Intelligent data processing method and device with AI digital person

Abstract

The embodiment of the application provides a kind of smart data processing method, device and electronic equipment with AI digital person, belong to data processing technical field, which comprises: obtaining the demand information Q1 of user;Based on the demand information Q1 of user, generate multidimensional data visualization chart T;Utilize deep learning model to analyze the interactive history Jh of user;Obtain the feedback information Q2 of user to data visualization chart T, historical analysis data D1 and future prediction data D2, and determine the social media account information M of user based on the feedback information Q2;The authenticity of the social media account information M of user is verified, and different types of AI digital person expression are generated based on the result of authenticity verification.The application can improve the interactivity and intelligent level of data board.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, and in particular to an intelligent data processing method, apparatus, and electronic device equipped with an AI digital human. Background Technology

[0002] Intelligent data dashboards are advanced data visualization tools that help enterprise users understand and utilize data more efficiently by integrating multiple data sources, applying advanced analytics, and providing rich interactive features. Intelligent data dashboards not only display data but also provide in-depth data insights in an intelligent way to support decision-making.

[0003] Existing technologies have the following problems in the development of interactivity and intelligence in intelligent data dashboard systems:

[0004] 1. Passivity of data presentation: Traditional data dashboards mainly present data in the form of charts and graphs, requiring users to actively browse and interpret the data, lacking intelligent guidance and interaction.

[0005] 2. Difficulty in interpreting complex data: For non-technical personnel, complex multidimensional data and cross-analysis results are difficult to understand, and traditional data dashboards cannot provide intuitive data interpretation.

[0006] 3. Insufficient real-time interaction: While existing data dashboards have basic real-time data update capabilities, they lack real-time interactive features based on Natural Language Processing (NLP) and artificial intelligence technologies, preventing users from quickly obtaining analysis conclusions through natural language.

[0007] 4. Limited personalized data analysis capabilities: Traditional systems typically lack the ability to automatically generate personalized data analysis and recommendations based on user needs and preferences.

[0008] The aforementioned issues have become technical problems that need to be solved. Summary of the Invention

[0009] In view of this, embodiments of the present invention provide an intelligent data processing method, apparatus, and electronic device equipped with an AI digital human, which at least partially solves the problems existing in the prior art.

[0010] In a first aspect, embodiments of the present invention provide an intelligent data processing method equipped with an AI digital human, comprising:

[0011] An AI digital human is embedded in a digital dashboard. The AI ​​digital human interacts with the user through voice, text, or gestures to obtain the user's needs information Q1.

[0012] Based on the user's needs information Q1, generate a multi-dimensional data visualization chart T, and based on the chart, select an adaptive algorithm to dynamically adjust the chart layout, style and interaction method in real time by analyzing data characteristics.

[0013] The user's interaction history Jh is analyzed using a deep learning model, and historical analysis data D1 and future prediction data D2 are generated for the user based on the interaction history.

[0014] Obtain user feedback information Q2 regarding data visualization chart T, historical analysis data D1, and future prediction data D2, and determine the user's social media account information M based on the feedback information Q2;

[0015] The authenticity of the user's social media account information M is verified, and different types of AI digital human expressions are generated based on the verification results.

[0016] According to a specific implementation of an embodiment of the present invention, embedding an AI digital human in a digital dashboard, wherein the AI ​​digital human interacts with the user through voice, text, or gestures to obtain the user's demand information Q1, includes:

[0017] Use the speech recognition API to convert the user's voice input into text in Qt;

[0018] Use natural language processing techniques to parse the user's text input Qr;

[0019] Capture user body movements and map user gestures to specific interaction commands Qm;

[0020] Based on text-based Qt, text input Qr, and interactive commands Qm, generate user requirement information Q1.

[0021] According to a specific implementation of an embodiment of the present invention, the step of generating a multi-dimensional data visualization chart T based on user demand information Q1, and dynamically adjusting the chart layout, style, and interaction method in real time based on a chart selection adaptive algorithm, includes:

[0022] Parse the user's requirement information Q1 to determine the data source, time range, data type, and chart type corresponding to the requirement information Q1;

[0023] Data is retrieved from a specified data source and preprocessed to obtain chart data;

[0024] Use a data visualization library to transform the chart data into a multi-dimensional data visualization chart T.

[0025] According to a specific implementation of an embodiment of the present invention, the step of generating a multi-dimensional data visualization chart T based on user demand information Q1, and dynamically adjusting the chart layout, style, and interaction method in real time based on a chart selection adaptive algorithm, further includes:

[0026] Real-time analysis of data characteristics, including data volume, data distribution characteristics, data change trends, and identification of key data features;

[0027] Automatically select the appropriate chart type based on data characteristics;

[0028] The layout of the chart is dynamically adjusted based on the amount of data and the chart type.

[0029] According to a specific implementation of an embodiment of the present invention, the step of analyzing a user's interaction history Jh using a deep learning model and generating historical analysis data D1 and future prediction data D2 for the user based on the interaction history includes:

[0030] The state update formula for building a deep learning model:

[0031]

[0032] For the output of the forget gate, The Sigmoid activation function, with an output range between 0 and 1, controls the degree to which information is preserved. Here is the weight matrix for the forget gate. This is the hidden state from the previous moment. For the current input, This is the bias vector for the forget gate;

[0033]

[0034] The output of the input gate, Here is the weight matrix of the input gate. The bias vector for the input gate;

[0035]

[0036] The candidate cell state is new information calculated using the current input and the hidden state from the previous time step. It is a hyperbolic tangent activation function, with an output range between -1 and 1. This is the weight matrix for the candidate cell states. is the bias vector for the candidate cell state;

[0037]

[0038] This refers to the current state of memory cells. For element-wise multiplication, This represents the state of memory cells at the previous moment;

[0039]

[0040] For the output of the output gate, Here is the weight matrix of the output gate. This is the bias vector for the output gate;

[0041]

[0042] This is the hidden state at the current moment, used to pass it to the next time step. Memory cell state The value after applying the hyperbolic tangent activation function;

[0043] The time series is decomposed into trend, seasonal, and random components using an additive model:

[0044]

[0045] The original time series data, The trend component represents the long-term trend of the data. The seasonal component indicates the periodic fluctuations in the data. The random component represents random noise or unpredictable parts of the data.

[0046] According to a specific implementation of an embodiment of the present invention, obtaining user feedback information Q2 regarding data visualization chart T, historical analysis data D1, and future prediction data D2, and determining the user's social media account information M based on the feedback information Q2, includes:

[0047] The logistic regression model was used to analyze the sentiment polarity of user feedback, and the probability P(y=1|x) of positive sentiment in user feedback was calculated:

[0048]

[0049] It is the Sigmoid activation function, w is the weight vector, x is the feature vector, and b is the bias term;

[0050] Based on the sentiment analysis results of user feedback information Q2, and combined with the user's behavioral data in the system, the user's social media account information M is determined.

[0051] According to a specific implementation of an embodiment of the present invention, the step of verifying the authenticity of the user's social media account information M, and generating different types of AI digital human expressions based on the verification results, includes:

[0052] Calculate the trust score S of a user's social media account information M:

[0053]

[0054] S is the trust score, α, β, γ, δ, ϵ are the feature weights, A is the user account age, C is the number of user account connections, P is the number of posts published by the user account, F is the number of users followed by the user account, and L is the number of likes on the user account.

[0055] Based on trust scoring, the authenticity of a user's social media account information M is determined.

[0056] According to a specific implementation of an embodiment of the present invention, the step of verifying the authenticity of the user's social media account information M, and generating different types of AI digital human expressions based on the verification results, further includes:

[0057] Predicting facial expression selection probability using a multilayer perceptual model:

[0058]

[0059] x is the input feature vector (e.g., user interaction features); W1 is the first-layer weight matrix, b1 is the first-layer bias vector; σ is the activation function (e.g., ReLU function, σ(z)=max(0,z)); W2 is the second-layer weight matrix, used to map the hidden layer output to the expression category space, b2 is the second-layer bias vector; y is the output vector, where each element yi represents the probability of selecting the i-th expression. This represents the probability of choosing the i-th expression. The Softmax function is used to transform an output vector into a probability distribution. The definition of the Softmax function is:

[0060]

[0061] Where z is the input vector and k is the number of output categories;

[0062] This is the weight matrix for the second layer, used to transform the output of the hidden layer into the final output layer. For activation function:

[0063]

[0064] This is the weight matrix for the first layer, used to transform the input feature vector x into the input for the hidden layer. x is the input feature vector, containing features extracted from user feedback and social media account information. This is the bias vector for the first layer, used to adjust the input to the hidden layers. This is the bias vector for the second layer, used to adjust the input to the output layer.

[0065] Secondly, embodiments of the present invention provide an intelligent data processing device equipped with an AI digital human, comprising:

[0066] The acquisition module embeds an AI digital human into the digital dashboard. The AI ​​digital human interacts with the user through voice, text, or gestures to acquire the user's demand information Q1.

[0067] The generation module generates multi-dimensional data visualization charts T based on the user's needs information Q1, and dynamically adjusts the chart layout, style and interaction method based on the chart selection adaptive algorithm and real-time analysis of data characteristics.

[0068] The analysis module uses a deep learning model to analyze the user's interaction history Jh, and generates historical analysis data D1 and future prediction data D2 for the user based on the interaction history.

[0069] The module determines the user's social media account information M based on the feedback information Q2 from the user regarding the data visualization chart T, historical analysis data D1, and future prediction data D2.

[0070] The verification module verifies the authenticity of the user's social media account information M, and generates different types of AI digital human expressions based on the verification results.

[0071] Thirdly, embodiments of the present invention also provide an electronic device, the electronic device comprising:

[0072] At least one processor; and,

[0073] The memory is communicatively connected to the at least one processor; wherein,

[0074] The memory stores instructions that can be executed by the at least one processor to enable the at least one processor to perform the intelligent data processing method for AI-equipped digital humans in any of the first aspects or any implementations thereof.

[0075] Fourthly, embodiments of the present invention also provide a non-transitory computer-readable storage medium storing computer instructions for causing the computer to execute the intelligent data processing method with AI digital human as described in the first aspect or any implementation thereof.

[0076] Fifthly, embodiments of the present invention also provide a computer program product, the computer program product including a computing program stored on a non-transitory computer-readable storage medium, the computer program including program instructions, which, when executed by a computer, cause the computer to execute the intelligent data processing method with AI digital human in the first aspect or any implementation thereof.

[0077] The intelligent data processing solution equipped with an AI digital human in this embodiment of the invention includes: embedding an AI digital human in a digital dashboard, wherein the AI ​​digital human interacts with the user through voice, text, or gestures to obtain the user's demand information Q1; generating a multi-dimensional data visualization chart T based on the user's demand information Q1, and dynamically adjusting the chart layout, style, and interaction method based on the chart by selecting an adaptive algorithm and analyzing data characteristics in real time; analyzing the user's interaction history Jh using a deep learning model, and generating historical analysis data D1 and future prediction data D2 for the user based on the interaction history; obtaining the user's feedback information Q2 on the data visualization chart T, historical analysis data D1, and future prediction data D2, and determining the user's social media account information M based on the feedback information Q2; verifying the authenticity of the user's social media account information M, and generating different types of AI digital human expressions based on the results of the authenticity verification. This invention has the following beneficial effects:

[0078] A deeply embedded AI digital human data analysis system:

[0079] Through advanced facial expression synthesis, speech generation, and natural language processing technologies, AI digital humans become intelligent data analysis assistants for users, breaking the passive display mode of traditional data dashboards.

[0080] b. Intelligent interaction based on affective computing:

[0081] The system analyzes user emotions in real time and dynamically adjusts interactive content and display methods to make the user experience more humanized and personalized.

[0082] c. Deep self-learning capabilities for intelligent recommendation and prediction:

[0083] By combining deep learning and collaborative filtering algorithms, the system can continuously learn user preferences, automatically generate future trend predictions, and help users make quick decisions.

[0084] d. Multimodal interaction technology innovation:

[0085] It supports the integration of multiple interaction methods such as voice, gesture, and text, allowing users to freely choose the most suitable interaction method, greatly improving the flexibility and experience of user operation. Attached Figure Description

[0086] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0087] Figure 1 This is a schematic flowchart of an intelligent data processing method equipped with an AI digital human, provided by an embodiment of the present invention.

[0088] Figure 2 This is a schematic diagram of another intelligent data processing method equipped with an AI digital human provided in an embodiment of the present invention;

[0089] Figure 3 This is a schematic diagram of the structure of an intelligent data processing device equipped with an AI digital human, provided in an embodiment of the present invention.

[0090] Figure 4 A schematic diagram of an electronic device provided in an embodiment of the present invention. Detailed Implementation

[0091] The embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0092] The following specific examples illustrate the implementation of this disclosure. Those skilled in the art can easily understand other advantages and effects of this disclosure from the content disclosed in this specification. Obviously, the described embodiments are only a part of the embodiments of this disclosure, and not all of them. This disclosure can also be implemented or applied through other different specific embodiments, and the details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of this disclosure. It should be noted that, in the absence of conflict, the following embodiments and features in the embodiments can be combined with each other. Based on the embodiments in this disclosure, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this disclosure.

[0093] It should be noted that various aspects of embodiments within the scope of the appended claims are described below. It will be apparent that the aspects described herein can be embodied in a wide variety of forms, and any particular structure and / or function described herein is merely illustrative. Based on this disclosure, those skilled in the art will understand that one aspect described herein can be implemented independently of any other aspect, and two or more of these aspects can be combined in various ways. For example, any number of aspects set forth herein can be used to implement the device and / or practice the method. Additionally, this device and / or method can be implemented using structures and / or functionalities other than one or more of the aspects set forth herein.

[0094] It should also be noted that the illustrations provided in the following embodiments are only schematic representations of the basic concept of this disclosure. The drawings only show the components related to this disclosure and are not drawn according to the number, shape and size of the components in actual implementation. In actual implementation, the form, quantity and proportion of each component can be arbitrarily changed, and the layout of the components may also be more complex.

[0095] Furthermore, specific details are provided in the following description to facilitate a thorough understanding of the examples. However, those skilled in the art will understand that the described aspects can be practiced without these specific details.

[0096] This disclosure provides an intelligent data processing method for an AI-enabled digital human. The intelligent data processing method for an AI-enabled digital human provided in this embodiment can be executed by a computing device, which can be implemented as software or as a combination of software and hardware. This computing device can be integrated into a server, terminal device, or the like.

[0097] See Figure 1 and Figure 2 This disclosure provides an intelligent data processing method equipped with an AI digital human, including:

[0098] S101, An AI digital human is embedded in the digital dashboard. The AI ​​digital human interacts with the user through voice, text or gestures to obtain the user's demand information Q1.

[0099] The AI ​​digital human is the core module for system interaction with users, communicating through voice, facial expressions, and natural language understanding. It can not only answer user questions but also proactively guide users through complex data analysis. The core technologies of this module include the following components:

[0100] Facial expression and speech synthesis technology:

[0101] This invention uses a multimodal fusion model based on deep neural networks (DNNs) to generate facial expressions and speech in AI digital humans. The model is implemented through the following steps:

[0102] Facial Expression Synthesis: The AI ​​digital human analyzes the user's emotional state using a deep convolutional neural network (DCNN) to automatically generate corresponding facial expressions. The DCNN adjusts the AI ​​digital human's facial expressions in real time based on the input emotional labels (such as "happy," "confused," "anxious," etc.), generating rich facial details to ensure natural interaction.

[0103] Speech Synthesis: A speech synthesis system based on Generative Adversarial Networks (GANs) is employed to convert text content into speech output. This system utilizes the generative network of the GAN to generate natural speech waveforms, while a discriminative network ensures the synchronization and naturalness of speech and facial expressions. Compared to traditional text-to-speech (TTS) systems, GAN-synthesized speech is more natural and fluent.

[0104] Synchronized generation of lip movements and speech: Using self-supervised learning techniques, the lip movements of the AI ​​digital human are modeled to synchronize lip movements and speech generation. By training a self-supervised model, this module can achieve high-precision lip movement generation without relying on a large amount of labeled data, ensuring the matching of speech and lip movements.

[0105] Intelligent interaction strategy:

[0106] In this invention, the AI ​​digital human not only passively responds to user requests but also proactively interprets and makes recommendations based on data. This process is implemented using reinforcement learning (RL), as detailed below:

[0107] User intent prediction: The system constructs a Bayesian Network based on the user's interaction history and current operation context to predict the data or operation the user might be interested in next. For example, when a user is viewing interaction data for a post, the system can use the Bayesian Network to infer that the user might also be interested in the geographical distribution or time trend of that interaction data and proactively recommend related data.

[0108] Interaction Strategy Optimization: To enhance the rationality of the AI ​​digital human's proactive recommendation data and user satisfaction, the system employs a reinforcement learning algorithm based on Deep Q-Network (DQN). This algorithm continuously iterates and learns from user feedback to optimize the interaction strategy. DQN uses each user interaction as a learning signal to gradually optimize recommended content and interaction methods, making the AI ​​digital human more aligned with user needs.

[0109] S102, based on the user's needs information Q1, generates a multi-dimensional data visualization chart T, and dynamically adjusts the chart layout, style and interaction method in real time based on the chart selection adaptive algorithm and data characteristics.

[0110] The data visualization generation module is responsible for generating suitable multi-dimensional visualization charts based on user needs and interactive guidance from the AI ​​digital human. This module not only supports static data display but also features dynamic adaptation and scene awareness capabilities.

[0111] Traditional visualization systems often only display fixed-format charts, making it difficult to flexibly adapt to changes in different data. This invention introduces an adaptive data visualization engine, whose core algorithm is a deep learning model based on chart self-optimization. The specific steps are as follows:

[0112] Chart Selection Model: Through a multi-task deep learning model, the system can dynamically select the optimal visualization format based on data complexity, dimensionality, and user preferences. For example, as the dimensionality of the dataset increases, the model automatically switches from a simple bar chart to a more suitable multi-dimensional heatmap or scatter plot. The model's training data includes the visualization types historically selected by users, the feature dimensions of the data, and the display effects.

[0113] Adaptive Visualization Framework: This engine is based on an adaptive chart optimization algorithm. By analyzing data characteristics in real time (such as time span, geographical distribution, and numerical changes), it dynamically adjusts the chart layout, style, and interaction methods. For example, the system can dynamically adjust chart parameters such as the time axis and color gradient based on data trends and density, ensuring that the data display is always optimal.

[0114] The scene-aware visualization technology of this invention enables the system to automatically adjust the complexity of charts and the way information is presented based on the user's emotions and current task status. Its core technology is an emotion-driven visualization model, and the implementation steps are as follows:

[0115] Emotional Perception: The system utilizes a convolutional neural network (CNN) combined with a long short-term memory network (LSTM) to perform sentiment analysis on users' facial expressions, speech, and behavior. By capturing users' facial expressions and speech tone in real time through a camera and microphone, the system can identify users' emotional states (such as anxiety, confusion, satisfaction, etc.) and provide a reference for the visualization process.

[0116] Emotion-driven data simplification: Based on sentiment analysis results, the system automatically simplifies charts and highlights the most essential data when users feel stressed or confused. For example, when it detects that a user is exhibiting confusion or doubt, the system will automatically switch to simplification mode, hiding unnecessary details and displaying only key information to reduce the user's cognitive burden.

[0117] S103, use a deep learning model to analyze the user's interaction history Jh, and generate historical analysis data D1 and future prediction data D2 for the user based on the interaction history.

[0118] The intelligent recommendation and prediction module provides users with intelligent analysis suggestions and trend predictions based on historical data and current context. The core implementation technologies include deep learning models, collaborative filtering, and a self-learning prediction system, detailed below:

[0119] 4.3.1 Personalized Data Recommendations:

[0120] The system provides personalized data recommendations to users through collaborative filtering algorithms and user profile modeling. Its workflow is as follows:

[0121] User profile modeling: Based on users' historical interaction data and operational behaviors, the system constructs user profiles using a multilayer perceptron (MLP) neural network. This model learns users' operational habits and preferences from multiple dimensions, such as interaction frequency, metrics they pay attention to, and preference charts.

[0122] Collaborative filtering recommendation: Utilizing content-based collaborative filtering algorithms, the system generates personalized data recommendation lists by comparing user profiles with the operational history of other similar users. This technology proactively recommends dimensions or analytical perspectives that users may be interested in, thereby enhancing the user experience.

[0123] 4.3.2 Self-learning prediction model:

[0124] To provide more forward-looking analysis, the system introduces a time series forecasting model based on an LSTM neural network to predict future data trends. The implementation steps are as follows:

[0125] Data preprocessing: Historical data is cleaned and features are extracted to ensure the model can correctly identify trends and periodic changes in the data. A time-series feature enhancement algorithm is used to add key features from the time series (such as periodic changes and seasonal fluctuations) to the model input.

[0126] Time series forecasting: LSTM models, through their long short-term memory structure, can capture long-term dependencies in data and predict key future data points. This model is suitable for predicting data growth trends and user behavior patterns, and generating specific suggestions (such as fan growth prediction, interaction data growth prediction, etc.).

[0127] S104, obtain user feedback information Q2 regarding the data visualization chart T, historical analysis data D1, and future prediction data D2, and determine the user's social media account information M based on the feedback information Q2.

[0128] The user interface can be used to collect user feedback (Q2) on the data visualization chart (T), historical analysis data (D1), and future prediction data (D2). Feedback can include text comments, ratings, and clicks.

[0129] Natural language processing (NLP) techniques are used to analyze the sentiment polarity of user feedback to determine user satisfaction and opinions on the data. Based on the sentiment analysis results and user behavior data within the system, the system identifies the user's social media account information (M). The system can identify and verify a user's social media account information through the content of their feedback, interaction frequency, and behavioral patterns.

[0130] S105 verifies the authenticity of the user's social media account information M, and generates different types of AI digital human expressions based on the verification results.

[0131] The user's social media account information M can be extracted from the user's feedback information Q2. Various verification methods (such as account activity records, number of followers, authenticity of published content, etc.) can be used to verify the authenticity of the social media account information M.

[0132] Based on the verification results, the user's trust score S is calculated. The trust score can take into account factors such as account age, number of connections, number of posts, number of followers, and number of likes.

[0133] If a user's trust score is high, indicating that the account information is authentic and reliable, the AI ​​digitizer can generate expressions of satisfaction and trust, such as smiling and nodding. If a user's trust score is low, indicating that the account information may be questionable, the AI ​​digitizer can generate expressions of caution and suspicion, such as frowning and shaking its head.

[0134] As a specific implementation method, verifying the authenticity of a user's social media account information M may include the following steps S1051-S1055:

[0135] S1051. Collect multi-dimensional behavioral data DY from users on social media platforms, and perform data cleaning, deduplication, and standardization on the collected behavioral data to obtain standard data DB.

[0136] S1052, extract feature vector X from standard data DB using feature extraction function Ft, X=Ft(DB)=[Zh,Xw,Sw,Nt], where Zh represents account information features, Xw represents behavioral pattern features, Sw represents social network features, and Nt represents content features.

[0137] The feature extraction module is responsible for extracting multi-dimensional features from user behavior data to provide training data for the decision tree model. The extracted features include:

[0138] Account information characteristics:

[0139] Account creation time: Fake accounts usually have a short lifespan, and newer accounts may pose a higher risk.

[0140] Pay attention to the user growth rate: The user growth rate of fake accounts may be abnormal, such as gaining a large number of followers in a short period of time or having a constant number of followers over a long period of time.

[0141] Profile picture settings: Fake accounts may not set a profile picture or use the default profile picture.

[0142] Is the account description complete? Accounts with simple or blank descriptions are more likely to be fake accounts.

[0143] b. Behavioral pattern characteristics:

[0144] Posting frequency: Fake accounts typically post a large amount of information in a short period of time, or remain inactive for a long time.

[0145] Content similarity: The published content is highly repetitive or templated, and is usually machine-generated false information.

[0146] Abnormal interaction: behaviors such as frequently forwarding specific content, excessively high number of likes, or an abnormal number of meaningless comments.

[0147] c. Social network characteristics:

[0148] Friends / followers network: Fake accounts often follow each other with a large number of other fake accounts, forming an unnatural network structure.

[0149] Interaction depth: Fake accounts interact with other users less frequently and less deeply, and the interactions are often one-way.

[0150] d. Content characteristics:

[0151] Textual characteristics: Natural language processing (NLP) techniques are used to analyze the complexity and diversity of user-posted content. Fake accounts typically produce simple, repetitive content with little emotional fluctuation.

[0152] Sentiment analysis: Sentiment analysis of post content reveals that fake accounts often lack natural emotional expression.

[0153] S1053, calculate the entropy H(DB) of the dataset DB, and the conditional entropies H(DB∣Zh), H(DB∣Xw), H(DB∣Sw), and H(DB∣Nt) of the features Zh, Xw, Sw, and Nt. Based on the entropy H(DB) and the conditional entropies H(DB∣Zh), H(DB∣Xw), H(DB∣Sw), and H(DB∣Nt), calculate the information gain IG(DB,X) of the feature vector X. Determine the root node of the decision tree based on the information gain IG(DB,X).

[0154] A model can be trained using decision tree algorithms to identify fake accounts.

[0155] Feature selection and construction:

[0156] The system uses multi-dimensional features extracted by the feature extraction module as input data to construct feature vectors. To improve the model's performance, the system further selects and constructs these features.

[0157] Calculate the entropy H(DB) of the dataset DB:

[0158]

[0159] DB is the standard dataset, m is the number of categories of the category label Y corresponding to the standard dataset DB, and pj is the frequency of the j-th category label in the dataset DB.

[0160] Based on the different values ​​v1 of feature Zh, the dataset DB is divided into multiple subsets D. v1 For each subset D v1 Calculate the entropy H(D) v1 ), calculate the conditional entropy H(DB|Zh) based on the entropy of the subset and the size of the subset:

[0161]

[0162] Based on the different values ​​v2 of feature Xw, the dataset DB is divided into multiple subsets D. v2 For each subset D v2 Calculate the entropy H(D) v2 ), calculate the conditional entropy H(DB|Xw) based on the entropy of the subset and the size of the subset:

[0163]

[0164] Based on the different values ​​of feature Sw (v3), the dataset DB is divided into multiple subsets D. v3 For each subset D v3 Calculate the entropy H(D) v3 ), calculate the conditional entropy H(DB|Sw) based on the entropy of the subset and the size of the subset:

[0165]

[0166] Based on the different values ​​v4 of feature Nt, the dataset DB is divided into multiple subsets Dv4. For each subset Dv4, the entropy H(Dv4) is calculated. The conditional entropy H(DB|Nt) is calculated based on the entropy and size of the subset.

[0167]

[0168] |DB|、|D v1 |、|D v2 |、|D v3 |and|D v4 | represent the dataset DB and subset D, respectively. v1 D v2 D v3 D v4 The size of H is given by H(), which is the entropy calculation function.

[0169] According to a specific implementation of an embodiment of the present invention, the calculation of the entropy H(DB) of the dataset DB, and the conditional entropies H(DB∣Zh), H(DB∣Xw), H(DB∣Sw), and H(DB∣Nt) of the features Zh, Xw, Sw, and Nt, and the calculation of the information gain IG(DB,X) of the feature vector X based on the entropy H(DB) and the conditional entropies H(DB∣Zh), H(DB∣Xw), H(DB∣Sw), and H(DB∣Nt), further includes:

[0170] Calculate the information gain of each feature:

[0171] IG(DB,Zh)=H(DB)-H(DB|Zh)

[0172] IG(DB,Xw)=H(DB)-H(DB|Xw)

[0173] IG(DB,Sw)=H(DB)-H(DB|Sw)

[0174] IG(DB,Nt)=H(DB)-H(DB|Nt)

[0175] Calculate the main effect information gain :

[0176]

[0177] Calculate the information gain of feature interaction effects :

[0178]

[0179] IG(DB,Zh), IG(DB,Xw), IG(DB,Sw), and IG(DB,Nt) are the information gains of features Zh, Xw, Sw, and Nt, respectively. These are the weights of each feature. IG(DB,Z·Xw), IG(DB,Zh·Sw), IG(DB,Zh·Nt), IG(DB,Xw·Sw), IG(DB,Xw·Nt), and IG(DB,Sw·Nt) are the information gains of the feature interaction effects, respectively. It is the weight of the feature interaction effect;

[0180] The information gain IG(DB,X) of the feature vector X is expressed as:

[0181] .

[0182] S1054, based on the feature value of the root node, the standard data is divided into multiple subsets Dz, and based on the process of dividing the subsets, the decision child node Jz and decision rule Rj of the decision tree are generated. Each path of the decision tree represents a decision rule from the root node to the leaf node. The leaf node of the decision tree represents the classification result Fz, thus forming the decision tree DT=Fj(Dz,Jz,Rj,Fz), where Fj is the decision tree processing function.

[0183] The training process of a decision tree includes the following steps:

[0184] Building the root node: The system first selects the feature with the largest information gain as the root node, and then divides the dataset into multiple subsets based on this feature.

[0185] Recursively construct child nodes: For each subset, the system repeats the process of feature selection and dataset partitioning to generate new nodes until a stopping condition is met (such as the data in the node meeting purity requirements or the maximum depth limit of the tree).

[0186] Forming decision rules: Each path in the decision tree represents a decision rule from the root node to the leaf node, and the final leaf node represents the classification result (real account or fake account).

[0187] Pruning: To avoid overfitting, the system prunes the trained decision tree, removing redundant nodes. Pruning can be based on the minimum description length (MDL) principle or cross-validation.

[0188] S1055, after receiving the behavior data Dn of a new user, the behavior data Dn is subjected to feature extraction to generate a feature vector Xn corresponding to the behavior data Dn. Based on the feature vector Xn, the decision tree DT and the preset classification function Fc, a prediction result y=Fc(DT,Xn) corresponding to the behavior data Dn of the new user is generated. The prediction result is used to determine whether the user account corresponding to the behavior data Dn of the new user is a fake account.

[0189] After training, the decision tree model is used to identify whether a new user is a fake account. The identification process is as follows:

[0190] Input feature data: After the new user's behavior data is processed by the feature extraction module, a corresponding feature vector is generated.

[0191] Model prediction: The feature vector passes through each node of the decision tree in sequence, and is matched to the corresponding child node according to the feature value, until the leaf node is reached.

[0192] Output results: The output of the leaf nodes is the predicted classification result of the account (real account or fake account).

[0193] Through the above process, the system can verify the authenticity of users' social media account information and generate different types of AI digital human expressions based on the verification results, thereby providing a more personalized interactive experience.

[0194] According to a specific implementation of an embodiment of the present invention, embedding an AI digital human in a digital dashboard, wherein the AI ​​digital human interacts with the user through voice, text, or gestures to obtain the user's demand information Q1, includes:

[0195] Use the speech recognition API to convert the user's voice input into text in Qt;

[0196] Use natural language processing techniques to parse the user's text input Qr;

[0197] Capture user body movements and map user gestures to specific interaction commands Qm;

[0198] Based on text-based Qt, text input Qr, and interactive commands Qm, generate user requirement information Q1.

[0199] According to a specific implementation of an embodiment of the present invention, the step of generating a multi-dimensional data visualization chart T based on user demand information Q1, and dynamically adjusting the chart layout, style, and interaction method in real time based on a chart selection adaptive algorithm, includes:

[0200] Parse the user's requirement information Q1 to determine the data source, time range, data type, and chart type corresponding to the requirement information Q1;

[0201] Data is retrieved from a specified data source and preprocessed to obtain chart data;

[0202] Use a data visualization library to transform the chart data into a multi-dimensional data visualization chart T.

[0203] According to a specific implementation of an embodiment of the present invention, the step of generating a multi-dimensional data visualization chart T based on user demand information Q1, and dynamically adjusting the chart layout, style, and interaction method in real time based on a chart selection adaptive algorithm, further includes:

[0204] Real-time analysis of data characteristics, including data volume, data distribution characteristics, data change trends, and identification of key data features;

[0205] Automatically select the appropriate chart type based on data characteristics;

[0206] The layout of the chart is dynamically adjusted based on the amount of data and the chart type.

[0207] According to a specific implementation of an embodiment of the present invention, the step of verifying the authenticity of the user's social media account information M, and generating different types of AI digital human expressions based on the verification results, includes:

[0208] Calculate the trust score S of a user's social media account information M:

[0209]

[0210] S is the trust score, α, β, γ, δ, ϵ are the feature weights, A is the user account age, C is the number of user account connections, P is the number of posts published by the user account, F is the number of users followed by the user account, and L is the number of likes on the user account.

[0211] Based on trust scoring, the authenticity of a user's social media account information M is determined.

[0212] According to a specific implementation of an embodiment of the present invention, the step of verifying the authenticity of the user's social media account information M, and generating different types of AI digital human expressions based on the verification results, further includes:

[0213] Predicting facial expression selection probability using a multilayer perceptual model:

[0214]

[0215] y is the output vector, representing the probability of selecting each expression. Each element... This represents the probability of choosing the i-th expression. The Softmax function is used to transform an output vector into a probability distribution. The definition of the Softmax function is:

[0216]

[0217] Where z is the input vector and k is the number of output categories;

[0218] This is the weight matrix for the second layer, used to transform the output of the hidden layer into the final output layer. For activation function:

[0219]

[0220] This is the weight matrix for the first layer, used to transform the input feature vector x into the input for the hidden layer. x is the input feature vector, containing features extracted from user feedback and social media account information. This is the bias vector for the first layer, used to adjust the input to the hidden layers. This is the bias vector for the second layer, used to adjust the input to the output layer.

[0221] For a corresponding method embodiment, see [link to relevant documentation]. Figure 3 The present invention also discloses an intelligent data processing device 30 equipped with an AI digital human, comprising:

[0222] The acquisition module 301 embeds an AI digital human in the digital dashboard. The AI ​​digital human interacts with the user through voice, text or gestures to acquire the user's demand information Q1.

[0223] The generation module 302 generates a multi-dimensional data visualization chart T based on the user's requirement information Q1, and dynamically adjusts the chart layout, style and interaction method based on the chart selection adaptive algorithm and real-time analysis of data characteristics.

[0224] Analysis module 303 uses a deep learning model to analyze the user's interaction history Jh, and generates historical analysis data D1 and future prediction data D2 for the user based on the interaction history;

[0225] The determination module 304 obtains user feedback information Q2 regarding the data visualization chart T, historical analysis data D1, and future prediction data D2, and determines the user's social media account information M based on the feedback information Q2;

[0226] The verification module 305 verifies the authenticity of the user's social media account information M, and generates different types of AI digital human expressions based on the verification results.

[0227] See Figure 4This invention also provides an electronic device 60, which includes:

[0228] At least one processor; and,

[0229] The memory is communicatively connected to the at least one processor; wherein,

[0230] The memory stores instructions that can be executed by the at least one processor, which enables the at least one processor to perform the intelligent data processing method with AI digital human in the foregoing method embodiments.

[0231] This invention also provides a non-transitory computer-readable storage medium that stores computer instructions for causing the computer to perform the aforementioned method embodiments.

[0232] This invention also provides a computer program product, which includes a computing program stored on a non-transitory computer-readable storage medium. The computer program includes program instructions, which, when executed by a computer, cause the computer to perform the intelligent data processing method with AI digital human described in the foregoing method embodiments.

[0233] The following is for reference. Figure 4 The diagram illustrates a structural schematic of an electronic device 60 suitable for implementing embodiments of the present disclosure. The electronic devices in the embodiments of the present disclosure may include, but are not limited to, mobile terminals such as mobile phones, laptops, digital broadcast receivers, PDAs (personal digital assistants), PADs (tablet computers), PMPs (portable multimedia players), in-vehicle terminals (e.g., in-vehicle navigation terminals), and fixed terminals such as digital TVs and desktop computers. Figure 4 The electronic device shown is merely an example and should not be construed as limiting the functionality and scope of the embodiments disclosed herein.

[0234] like Figure 4 As shown, electronic device 60 may include a processing device (e.g., a central processing unit, a graphics processor, etc.) 601, which can perform various appropriate actions and processes according to a program stored in read-only memory (ROM) 602 or a program loaded from storage device 608 into random access memory (RAM) 603. RAM 603 also stores various programs and data required for the operation of electronic device 60. Processing device 601, ROM 602, and RAM 603 are interconnected via bus 604. Input / output (I / O) interface 605 is also connected to bus 604.

[0235] Typically, the following devices can be connected to I / O interface 605: input devices 606 including, for example, touchscreens, touchpads, keyboards, mice, image sensors, microphones, accelerometers, gyroscopes, etc.; output devices 607 including, for example, liquid crystal displays (LCDs), speakers, vibrators, etc.; storage devices 608 including, for example, magnetic tapes, hard disks, etc.; and communication devices 609. Communication device 609 allows electronic device 60 to communicate wirelessly or wiredly with other devices to exchange data. Although Figure 4 An electronic device 60 with various devices is shown; however, it should be understood that it is not required to implement or possess all of the devices shown. More or fewer devices may be implemented or possessed alternatively.

[0236] In particular, according to embodiments of this disclosure, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of this disclosure include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication device 609, or installed from a storage device 608, or installed from a ROM 602. When the computer program is executed by the processing device 601, it performs the functions defined in the methods of embodiments of this disclosure.

[0237] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for intelligent data processing equipped with an AI digital human, characterized in that, include: An AI digital human is embedded in a digital dashboard. The AI ​​digital human interacts with the user through voice, text, or gestures to obtain the user's needs information Q1. Based on the user's needs information Q1, generate a multi-dimensional data visualization chart T, and based on the chart, select an adaptive algorithm to dynamically adjust the chart layout, style and interaction method in real time by analyzing data characteristics. The user's interaction history Jh is analyzed using a deep learning model, and historical analysis data D1 and future prediction data D2 are generated for the user based on the interaction history. Obtain user feedback information Q2 regarding data visualization chart T, historical analysis data D1, and future prediction data D2, and determine the user's social media account information M based on the feedback information Q2; The authenticity of the user's social media account information M is verified, and based on the verification results, different types of AI digital human expressions are generated; among them The step of using a deep learning model to analyze the user's interaction history Jh, and generating historical analysis data D1 and future prediction data D2 for the user based on the interaction history, includes: The state update formula for building a deep learning model: For the output of the forget gate, The Sigmoid activation function, with an output range between 0 and 1, controls the degree to which information is preserved. Here is the weight matrix for the forget gate. This is the hidden state from the previous moment. For the current input, This is the bias vector for the forget gate; The output of the input gate, Here is the weight matrix of the input gate. The bias vector for the input gate; The candidate cell state is new information calculated using the current input and the hidden state from the previous time step. It is a hyperbolic tangent activation function, with an output range between -1 and 1. This is the weight matrix for the candidate cell states. is the bias vector for the candidate cell state; This refers to the current state of memory cells. For element-wise multiplication, This represents the state of memory cells at the previous moment; For the output of the output gate, This is the weight matrix of the output gate. This is the bias vector for the output gate; This is the hidden state at the current moment, used to pass it to the next time step. Memory cell state The value after applying the hyperbolic tangent activation function; The time series is decomposed into trend, seasonal, and random components using an additive model: The original time series data, The trend component represents the long-term trend of the data. The seasonal component indicates the periodic fluctuations in the data. The random component represents random noise or unpredictable parts of the data.

2. The method according to claim 1, characterized in that, The embedding of an AI digital human in the digital dashboard, wherein the AI ​​digital human interacts with the user through voice, text, or gestures to obtain the user's demand information Q1, includes: Use the speech recognition API to convert the user's voice input into text in Qt; Use natural language processing techniques to parse the user's text input Qr; Capture user body movements and map user gestures to specific interaction commands Qm; Based on text-based Qt, text input Qr, and interactive commands Qm, generate user requirement information Q1.

3. The method according to claim 2, characterized in that, Based on the user's demand information Q1, a multi-dimensional data visualization chart T is generated. Then, based on a chart selection adaptive algorithm, the chart layout, style, and interaction methods are dynamically adjusted in real time by analyzing data characteristics. This includes: Parse the user's requirement information Q1 to determine the data source, time range, data type, and chart type corresponding to the requirement information Q1; Data is retrieved from a specified data source and preprocessed to obtain chart data; Use a data visualization library to transform the chart data into a multi-dimensional data visualization chart T.

4. The method according to claim 3, characterized in that, The process of generating a multi-dimensional data visualization chart T based on user demand information Q1, and dynamically adjusting the chart layout, style, and interaction method based on a chart selection adaptive algorithm to analyze data characteristics in real time, also includes: Real-time analysis of data characteristics, including data volume, data distribution characteristics, data change trends, and identification of key data features; Automatically select the appropriate chart type based on data characteristics; The layout of the chart is dynamically adjusted based on the amount of data and the chart type.

5. The method according to claim 1, characterized in that, The step of obtaining user feedback information Q2 regarding data visualization chart T, historical analysis data D1, and future prediction data D2, and determining the user's social media account information M based on the feedback information Q2, includes: The logistic regression model was used to analyze the sentiment polarity of user feedback, and the probability P(y=1|x) of positive sentiment in user feedback was calculated: It is the Sigmoid activation function, w is the weight vector, x is the feature vector, and b is the bias term; Based on the sentiment analysis results of user feedback information Q2, and combined with the user's behavioral data in the system, the user's social media account information M is determined.

6. The method according to claim 5, characterized in that, The process involves verifying the authenticity of the user's social media account information M, and generating different types of AI digital human expressions based on the verification results, including: Calculate the trust score S of a user's social media account information M: S is the trust score, α, β, γ, δ, ϵ are the feature weights, A is the user account age, C is the number of user account connections, P is the number of posts published by the user account, F is the number of users followed by the user account, and L is the number of likes on the user account. Based on trust scoring, the authenticity of a user's social media account information M is determined.

7. The method according to claim 6, characterized in that, The process of verifying the authenticity of the user's social media account information M, and generating different types of AI digital human expressions based on the verification results, also includes: Predicting facial expression selection probability using a multilayer perceptual model: y is the output vector, representing the probability of selecting each expression. Each element... This represents the probability of choosing the i-th expression. The Softmax function is used to transform an output vector into a probability distribution. The definition of the Softmax function is: Where z is the input vector and k is the number of output categories; This is the weight matrix for the second layer, used to transform the output of the hidden layer into the final output layer. For activation function: This is the weight matrix for the first layer, used to transform the input feature vector x into the input for the hidden layer. x is the input feature vector, containing features extracted from user feedback and social media account information. This is the bias vector for the first layer, used to adjust the input to the hidden layers. This is the bias vector for the second layer, used to adjust the input to the output layer.

8. An intelligent data processing device equipped with an AI digital human, characterized in that, include: The acquisition module embeds an AI digital human into the digital dashboard. The AI ​​digital human interacts with the user through voice, text, or gestures to acquire the user's demand information Q1. The generation module generates multi-dimensional data visualization charts T based on the user's needs information Q1, and dynamically adjusts the chart layout, style and interaction method based on the chart selection adaptive algorithm and real-time analysis of data characteristics. The analysis module uses a deep learning model to analyze the user's interaction history Jh, and generates historical analysis data D1 and future prediction data D2 for the user based on the interaction history. The module determines the user's social media account information M based on the feedback information Q2 from the user regarding the data visualization chart T, historical analysis data D1, and future prediction data D2. The verification module verifies the authenticity of the user's social media account information M, and generates different types of AI digital human expressions based on the verification results; among which... The step of using a deep learning model to analyze the user's interaction history Jh, and generating historical analysis data D1 and future prediction data D2 for the user based on the interaction history, includes: The state update formula for building a deep learning model: For the output of the forget gate, The Sigmoid activation function, with an output range between 0 and 1, controls the degree to which information is preserved. Here is the weight matrix for the forget gate. This is the hidden state from the previous moment. For the current input, This is the bias vector for the forget gate; The output of the input gate, Here is the weight matrix of the input gate. The bias vector for the input gate; The candidate cell state is new information calculated using the current input and the hidden state from the previous time step. It is a hyperbolic tangent activation function, with an output range between -1 and 1. This is the weight matrix for the candidate cell states. is the bias vector for the candidate cell state; This refers to the current state of memory cells. For element-wise multiplication, This represents the state of memory cells at the previous moment; For the output of the output gate, This is the weight matrix of the output gate. This is the bias vector for the output gate; This is the hidden state at the current moment, used to pass it to the next time step. Memory cell state The value after applying the hyperbolic tangent activation function; The time series is decomposed into trend, seasonal, and random components using an additive model: The original time series data, The trend component represents the long-term trend of the data. The seasonal component indicates the periodic fluctuations in the data. The random component represents random noise or unpredictable parts of the data.

9. An electronic device, characterized in that, The electronic device includes: At least one processor; and, A memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor, which, when executed by the at least one processor, enables the at least one processor to perform the intelligent data processing method for an AI-equipped digital human as described in any one of claims 1 to 7.

Images