target data indicators; and a data display unit 705, used to perform data extraction operations according to the target data indicators and target data parameters to obtain raw analysis data, and display the raw analysis data in an analysis display form, the analysis display form being a front-end interface form for visual display and interactive analysis of the raw analysis data.

[0098] In one possible example, intent ambiguity is categorized into first ambiguity, second ambiguity, or third ambiguity based on the type of ambiguity. First ambiguity indicates that there are ambiguous or similar indicators among the candidate data indicators; second ambiguity indicates that the user specifies an analysis direction but does not specify a candidate data indicator; and third ambiguity indicates that the user does not specify either a candidate data indicator or an analysis direction. Regarding intent understanding of data analysis instructions to obtain the instruction analysis state, the state analysis unit 702 is specifically used for: indicator identification of the data analysis instruction, determining whether candidate data indicators are identified; if candidate data indicators are identified, and the candidate data indicators are ambiguous or similar, then first ambiguity is determined; if no candidate data indicators are identified, then it is determined whether the user specifies an analysis direction: if the user specifies an analysis direction, then based on a preset indicator recommendation model, candidate data indicators corresponding to the analysis direction are determined, and second ambiguity is determined; if no candidate data indicators are identified, and the user does not specify an analysis direction, then third ambiguity is determined; based on the existence of first, second, or third ambiguity, and whether parameters are missing, an instruction analysis state is generated.

[0099] In one possible example, the intent interaction card includes an indicator confirmation card, a parameter supplementation card, and a data extraction confirmation card. The indicator confirmation card is a front-end interface component used to display candidate data indicators and receive user confirmation or correction operations for these indicators. The parameter supplementation card is a front-end interface component used to receive user input of missing data parameters. The data extraction confirmation card is a front-end interface component used to display data extraction rules composed of target data indicators and target data parameters and receive user confirmation of these rules. Regarding the generation and display of intent interaction cards, the card generation unit 703 is specifically used to: generate and display an indicator confirmation card when the instruction analysis status indication has intent ambiguity; generate and display a parameter supplementation card when the instruction analysis status indication has missing parameters; and generate and display a data extraction confirmation card after confirming the target data indicators and target data parameters.

[0100] In one possible example, in responding to a user's interaction to confirm or correct a candidate data indicator to obtain a target data indicator, and / or to supplement the target data parameters corresponding to the target data indicator, the interaction response unit 704 is specifically configured to: receive a user's confirmation operation on a candidate data indicator displayed in the indicator confirmation card, and determine the candidate data indicator as the target data indicator; or, receive a user's correction operation on a candidate data indicator displayed in the indicator confirmation card, and determine the corrected candidate data indicator as the target data indicator; receive missing data parameters entered by the user in the parameter supplement card, and determine the missing data parameters as target data parameters; receive a user's confirmation operation on the data extraction rules displayed in the data extraction confirmation card, and trigger a data extraction operation based on the confirmation operation.

[0101] In one possible example, regarding the display of raw analytical data in an analytical presentation format, the data presentation unit 705 is specifically used to: determine the data characteristics of the raw analytical data, which include at least one of data type, data dimension, and data volume; the data type includes types with object identifiers and types without object identifiers; the type with object identifiers is used to characterize that the raw analytical data contains analytical object identifiers, and the type without object identifiers is used to characterize that the raw analytical data is general statistical data; the data dimension includes the dimension of the number of indicators, the dimension of the number of target codes, and the dimension of the number of analysis periods; and determine the analytical presentation format from multiple preset analytical presentation formats based on the combination of the data type, data dimension, and data volume of the raw analytical data, which include at least one of the following: data table format, trend chart format, comparison chart format, and comprehensive dashboard format.

[0102] In one possible example, after determining the analysis display format from multiple preset analysis display formats based on the combination of data type, data dimension, and data volume of the original analysis data, the data display unit 705 is further used to: generate a data summary text corresponding to the original analysis data, the data summary text being used to describe the core features of the original analysis data in natural language; and display the data summary text in the analysis display format.

[0103] In one possible example, after displaying the raw analysis data in an analytical presentation format, the data presentation unit 705 is further used to: generate a recommendation analysis model or recommendation data indicator associated with the current analysis topic based on the data characteristics of the raw analysis data and the user's historical analysis records, and display it in a preset area of ​​the analytical presentation format. The historical analysis records include the user's historical question data and historical query result data. When a user triggers an operation on the recommendation analysis model or recommendation data indicator, the current analysis context is associated with the new analysis intent corresponding to the trigger operation, and a new round of data extraction and analysis presentation is performed based on the data indicators and data parameters corresponding to the associated analysis intent.

[0104] When using integrated units, such as Figure 8 As shown, Figure 8 This is a functional unit block diagram of another data analysis device based on an interactive smart card provided in this application embodiment. Figure 8The interactive smart card-based data analysis device 70 includes a processing module 802 and a communication module 801. The processing module 802 controls and manages the actions of the interactive smart card-based data analysis device 70, such as the steps of the instruction acquisition unit 701, the status analysis unit 702, the card generation unit 703, the interaction response unit 704, and the data display unit 705, and / or performs other processes described herein. The communication module 801 supports interaction between the interactive smart card-based data analysis device and other devices. Figure 8 As shown, the data analysis device based on interactive smart cards may include a storage module 803, which is used to store the program code and data of the data analysis device based on interactive smart cards.

[0105] The processing module 802 can be a processor or processing module, such as a central processing unit, a general-purpose processor, a digital signal processor, an ASIC, an FPGA, or other programmable logic devices, transistor logic devices, hardware components, or any combination thereof. It can implement or execute the various exemplary logic blocks, modules, and circuits described in conjunction with the disclosure of this application. The processor can also be a combination that implements computational functions, such as a combination of one or more microprocessors, a combination of a DSP and a microprocessor, etc. The communication module 801 can be a transceiver, RF circuitry, or a communication interface, etc. The storage module 803 can be a memory.

[0106] All relevant content in each scenario involved in the above method embodiments can be referenced from the functional descriptions of the corresponding functional modules, and will not be repeated here. The above-mentioned data analysis device 70 based on interactive smart cards can all perform the above-mentioned... Figure 3 The data analysis method based on interactive smart cards is shown.

[0107] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, as a computer program product. A computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, all or part of the processes or functions according to the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. Computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired or wireless means. A computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more sets of available media. Available media can be magnetic media (e.g., floppy disks, hard disks, magnetic tapes), optical media (e.g., DVDs), or semiconductor media. Semiconductor media can be solid-state drives.

[0108] Figure 9 This is a structural block diagram of an electronic device provided in an embodiment of this application. For example... Figure 9 As shown, the electronic device 90 may include one or more of the following components: a processor 901 and a memory 902 coupled to the processor 901, wherein the memory 902 may store one or more computer programs 903, and the one or more computer programs 903 may be configured to implement the methods described in the above embodiments when executed by one or more processors 901. The electronic device 90 here is the client 110 in the above embodiments.

[0109] Processor 901 may include one or more processing cores. Processor 901 connects to various parts within the electronic device 90 using various interfaces and lines, and performs various functions and processes data of the electronic device 90 by running or executing instructions, programs, code sets, or instruction sets stored in memory 902, and by calling data stored in memory 902. Optionally, processor 901 may be implemented using at least one hardware form of Digital Signal Processing (DSP), Field-Programmable Gate Array (FPGA), or Programmable Logic Array (PLA). Processor 901 may integrate one or more of the following: central processing unit, image processor, and modem.

[0110] The memory 902 may include random access memory (RAM) or read-only memory (ROM). The memory 902 can be used to store instructions, programs, code, code sets, or instruction sets. The memory 902 may include a program storage area and a data storage area. The program storage area may store instructions for implementing an operating system, instructions for implementing at least one function (such as touch functionality, sound playback functionality, image playback functionality, etc.), and instructions for implementing the various method embodiments described above. The data storage area may also store data created by the electronic device 90 during use.

[0111] It is understood that the electronic device 90 may include more or fewer structural elements than those shown in the above block diagram, without limitation herein.

[0112] This application also provides a computer storage medium storing a computer program / instructions thereon, which, when executed by a processor, implements some or all of the steps of any of the methods described in the above method embodiments.

[0113] This application also provides a computer program product, which includes a non-transitory computer-readable storage medium storing a computer program operable to cause a computer to perform some or all of the steps of any of the methods described in the above method embodiments.

[0114] It should be understood that in the various embodiments of this application, the order of the above-mentioned processes does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0115] In the several embodiments provided in this application, it should be understood that the disclosed methods, apparatuses, and systems can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for example, the division of units is merely a logical functional division, and there may be other division methods in actual implementation; for example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces, and the indirect coupling or communication connection between devices or units may be electrical, mechanical, or other forms.

[0116] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0117] Furthermore, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can be physically comprised separately, or two or more units can be integrated into one unit. The integrated unit described above can be implemented in hardware or in the form of hardware plus software functional units.

[0118] The integrated units implemented as software functional units described above can be stored in a computer-readable storage medium. These software functional units, stored in a storage medium, include several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute partial steps of the methods of the various embodiments of the present invention. The aforementioned storage medium includes: a USB flash drive, a portable hard drive, a magnetic disk, an optical disk, volatile memory, or non-volatile memory. The non-volatile memory can be read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), or flash memory. The volatile memory can be random access memory (RAM), which is used as an external cache. By way of example, but not limitation, many forms of random access memory (RAM) are available, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDR SDRAM), enhanced SDRAM (ESDRAM), synchronous linked DRAM (SLDRAM), and direct rambus RAM (DR RAM), etc., which are various media capable of storing program code.

[0119] While the present invention has been disclosed above, it is not limited thereto. Any person skilled in the art can easily conceive of variations or substitutions without departing from the spirit and scope of the present invention, and various modifications and alterations can be made, including combinations of the different functions and implementation steps described above, as well as software and hardware implementation methods, all of which are within the protection scope of the present invention.

Claims

1. A data analysis method based on interactive smart cards, characterized in that, include: Obtain data analysis commands input by the user; The intent of the data analysis instruction is understood to obtain the instruction analysis status. The instruction analysis status is used to indicate whether there is intent ambiguity or missing parameters in the data analysis instruction, as well as the candidate data indicators identified from the data analysis instruction. The intent ambiguity is categorized into first ambiguity, second ambiguity, and third ambiguity based on the type of ambiguity. The process of understanding the intent of the data analysis instruction to obtain the instruction analysis state includes: identifying indicators in the data analysis instruction and determining whether candidate data indicators are identified; if candidate data indicators are identified, and the candidate data indicators contain ambiguous or similar indicators, then a first ambiguity is determined; if no candidate data indicators are identified, then it is determined whether the user has specified an analysis direction; if the user has specified an analysis direction, then a candidate data indicator corresponding to the analysis direction is determined based on a preset indicator recommendation model, and a second ambiguity is determined; if the user has not specified an analysis direction, a third ambiguity is determined; and an instruction analysis state is generated based on the existence of the first ambiguity, the second ambiguity, or the third ambiguity, and whether any parameters are missing. If the instruction analysis status indicates that there is intent ambiguity or missing parameters, an intent interaction card is generated and displayed. The intent interaction card refers to a front-end interface interaction component used to guide the user to confirm the candidate data indicators and / or supplement the missing data parameters. In response to the user's interactive operation, the candidate data indicators are confirmed or corrected to obtain the target data indicators, and / or the target data parameters corresponding to the target data indicators are supplemented. Data extraction is performed based on the target data indicators and the target data parameters to obtain raw analysis data, which is then displayed in an analysis display format. The analysis display format refers to the front-end interface format for visualizing and interactively analyzing the raw analysis data.

2. The method according to claim 1, characterized in that, The intent interaction card includes an indicator confirmation card, a parameter supplementation card, and a data extraction confirmation card. The indicator confirmation card is a front-end interface component used to display the candidate data indicators and receive the user's confirmation or correction operation for the candidate data indicators. The parameter supplementation card is a front-end interface component used to receive the user's operation to supplement missing data parameters. The data extraction confirmation card is a front-end interface component used to display the data extraction rules composed of the target data indicators and the target data parameters and receive the user's confirmation operation for the data extraction rules. The generation and display of intent interaction cards includes: When the instruction analysis status indicates that there is an ambiguity in the intent, the indicator confirmation card is generated and displayed; When the instruction analysis status indicates that a parameter is missing, a parameter supplement card is generated and displayed. After confirming the target data indicators and the target data parameters, the data extraction confirmation card is generated and displayed.

3. The method according to claim 2, characterized in that, The step of responding to the user's interaction by confirming or correcting the candidate data metrics to obtain the target data metrics, and / or supplementing the target data parameters corresponding to the target data metrics, includes: The system receives a confirmation operation from the user for the candidate data indicator displayed in the indicator confirmation card, and determines the candidate data indicator as the target data indicator; or, it receives a correction operation from the user for the candidate data indicator displayed in the indicator confirmation card, and determines the corrected candidate data indicator as the target data indicator. Receive the missing data parameter input by the user in the parameter supplement card, and determine the missing data parameter as the target data parameter; Receive confirmation from the user regarding the data extraction rules displayed in the data extraction confirmation card, and trigger the data extraction operation based on the confirmation.

4. The method according to claim 3, characterized in that, The presentation of the raw analysis data in an analytical display format includes: Determine the data characteristics of the original analysis data. The data characteristics include at least one of data type, data dimension, and data volume. The data type includes a type with object identifier and a type without object identifier. The type with object identifier is used to characterize that the original analysis data contains an analysis object identifier. The type without object identifier is used to characterize that the original analysis data is general statistical data. The data dimension includes the dimension of the number of indicators, the dimension of the number of target codes, and the dimension of the number of analysis periods. Based on the combination of data type, data dimension, and data volume of the original analysis data, the analysis display format is determined from multiple preset analysis display formats, including at least one of data table format, trend chart format, comparison chart format, and comprehensive dashboard format.

5. The method according to claim 4, characterized in that, After determining the analysis display format from multiple preset analysis display formats based on a combination of the data type, data dimension, and data volume of the original analysis data, the method further includes: Generate a data summary text corresponding to the original analysis data, the data summary text being used to describe the core features of the original analysis data in natural language; The data summary text is displayed in the aforementioned analysis presentation format.

6. The method according to any one of claims 1-5, characterized in that, After displaying the raw analysis data in an analytical presentation format, the method further includes: Based on the data characteristics of the original analysis data and the user's historical analysis records, a recommendation analysis model or recommendation data index associated with the current analysis topic is generated and displayed in a preset area of ​​the analysis display format. The historical analysis records include the user's historical question data and historical query result data. When a user triggers an operation on the recommendation analysis model or the recommendation data metrics, the current analysis context is associated with the new analysis intent corresponding to the trigger operation, and a new round of data extraction, analysis, and display is performed based on the data metrics and data parameters corresponding to the associated analysis intent.

7. A data analysis device based on an interactive smart card, characterized in that, include: The instruction acquisition unit is used to acquire data analysis instructions input by the user. The state analysis unit is used to understand the intent of the data analysis instruction and obtain the instruction analysis state. The instruction analysis state is used to indicate whether there is intent ambiguity or missing parameters in the data analysis instruction, as well as the candidate data indicators identified from the data analysis instruction. The intent ambiguity is categorized into first ambiguity, second ambiguity, and third ambiguity based on the type of ambiguity. The process of understanding the intent of the data analysis instruction to obtain the instruction analysis state includes: identifying indicators in the data analysis instruction and determining whether candidate data indicators are identified; if candidate data indicators are identified, and the candidate data indicators contain ambiguous or similar indicators, then a first ambiguity is determined; if no candidate data indicators are identified, then it is determined whether the user has specified an analysis direction; if the user has specified an analysis direction, then a candidate data indicator corresponding to the analysis direction is determined based on a preset indicator recommendation model, and a second ambiguity is determined; if the user has not specified an analysis direction, a third ambiguity is determined; and an instruction analysis state is generated based on the existence of the first ambiguity, the second ambiguity, or the third ambiguity, and whether any parameters are missing. The card generation unit is used to generate and display an intent interaction card if the instruction analysis status indication has intent ambiguity or missing parameters. The intent interaction card refers to a front-end interface interaction component used to guide the user to confirm the candidate data indicators and / or supplement missing data parameters. An interactive response unit is used to respond to the user's interactive operation, confirm or correct the candidate data indicators to obtain the target data indicators, and / or supplement the target data parameters corresponding to the target data indicators; The data display unit is used to perform data extraction operations based on the target data indicators and the target data parameters to obtain raw analysis data, and to display the raw analysis data in an analysis display format, wherein the analysis display format refers to the front-end interface format for visual display and interactive analysis of the raw analysis data.

8. An electronic device, characterized in that, It includes a processor, a memory, and one or more programs, said one or more programs being stored in the memory and configured to be executed by the processor, said programs including instructions for performing the steps in the method as claimed in any one of claims 1-6.

9. A computer-readable storage medium having a computer program / instructions stored thereon, characterized in that, When the computer program / instructions are executed by the processor, they implement the steps of the method according to any one of claims 1-6.

Images