Digital transformation diagnosis method and system based on artificial intelligence

Abstract

The invention relates to the technical field of artificial intelligence, and discloses a digital transformation diagnosis method and system based on artificial intelligence, and the method comprises the steps: automatically extracting the multi-dimensional attributes of a user, such as the industry, the scale, the control mode and the listed place through a preset information extraction model; the target index system most conforming to the user characteristics is accurately matched in a certain number of preset index systems, the associated questions are obtained from the question bank based on the target index system, configuration such as weight distribution and question order optimization is completed, the information extraction model processes multi-modal data at the same time, the user attribute recognition accuracy is improved, and the user experience is improved. And the accuracy of enterprise digital transformation diagnosis is improved.

Description

Technical Field

[0001] This application relates to the field of artificial intelligence technology, and in particular to a digital transformation diagnostic method and system based on artificial intelligence. Background Technology

[0002] With the rapid development of the digital economy, governments at all levels, industry organizations, and large corporations need to conduct large-scale, periodic assessments of the digital transformation level of their jurisdictions or affiliated enterprises in order to accurately formulate support policies, allocate resources, and track results. Common practices currently include:

[0003] Offline evaluations rely on consulting teams to conduct in-depth on-site assessments of companies, generating evaluation reports through interviews, document review, and manual scoring. This method is labor-intensive, time-consuming, and costly, making it difficult to cover a large number of companies in a short period. Single-point questionnaire tools: Some organizations have developed online questionnaire systems, but these only support fixed questions and simple scoring, lacking dynamic indicator systems, hierarchical weight models, and industry benchmarking functions, resulting in highly subjective evaluation results and poor comparability. Static Excel / Word templates: Standardized forms are distributed, and companies fill them out offline before summarizing. This results in inconsistent data formats, low response rates, and the inability to update industry benchmark data in real time, making it difficult to support dynamic decision-making.

[0004] Existing technologies often suffer from drawbacks such as a lack of hierarchical management of indicator systems, high coupling between question banks and questionnaires, rigid scoring rules, and a single source of benchmarking data. Therefore, improving the accuracy of enterprise digital transformation diagnosis has become an urgent technical problem to be solved. Summary of the Invention

[0005] This application provides an artificial intelligence-based digital transformation diagnostic method and system to improve the accuracy of enterprise digital transformation diagnostics.

[0006] Firstly, this application provides a digital transformation diagnostic method based on artificial intelligence, the method comprising:

[0007] The target user attribute information of the target user is extracted by a preset information extraction model. The target indicator system is determined from each preset indicator system based on the target user attribute information. The target questionnaire is determined from each preset questionnaire based on the target indicator system.

[0008] Based on the target questionnaire and the target user attribute information, a digital diagnostic result is generated;

[0009] A target digital transformation diagnostic strategy is generated based on the preset digital transformation diagnostic model and the digital diagnostic results.

[0010] Secondly, this application also provides an artificial intelligence-based digital transformation diagnostic system, the system comprising:

[0011] The main site management module is used to manage the indicator system, question bank, questionnaires, and evaluation activities.

[0012] The sub-site service module is used by target users to fill out questionnaires and view diagnostic reports;

[0013] The service module provides score calculation service and evaluation report generation service. The score calculation is based on the model of basic indicator score = Σ (question score * weight) to aggregate and calculate the total score.

[0014] The integration module connects to the user center console and data dictionary system, supporting single sign-on and basic data synchronization.

[0015] This application discloses an AI-based digital transformation diagnostic method and system. The AI-based digital transformation diagnostic method includes: extracting target user attribute information from a preset information extraction model; determining a target indicator system from various preset indicator systems using the target user attribute information; determining a target questionnaire from various preset questionnaires based on the target indicator system; generating a digital diagnostic result based on the target questionnaire and the target user attribute information; and generating a target digital transformation diagnostic strategy based on the preset digital transformation diagnostic model and the digital diagnostic result. Through this method, this application automatically extracts multi-dimensional attributes such as user industry, scale, management model, and listing location using a preset information extraction model. It accurately matches the target indicator system that best matches the user characteristics from a certain number of preset indicator systems. Based on the target indicator system, it obtains related questions from a question bank, completes configurations such as weight allocation and question order optimization, and the information extraction model simultaneously processes multimodal data, improving the accuracy of user attribute identification and thus enhancing the accuracy of enterprise digital transformation diagnostics. Attached Figure Description

[0016] To more clearly illustrate the technical solutions of the embodiments of this application, the drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1 This is a schematic flowchart of an artificial intelligence-based digital transformation diagnostic method provided in an embodiment of this application;

[0018] Figure 2 A schematic block diagram of an artificial intelligence-based digital transformation diagnostic system provided for embodiments of this application;

[0019] Figure 3A schematic block diagram of the structure of a computer device provided for an embodiment of this application. Detailed Implementation

[0020] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0021] The flowchart shown in the attached diagram is for illustrative purposes only and does not necessarily include all content and operations / steps, nor does it necessarily have to be performed in the order described. For example, some operations / steps can be broken down, combined, or partially merged, so the actual execution order may change depending on the actual situation.

[0022] It should be understood that the terminology used in this specification is for the purpose of describing particular embodiments only and is not intended to limit the scope of the application. As used in this specification and the appended claims, the singular forms “a,” “an,” and “the” are intended to include the plural forms unless the context clearly indicates otherwise.

[0023] It should also be understood that the term “and / or” as used in this application specification and the appended claims means any combination of one or more of the associated listed items and all possible combinations, and includes such combinations.

[0024] This application provides an AI-based digital transformation diagnostic method and system. The AI-based digital transformation diagnostic method can be applied to a server. It automatically extracts multi-dimensional attributes such as user industry, scale, management model, and listing location through a preset information extraction model. It accurately matches the target indicator system that best matches the user's characteristics from a certain number of preset indicator systems. Based on the target indicator system, it retrieves related questions from a question bank, completes configurations such as weight allocation and question order optimization. The information extraction model simultaneously processes multi-modal data, improving the accuracy of user attribute identification and thus enhancing the accuracy of enterprise digital transformation diagnostics.

[0025] The following detailed description of some embodiments of this application is provided in conjunction with the accompanying drawings. Unless otherwise specified, the following embodiments and features can be combined with each other.

[0026] Please see Figure 1 , Figure 1This is a schematic flowchart illustrating an AI-based digital transformation diagnostic method provided in an embodiment of this application. This AI-based digital transformation diagnostic method can be applied to servers to improve the accuracy of enterprise digital transformation diagnostics.

[0027] like Figure 1 As shown, the AI-based digital transformation diagnostic method specifically includes steps S10 to S30.

[0028] Step S10: Extract target user attribute information of target users through a preset information extraction model, determine target indicator system from each preset indicator system through the target user attribute information, and determine target questionnaire from each preset questionnaire according to the target indicator system;

[0029] Specifically, define user attribute fields (such as company type, industry, size, etc.). These fields are pre-defined through a data dictionary table to ensure standardized attribute values. For example, the industry attribute is maintained through a tree structure, supporting multi-level classification (e.g., manufacturing > automotive manufacturing).

[0030] When an evaluation activity is created, the system automatically extracts target user attributes using a preset model:

[0031] Input: Select the enterprise scope (such as subsite, enterprise type, industry, etc.) on the evaluation activity interface;

[0032] Processing: The model matches attribute values ​​in real time based on the data dictionary. For example, when the "Automotive Manufacturing" industry is selected, the model maps to the corresponding data dictionary ID.

[0033] Output: Generates a structured attribute set, such as {Industry:"Automotive Manufacturing", Size:"Large", Enterprise Type:"Production / Service"};

[0034] Key logic: Attribute extraction relies on the system's built-in validation rules to ensure that the attributes are valid (e.g., industry values ​​must exist in the dictionary).

[0035] Map user attributes (such as industry) to the "Application Scenario" field of the indicator system, for example:

[0036] If the user's industry is "automobile manufacturing", then the matching indicator system is "industrial big data" or "integration of informatization and industrialization".

[0037] Matching logic: Ensure accurate association through database queries.

[0038] Perform the filtering in the "Indicator System Management" interface:

[0039] Input the application scenario value, query the list of enabled indicator systems, filter out the system that matches the scenario (such as the system with the status "enabled"), and finally output the unique target indicator system ID (such as ZBTX1902250002).

[0040] Step S20: Generate digital diagnostic results based on the target questionnaire and the target user attribute information;

[0041] Specifically, the system receives the identifier for the target questionnaire and the target user attribute information (such as company industry, size, management model, etc.). The target questionnaire is defined in the questionnaire management module, including the indicator system, item weights, and disabling rules; user attributes are obtained from the enterprise management module and are used to customize diagnostic scenarios.

[0042] The system queries the questionnaire database, loads the questionnaire content (such as indicator system and item weights), verifies user attributes (such as industry code and company size), and confirms whether it matches the questionnaire's preset "company type" or "industry" filtering conditions. For example, manufacturing company attributes will trigger relevant industry questionnaires. If the attributes do not match, the system will display an error message and pause the process; otherwise, it will proceed to the data collection stage.

[0043] Based on the target questionnaire, the system guides users (or automatically) to fill in the questionnaire data. Data collection is completed through an online platform (such as a sub-site filling module) to ensure real-time performance and accuracy. Specifically, the system pushes questionnaires based on user attributes. For example, group enterprise users will receive a questionnaire containing questions on "value chain integration". Data collection includes question scores (single choice, multiple choice, numerical fill-in-the-blank, etc.) and additional information. Numerical questions are scored using a formula editor.

[0044] The score for each final-level indicator is calculated as ∑(question score * question weight). For example, the score for the production digitization indicator is based on questions such as the CNC rate of key processes.

[0045] Non-final indicator score calculation: Higher-level indicator score = ∑(lower-level indicator score * weight). For example, the score of the first-level indicator of business innovation and transformation is obtained by weighting its lower-level indicators (such as digital production and digital R&D design).

[0046] Total score calculation: Total score = ∑(score of primary indicator * weight), where the weight comes from the indicator system settings, such as development strategy accounting for 20%.

[0047] By querying sample databases, such as comparing data from 99% of companies in the metal products industry, users can identify indicators that are better or worse than their peers. A tiered model (the table in Document 1) is then applied to map the total score to tiers such as standard level and scenario level. For example, a total score of 65 points corresponds to tier 6, "Domain Level: Data-Driven Across the Entire Organization."

[0048] Step S30: Generate a target digital transformation diagnostic strategy based on the preset digital transformation diagnostic model and the digital diagnostic results.

[0049] Specifically, the system analyzes the basic information (such as industry, size, nature, region, etc.) submitted by enterprises through a preset information extraction model. Based on this, it matches the target indicator system from a general indicator system and a preset number of industry-differentiated templates, and automatically generates related questions to form a personalized questionnaire. Enterprises complete the questionnaire online, and the system calculates the scores of each indicator according to the established scoring rules. It also benchmarks the data against a multi-dimensional database of 130,000 enterprises in terms of industry, region, size, etc., in real time, and outputs structured diagnostic results and shortcomings. The system calls the built-in digital transformation diagnostic model and, based on the diagnostic results and enterprise attributes, intelligently recommends the most suitable advanced strategy from five levels and ten grades from the standard level to the ecosystem level. It generates a personalized diagnostic report that includes stage goals, improvement measures, and expected benefits, realizing full automation of the process from "attribute extraction → questionnaire matching → result generation → strategy output".

[0050] This embodiment discloses an AI-based digital transformation diagnostic method and system. The AI-based digital transformation diagnostic method includes extracting target user attribute information of target users through a preset information extraction model; determining a target indicator system from various preset indicator systems based on the target user attribute information; determining a target questionnaire from various preset questionnaires based on the target indicator system; generating digital diagnostic results based on the target questionnaire and the target user attribute information; and generating a target digital transformation diagnostic strategy based on the preset digital transformation diagnostic model and the digital diagnostic results. Through the above method, this application automatically extracts multi-dimensional attributes such as user industry, scale, management model, and listing location through a preset information extraction model, accurately matches the target indicator system that best matches the user characteristics from a certain number of preset indicator systems, obtains related questions from a question bank based on the target indicator system, and completes configurations such as weight allocation and question order optimization. The information extraction model simultaneously processes multimodal data, improving the accuracy of user attribute identification and thus improving the accuracy of enterprise digital transformation diagnosis.

[0051] based on Figure 1 In the illustrated embodiment, before step S10, the following steps are included:

[0052] Construct a hierarchical indicator library, wherein the hierarchical indicator library includes at least one hierarchical indicator and at least one basic indicator;

[0053] Specifically, a hierarchical indicator library will be created in the system, and all indicators will be divided into two categories, "hierarchical indicators" and "basic indicators", according to business needs.

[0054] Hierarchical indicators serve only as directory nodes and are not directly associated with questions. They are used to form a multi-level tree structure. Basic indicators are attached to hierarchical indicators and serve as the smallest unit for scoring and linking questions.

[0055] The system automatically generates unique numbers according to rules (such as ZB + date + serial number), supports category retrieval, enable / freeze status management, and allows adjustment of hierarchical relationships at any time.

[0056] Based on the preset weight configuration rules and the hierarchical indicator library, at least one preset indicator system is generated.

[0057] Specifically, in the "Indicator System Management" function, create a new system:

[0058] Fill in the system name, application scenario, and description. Drag and drop the required level indicators from the indicator library to the system tree, and then attach the corresponding basic indicators under it. According to the principle of "the sum of weights of the same level = 100", fill in the weight for each indicator. If you need to judge the maturity level by level, then configure the stage score and downgrade rules for the first-level indicators.

[0059] Based on the preset question bank and the hierarchical indicator bank, a preset questionnaire corresponding to each preset indicator system is generated.

[0060] Specifically, the system is managed in three levels: "Category - Question Bank - Question". Questions support four types: single choice, multiple choice, numerical fill-in-the-blank, and open-ended questions. Each question can be associated with zero or more basic indicators. A new questionnaire can be created in "Questionnaire Management".

[0061] First, select one or more established indicator systems. The system will automatically bring up the questions associated with these basic indicators. You can manually add or delete questions, adjust their order, and assign a weight to each question (the sum of weights under the same basic indicator = 100). Set parameters such as scoring / disabling relationships, questionnaire introduction, and whether to provide feedback on scores.

[0062] In a specific embodiment, a hierarchical indicator library is constructed, including:

[0063] Non-final level indicators that contain at least one sub-indicator are defined as the hierarchical indicators, and final level indicators that do not contain the sub-indicators are defined as the basic indicators.

[0064] Configure indicator codes for each of the aforementioned hierarchical indicators and each of the aforementioned basic indicators, and create the hierarchical indicator library based on each of the aforementioned hierarchical indicators, each of the aforementioned basic indicators, and each of the aforementioned indicator codes.

[0065] Specifically, on the indicator library management page, click "Add Indicator". First, enter all the bottom-level indicators (such as "CNC rate of key processes" and "automatic data collection rate") according to the business logic. The system will automatically mark them as basic indicators. Then, enter the upper-level group indicators (such as "digitalization of production" and "data governance"). Check the "hierarchical indicators" checkbox to make them directory nodes that can be attached to sub-indicators.

[0066] Click "Add Sub-indicator" on the hierarchical indicator node, select the corresponding basic indicator or other hierarchical indicators from the list to form a multi-level tree structure; the system verifies in real time: each hierarchical indicator must contain at least one sub-indicator, and no sub-indicators can be attached to a basic indicator.

[0067] The system automatically generates a unique code according to the following rules:

[0068] The hierarchical / basic indicators are uniformly prefixed with "ZB" + 8-digit date + 4-digit serial number;

[0069] If human-readable functionality is required, a business abbreviation can be entered in the alias field. Once generated, the code cannot be modified to ensure consistency in subsequent references. The system adds a name, description, and category to each indicator and saves it. All hierarchical indicators, basic indicators, and their codes are then synchronously written to the indicator library. It supports enabling / freezing, category retrieval, and tree-view display, completing the creation of a hierarchical indicator library.

[0070] In a specific embodiment, based on preset weight configuration rules and the hierarchical indicator library, at least one preset indicator system is generated, including:

[0071] Based on the preset weight configuration rules, assign indicator weights to each hierarchical indicator and each basic indicator in the hierarchical indicator library;

[0072] Based on the hierarchical indicator library and the weights of each indicator, at least one of the preset indicator systems is generated.

[0073] Specifically, open "Indicator System Management" in the main menu, click "Create Indicator System," fill in the system name, application scenario, and description, and save it as a draft. In the new system tree interface, click the root node "Add Indicator," select the level indicator to be used as a first-level directory from the indicator library (such as "Development Strategy"), and enter the weight of the indicator at this level (e.g., 25%). Continue to add other first-level indicators in the same way, ensuring that the sum of the first-level weights is 100.

[0074] Select a primary-level indicator, click "Add Sub-indicator," select its underlying basic indicator from the indicator library (such as "Strategic Planning Deployment"), and enter the weight of that basic indicator under its parent level (e.g., 40%). Repeat this process until all basic indicators are linked, ensuring that the sum of the weights of all sub-indicators under the same parent level is 100.

[0075] After completing the weight configuration, click "Save" → "Enable", and the system will solidify the hierarchical indicator library and its weights into a preset indicator system.

[0076] In a specific embodiment, based on the preset question bank and the hierarchical indicator library, a preset questionnaire corresponding to each of the preset indicator systems is generated, including:

[0077] Convert each preset question in the preset question bank into a question code;

[0078] The indicator code is matched with the question code to generate a code matching result;

[0079] The preset questionnaire is generated based on the coding matching results.

[0080] Specifically, on the question bank management page, a unique code is automatically generated for each preset question. For example, the first multiple-choice question is "TM100000001". Enter the "Question Editing" interface, select the basic or hierarchical indicator to be bound in the "Associated Indicators" column, and the system will automatically write the code of the indicator into the question's metadata, forming a one-to-one or one-to-many matching record of "indicator code - question code".

[0081] When creating a new questionnaire, select the enabled indicator system. The system will automatically pull all related questions into the questionnaire based on the matching records. Administrators can add or delete questions, adjust the order, set question weights and disable rules as needed, and then save the results to obtain a preset questionnaire corresponding to the indicator system.

[0082] based on Figure 1 In the illustrated embodiment, step S10 includes:

[0083] Obtain the multimodal information of the target user, and extract structured data from the multimodal information using the preset information extraction model;

[0084] The structured data is identified as the target user attribute information, and the target user attribute information is converted into target user attribute encoding;

[0085] Obtain the preset indicator system code corresponding to each preset indicator system, and determine the preset indicator system corresponding to the preset indicator system code with the highest matching degree with the target user attribute code as the target indicator system;

[0086] The preset questionnaire corresponding to the target indicator system is determined as the target questionnaire.

[0087] Specifically, the system automatically collects the materials submitted by target users at each entry point: text in online forms, scanned copies (images) of uploaded business licenses, supplementary explanations (audio) for voice-filled forms, and historical filling records (structured data).

[0088] The above multimodal information is then fed into a preset information extraction model:

[0089] Using OCR to identify company names, industries, and sizes in images;

[0090] Convert speech to text using ASR;

[0091] Key fields are extracted using BERT+ rules to output unified structured data.

[0092] Each field is mapped to a fixed-length code segment according to a predetermined encoding table and then concatenated to form a unique target user attribute code.

[0093] Obtain all enabled preset indicator system codes (including industry, scale, nature and other dimension identifiers), calculate the cosine similarity or exact match with the target user attribute codes, and select the system with the highest match as the target indicator system.

[0094] In a specific embodiment, extracting structured data from the multimodal information using the preset information extraction model includes:

[0095] The text semantic vectors in the text information are extracted using the pre-trained language sub-model built into the preset information extraction model;

[0096] Specifically, the text information, such as the text description and speech-to-text results filled in by the enterprise online, is input into the pre-trained language sub-model (such as RoBERTa) built into the preset information extraction model to obtain a fixed-dimensional text semantic vector V_txt.

[0097] The image semantic vector of the image information is extracted by the convolutional neural network built into the preset information extraction model;

[0098] Specifically, the uploaded business license, on-site photos, report screenshots, and other image information are fed into a convolutional neural network (such as ResNet-50), and the image semantic vector V_img is output after a global average pooling layer.

[0099] The temporal features in the video information are extracted using the time series network built into the preset information extraction model.

[0100] Specifically, for video information containing operation recordings or monitoring segments, after extracting pixel sequences frame by frame, temporal features are extracted through a time series network (such as 3D-CNN or Transformer encoder) to obtain a temporal semantic vector V_seq.

[0101] The text semantic vector, the image semantic vector, and the temporal features are mapped to the same dimension through the cross-modal alignment network of the preset information extraction model to generate the structured data.

[0102] Specifically, V_txt, V_img, and V_seq are simultaneously input into a cross-modal alignment network (shared projection layer + contrastive learning or attention mechanism), which maps the three vectors to the same dimensional space and merges them into a unified structured data output for subsequent attribute encoding and indicator system matching.

[0103] In a specific embodiment, step S20 includes:

[0104] The original scores of each target question in the target questionnaire are obtained in the pre-selection, and each of the pre-selected questions is converted into a pre-selected question vector;

[0105] The target user attribute information is mapped to a target user attribute vector, and the target user attribute vector is concatenated to the blank fields of each preset question vector to generate a fusion vector;

[0106] The target score for each target problem is determined based on the original scores, the fusion vectors, and the weights of the hierarchical and basic indicators in the target indicator system.

[0107] The digital diagnostic results are generated based on the target scores and the target indicator system.

[0108] Specifically, the preset raw score (0–100 points range) for each target question is read out sequentially from the target questionnaire. At the same time, the same vocabulary / encoder is used to convert the question stem, options, and explanatory text into a preset question vector q of fixed length.

[0109] The obtained target user attribute information (industry, size, region, etc.) is first mapped to a unified target user attribute vector u, and then this vector is directly concatenated to the reserved blank field of the vector q of each target question to form a "question-attribute" fusion vector f=[q|u].

[0110] The fusion vector f is processed by a multilayer perceptron (MLP) to output an adjustment coefficient α (between 0.8 and 1.2). This coefficient is multiplied by the original score to obtain a temporary score. Then, it is multiplied by the weight w_base of the basic indicator in the target indicator system and the weight w_level of the indicator at the level to which the basic indicator belongs, to obtain the final target score s = original score * α * w_base * w_level.

[0111] Weighted summation from bottom to top according to hierarchy;

[0112] Basic indicator score = Σ(target problem s);

[0113] Hierarchical indicator score = Σ(sub-indicator score * weight of that level);

[0114] Total system score = Σ(score of first-level indicators).

[0115] In a specific embodiment, step S30 includes:

[0116] The preset digital transformation diagnostic model is invoked, wherein the preset digital transformation diagnostic model defines the mapping relationship between each preset indicator system and each preset digital transformation diagnostic strategy;

[0117] Based on the target indicator system, at least one digital transformation diagnostic strategy to be determined is selected from each of the preset digital transformation diagnostic strategies.

[0118] Based on each of the digital transformation diagnostic strategies to be determined and the digital diagnostic results, determine the optimization information to be determined corresponding to each of the digital transformation diagnostic strategies to be determined;

[0119] Target optimization information is determined from the various optimization information to be determined, and the digital transformation diagnostic strategy to be determined corresponding to the target optimization information is determined as the target digital transformation diagnostic strategy.

[0120] Specifically, the "digital transformation diagnostic model" pre-trained and embedded in the platform is loaded into memory. The model contains a complete mapping table of "preset indicator system code → preset diagnostic strategy", as well as metadata such as thresholds, weights, and template paths required by the strategy.

[0121] Using the obtained target indicator system code as the key, all the preset diagnostic strategies bound to it are retrieved in the mapping table at once to form a candidate list (usually 2-4 items, such as "domain-level → platform-level transition strategy" "scenario-level → domain-level reinforcement strategy", etc.), which is recorded as the set of strategies to be determined.

[0122] For each strategy to be determined, read the "key indicator threshold" and "improvement path template" defined in the strategy template, substitute the generated digital diagnostic results (total score, scores of each primary indicator, and ranking of weaknesses) into the template formula, and calculate the "optimization information to be determined" under the strategy.

[0123] The system sorts all optimization information to be determined according to a preset priority rule (compares the least number of shortcomings, the highest ROI, and the lowest risk in that order), takes the optimization information ranked first as the "target optimization information", and marks the corresponding strategy to be determined as the "target digital transformation diagnostic strategy".

[0124] Please see Figure 2 , Figure 2 This application provides a schematic block diagram of an AI-based digital transformation diagnostic system, which is used to perform the aforementioned AI-based digital transformation diagnostic method.

[0125] like Figure 2 As shown, the AI-based digital transformation diagnostic system 400 includes:

[0126] The main site management module is used to manage the indicator system, question bank, questionnaires, and evaluation activities.

[0127] The sub-site service module is used by target users to fill out questionnaires and view diagnostic reports;

[0128] The service module provides score calculation service and evaluation report generation service. The score calculation is based on the model of basic indicator score = Σ (question score * weight) to aggregate and calculate the total score.

[0129] The integration module connects to the user center console and data dictionary system, supporting single sign-on and basic data synchronization.

[0130] It should be noted that those skilled in the art will understand that, for the sake of convenience and brevity, the specific working processes of the system and each module described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0131] The above-described system can be implemented as a computer program, which can be used in, for example... Figure 3 It runs on the computer device shown.

[0132] Please see Figure 3 , Figure 3 This is a schematic block diagram illustrating the structure of a computer device according to an embodiment of this application. The computer device may be a server.

[0133] See Figure 3 The computer device includes a processor, memory, and network interface connected via a system bus, wherein the memory may include non-volatile storage media and internal memory.

[0134] Non-volatile storage media can store operating systems and computer programs. These computer programs include program instructions that, when executed, cause the processor to perform any AI-based digital transformation diagnostic method.

[0135] The processor provides computing and control capabilities, supporting the operation of the entire computer device.

[0136] Internal memory provides an environment for the execution of computer programs in non-volatile storage media. When executed by a processor, the computer program can enable the processor to perform any kind of AI-based digital transformation diagnostic method.

[0137] This network interface is used for network communication, such as sending assigned tasks. Those skilled in the art will understand that... Figure 3 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0138] It should be understood that the processor can be a Central Processing Unit (CPU), but it can also be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. Among these, a general-purpose processor can be a microprocessor or any conventional processor.

[0139] In one embodiment, the processor is configured to run a computer program stored in memory to perform the following steps:

[0140] The target user attribute information of the target user is extracted by a preset information extraction model. The target indicator system is determined from each preset indicator system based on the target user attribute information. The target questionnaire is determined from each preset questionnaire based on the target indicator system.

[0141] Based on the target questionnaire and the target user attribute information, a digital diagnostic result is generated;

[0142] A target digital transformation diagnostic strategy is generated based on the preset digital transformation diagnostic model and the digital diagnostic results.

[0143] In one embodiment, before extracting target user attribute information from a preset information extraction model, determining a target indicator system from various preset indicator systems based on the target user attribute information, and determining the target questionnaire from various preset questionnaires based on the target indicator system, the following steps are performed:

[0144] Construct a hierarchical indicator library, wherein the hierarchical indicator library includes at least one hierarchical indicator and at least one basic indicator;

[0145] Based on the preset weight configuration rules and the hierarchical indicator library, at least one preset indicator system is generated.

[0146] Based on the preset question bank and the hierarchical indicator bank, a preset questionnaire corresponding to each preset indicator system is generated.

[0147] In one embodiment, a hierarchical indicator library is constructed to achieve the following:

[0148] Non-final level indicators that contain at least one sub-indicator are defined as the hierarchical indicators, and final level indicators that do not contain the sub-indicators are defined as the basic indicators.

[0149] Configure indicator codes for each of the aforementioned hierarchical indicators and each of the aforementioned basic indicators, and create the hierarchical indicator library based on each of the aforementioned hierarchical indicators, each of the aforementioned basic indicators, and each of the aforementioned indicator codes.

[0150] In one embodiment, at least one preset indicator system is generated based on preset weight configuration rules and the hierarchical indicator library to achieve:

[0151] Based on the preset weight configuration rules, assign indicator weights to each hierarchical indicator and each basic indicator in the hierarchical indicator library;

[0152] Based on the hierarchical indicator library and the weights of each indicator, at least one of the preset indicator systems is generated.

[0153] In one embodiment, a preset questionnaire corresponding to each of the preset indicator systems is generated based on a preset question bank and the hierarchical indicator bank, for the purpose of:

[0154] Convert each preset question in the preset question bank into a question code;

[0155] The indicator code is matched with the question code to generate a code matching result;

[0156] The preset questionnaire is generated based on the coding matching results.

[0157] In one embodiment, target user attribute information of the target user is extracted using a preset information extraction model; a target indicator system is determined from various preset indicator systems based on the target user attribute information; and a target questionnaire is determined from various preset questionnaires based on the target indicator system. This is used to achieve the following:

[0158] Obtain the multimodal information of the target user, and extract structured data from the multimodal information using the preset information extraction model;

[0159] The structured data is identified as the target user attribute information, and the target user attribute information is converted into target user attribute encoding;

[0160] Obtain the preset indicator system code corresponding to each preset indicator system, and determine the preset indicator system corresponding to the preset indicator system code with the highest matching degree with the target user attribute code as the target indicator system;

[0161] The preset questionnaire corresponding to the target indicator system is determined as the target questionnaire.

[0162] In one embodiment, structured data is extracted from the multimodal information using the preset information extraction model to achieve the following:

[0163] The text semantic vectors in the text information are extracted using the pre-trained language sub-model built into the preset information extraction model;

[0164] The image semantic vector of the image information is extracted by the convolutional neural network built into the preset information extraction model;

[0165] The temporal features in the video information are extracted using the time series network built into the preset information extraction model.

[0166] The text semantic vector, the image semantic vector, and the temporal features are mapped to the same dimension through the cross-modal alignment network of the preset information extraction model to generate the structured data.

[0167] In one embodiment, a digital diagnostic result is generated based on the target questionnaire and the target user attribute information to achieve:

[0168] The original scores of each target question in the target questionnaire are obtained in the pre-selection, and each of the pre-selected questions is converted into a pre-selected question vector;

[0169] The target user attribute information is mapped to a target user attribute vector, and the target user attribute vector is concatenated to the blank fields of each preset question vector to generate a fusion vector;

[0170] The target score for each target problem is determined based on the original scores, the fusion vectors, and the weights of the hierarchical and basic indicators in the target indicator system.

[0171] The digital diagnostic results are generated based on the target scores and the target indicator system.

[0172] In one embodiment, a target digital transformation diagnostic strategy is generated based on a preset digital transformation diagnostic model and the digital diagnostic results, for the purpose of:

[0173] The preset digital transformation diagnostic model is invoked, wherein the preset digital transformation diagnostic model defines the mapping relationship between each preset indicator system and each preset digital transformation diagnostic strategy;

[0174] Based on the target indicator system, at least one digital transformation diagnostic strategy to be determined is selected from each of the preset digital transformation diagnostic strategies.

[0175] Based on each of the digital transformation diagnostic strategies to be determined and the digital diagnostic results, determine the optimization information to be determined corresponding to each of the digital transformation diagnostic strategies to be determined;

[0176] Target optimization information is determined from the various optimization information to be determined, and the digital transformation diagnostic strategy to be determined corresponding to the target optimization information is determined as the target digital transformation diagnostic strategy.

[0177] The embodiments of this application also provide a computer-readable storage medium storing a computer program, the computer program including program instructions, and the processor executing the program instructions to implement any of the artificial intelligence-based digital transformation diagnostic methods provided in the embodiments of this application.

[0178] The computer-readable storage medium may be an internal storage unit of the computer device described in the foregoing embodiments, such as the hard disk or memory of the computer device. The computer-readable storage medium may also be an external storage device of the computer device, such as a plug-in hard disk, SmartMedia Card (SMC), Secure Digital (SD) card, or Flash Card equipped on the computer device.

[0179] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in this application, and these modifications or substitutions should all be covered within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A digital transformation diagnostic method based on artificial intelligence, characterized in that, include: The target user attribute information of the target user is extracted by a preset information extraction model. The target indicator system is determined from each preset indicator system based on the target user attribute information. The target questionnaire is determined from each preset questionnaire based on the target indicator system. Based on the target questionnaire and the target user attribute information, generate digital diagnostic results; A target digital transformation diagnostic strategy is generated based on the preset digital transformation diagnostic model and the digital diagnostic results.

2. The AI-based digital transformation diagnostic method according to claim 1, characterized in that, Before extracting target user attribute information from a preset information extraction model, determining a target indicator system from various preset indicator systems based on the target user attribute information, and determining the target questionnaire from various preset questionnaires based on the target indicator system, the process includes: Construct a hierarchical indicator library, wherein the hierarchical indicator library includes at least one hierarchical indicator and at least one basic indicator; Based on the preset weight configuration rules and the hierarchical indicator library, at least one preset indicator system is generated. Based on the preset question bank and the hierarchical indicator bank, a preset questionnaire corresponding to each preset indicator system is generated.

3. The AI-based digital transformation diagnostic method according to claim 2, characterized in that, The construction of the hierarchical indicator library includes: Non-final level indicators that contain at least one sub-indicator are defined as the hierarchical indicators, and final level indicators that do not contain the sub-indicators are defined as the basic indicators. Configure indicator codes for each of the aforementioned hierarchical indicators and each of the aforementioned basic indicators, and create the hierarchical indicator library based on each of the aforementioned hierarchical indicators, each of the aforementioned basic indicators, and each of the aforementioned indicator codes.

4. The AI-based digital transformation diagnostic method according to claim 3, characterized in that, The generation of at least one preset indicator system based on preset weight configuration rules and the hierarchical indicator library includes: Based on the preset weight configuration rules, assign indicator weights to each hierarchical indicator and each basic indicator in the hierarchical indicator library; Based on the hierarchical indicator library and the weights of each indicator, at least one of the preset indicator systems is generated.

5. The AI-based digital transformation diagnostic method according to claim 4, characterized in that, The step of generating a preset questionnaire corresponding to each of the preset indicator systems based on the preset question bank and the hierarchical indicator bank includes: Convert each preset question in the preset question bank into a question code; The indicator code is matched with the question code to generate a code matching result; The preset questionnaire is generated based on the coding matching results.

6. The AI-based digital transformation diagnostic method according to claim 1, characterized in that, The step of extracting target user attribute information from a preset information extraction model, determining a target indicator system from various preset indicator systems based on the target user attribute information, and determining a target questionnaire from various preset questionnaires based on the target indicator system includes: Obtain the multimodal information of the target user, and extract structured data from the multimodal information using the preset information extraction model; The structured data is identified as the target user attribute information, and the target user attribute information is converted into target user attribute encoding; Obtain the preset indicator system code corresponding to each preset indicator system, and determine the preset indicator system corresponding to the preset indicator system code with the highest matching degree with the target user attribute code as the target indicator system; The preset questionnaire corresponding to the target indicator system is determined as the target questionnaire.

7. The AI-based digital transformation diagnostic method according to claim 6, characterized in that, The multimodal information includes text information, image information, and video information. The step of extracting structured data from the multimodal information using the preset information extraction model includes: The text semantic vectors in the text information are extracted using the pre-trained language sub-model built into the preset information extraction model; The image semantic vector of the image information is extracted by the convolutional neural network built into the preset information extraction model; The temporal features in the video information are extracted using the time-series network built into the preset information extraction model. The text semantic vector, the image semantic vector, and the temporal features are mapped to the same dimension through the cross-modal alignment network of the preset information extraction model to generate the structured data.

8. The AI-based digital transformation diagnostic method according to claim 2, characterized in that, The step of generating a digital diagnostic result based on the target questionnaire and the target user attribute information includes: The original scores of each target question in the target questionnaire are obtained in the pre-selection, and each of the pre-selected questions is converted into a pre-selected question vector; The target user attribute information is mapped to a target user attribute vector, and the target user attribute vector is concatenated to the blank fields of each preset question vector to generate a fusion vector; The target score for each target problem is determined based on the original scores, the fusion vectors, and the weights of the hierarchical and basic indicators in the target indicator system. The digital diagnostic results are generated based on the target scores and the target indicator system.

9. The AI-based digital transformation diagnostic method according to claim 1, characterized in that, The step of generating a target digital transformation diagnostic strategy based on a preset digital transformation diagnostic model and the digital diagnostic results includes: The preset digital transformation diagnostic model is invoked, wherein the preset digital transformation diagnostic model defines the mapping relationship between each preset indicator system and each preset digital transformation diagnostic strategy; Based on the target indicator system, at least one digital transformation diagnostic strategy to be determined is selected from each of the preset digital transformation diagnostic strategies. Based on each of the digital transformation diagnostic strategies to be determined and the digital diagnostic results, determine the optimization information to be determined corresponding to each of the digital transformation diagnostic strategies to be determined; Target optimization information is determined from the various optimization information to be determined, and the digital transformation diagnostic strategy to be determined corresponding to the target optimization information is determined as the target digital transformation diagnostic strategy.

10. A digital transformation diagnostic system based on artificial intelligence, characterized in that, include: The main site management module is used to manage the indicator system, question bank, questionnaires, and evaluation activities. The sub-site service module is used by target users to fill out questionnaires and view diagnostic reports; The service module provides score calculation service and evaluation report generation service. The score calculation is based on the model of basic indicator score = Σ (question score * weight) to aggregate and calculate the total score. The integration module connects to the user center console and data dictionary system, supporting single sign-on and basic data synchronization.

Images