Intelligent data processing method and device carrying AI digital human

Abstract

The embodiment of the invention provides an intelligent data processing method and device carrying an AI digital human and electronic equipment, and belongs to the technical field of data processing.The method comprises the steps that demand information Q1 of a user is obtained; generating a multi-dimensional data visualization chart T based on the demand information Q1 of the user; analyzing the interaction history Jh of the user by using a deep learning model; obtaining feedback information Q2 of the user for the data visualization chart T, the historical analysis data D1 and the future prediction data D2, and determining social media account information M of the user based on the feedback information Q2; and carrying out authenticity verification on the social media account information M of the user, and generating different types of AI digital human expressions based on an authenticity verification result. According to the invention, the interactivity and the intelligent level of the data billboard can be improved.

Description

Technical Field

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

[0002] Smart Data Dashboard is an advanced data visualization tool that helps enterprise users understand and use data more efficiently by integrating multiple data sources, applying advanced analytical techniques, and providing rich interactive functions. Smart Data Dashboard not only displays data, but also provides in-depth data insights in an intelligent way to support decision-making.

[0003] The existing technology has the following problems in the interactive and intelligent development process of the intelligent data dashboard system:

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

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

[0006] 3. Insufficient real-time interaction / interaction: Although existing data dashboards have basic real-time data update functions, they lack real-time interactive functions based on natural language processing (NLP) and artificial intelligence technologies, and users cannot quickly obtain analytical conclusions through natural language.

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

[0008] The above problems become technical problems that need to be solved. Summary of the invention

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

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

[0011] An AI digital human is embedded in the digital signage, and the AI ​​digital human interacts with the user through voice, text or gesture to obtain the user's demand information Q1;

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

[0013] Analyze the user's interaction history Jh using a deep learning model, and generate historical analysis data D1 and future prediction data D2 for the user based on the interaction history;

[0014] Obtain user feedback information Q2 regarding the data visualization chart T, the historical analysis data D1, and the 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 results of the authenticity verification.

[0016] According to a specific implementation of an embodiment of the present invention, the AI ​​digital human is embedded in the digital signage, and the AI ​​digital human interacts with the user through voice, text or gesture to obtain the user's demand information Q1, including:

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

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

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

[0020] Based on the text Qt, text input Qr and interactive command Qm, the user's demand information Q1 is generated.

[0021] According to a specific implementation of an embodiment of the present invention, based on the user's demand information Q1, a multi-dimensional data visualization chart T is generated, and an adaptive algorithm is selected based on the chart to dynamically adjust the chart layout, style and interaction mode by analyzing data characteristics in real time, including:

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

[0023] Get data from the specified data source, and pre-process the acquired data to obtain chart data;

[0024] The chart data is converted into a multi-dimensional data visualization chart T using a data visualization library.

[0025] According to a specific implementation of an embodiment of the present invention, based on the user's demand information Q1, a multi-dimensional data visualization chart T is generated, and an adaptive algorithm is selected based on the chart to dynamically adjust the chart layout, style and interaction mode by analyzing data characteristics in real time, and further includes:

[0026] Analyze data characteristics in real time, including data volume, data distribution characteristics, data change trends, and identify key features of data;

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

[0028] Dynamically adjust the layout of the chart based on the amount of data and chart type.

[0029] According to a specific implementation of an embodiment of the present invention, the method of analyzing the user's interaction history Jh by 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 is:

[0031]

[0032] is the output of the forget gate, It is a Sigmoid activation function with an output range between 0 and 1, which is used to control the degree of information retention. is the weight matrix of the forget gate, is the hidden state at the previous moment, is the input at the current moment, is the bias vector of the forget gate;

[0033]

[0034] is the output of the input gate, is the weight matrix of the input gate, is the bias vector of the input gate;

[0035]

[0036] is the candidate cell state, which is the new information calculated by the current input and the hidden state of the previous moment. is the hyperbolic tangent activation function, and the output range is between -1 and 1. is the weight matrix of the candidate cell state, is the bias vector of the candidate cell state;

[0037]

[0038] is the current state of the memory cell, is an element-by-element multiplication operation, is the state of the memory cell at the previous moment;

[0039]

[0040] is the output of the output gate, is the weight matrix of the output gate, is the bias vector of the output gate;

[0041]

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

[0043] Use an additive model to decompose the time series into trend, seasonal, and random components:

[0044]

[0045] is the original time series data, is the trend component, which indicates the long-term trend of the data. is the seasonal component, indicating the periodic fluctuation of the data. is the random component, which represents the random noise or unpredictable part of the data.

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

[0047] Use the logistic regression model to analyze the sentiment polarity of user feedback and calculate the probability P(y=1|x) that the user feedback is positive:

[0048]

[0049] 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 the user feedback information Q2 and combined with the user's behavior 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 authenticity verification of the user's social media account information M is performed, and based on the authenticity verification result, different types of AI digital human expressions are generated, including:

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

[0053]

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

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

[0056] According to a specific implementation of an embodiment of the present invention, the authenticity verification of the user's social media account information M and the generation of different types of AI digital human expressions based on the authenticity verification result also include:

[0057] Use a multi-layer perceptron model to predict the probability of expression selection:

[0058]

[0059] y is the output vector, which represents the selection probability of each expression. Each element represents the probability of selecting the i-th expression, is the Softmax function, which is used to convert the output vector into a probability distribution. The Softmax function is defined as:

[0060]

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

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

[0063]

[0064] is the weight matrix of the first layer, which is used to convert the input feature vector x into the input of the hidden layer. x is the input feature vector, which contains features extracted from user feedback and social media account information. is the bias vector of the first layer, which is used to adjust the input of the hidden layer. is the bias vector of the second layer, which is used to adjust the input of the output layer.

[0065] In a second aspect, an embodiment of the present invention provides an intelligent data processing device equipped with an AI digital human, comprising:

[0066] An acquisition module embeds an AI digital human in the digital signage, and the AI ​​digital human interacts with the user through voice, text or gesture to acquire the user's demand information Q1;

[0067] The generation module generates a multi-dimensional data visualization chart T based on the user's demand information Q1, and selects an adaptive algorithm based on the chart to dynamically adjust the chart layout, style and interaction method by analyzing data characteristics in real time;

[0068] An analysis module, which 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] A determination module obtains user feedback information Q2 regarding the data visualization chart T, the historical analysis data D1, and the future prediction data D2, and determines the user's social media account information M based on the feedback information Q2;

[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 results of the authenticity verification.

[0071] In a third aspect, an embodiment of the present invention further provides an electronic device, the electronic device comprising:

[0072] at least one processor; and,

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

[0074] The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the intelligent data processing method equipped with an AI digital human in any of the aforementioned first aspects or any of the implementations of the first aspect.

[0075] In a fourth aspect, an embodiment of the present invention further provides a non-transitory computer-readable storage medium, which stores computer instructions, and the computer instructions are used to enable the computer to execute the intelligent data processing method equipped with an AI digital human in the aforementioned first aspect or any implementation method of the first aspect.

[0076] In the fifth aspect, an embodiment of the present invention further provides a computer program product, which includes a computer program stored on a non-transitory computer-readable storage medium, and the computer program includes program instructions. When the program instructions are executed by a computer, the computer executes the intelligent data processing method equipped with an AI digital human in the aforementioned first aspect or any implementation method of the first aspect.

[0077] The intelligent data processing scheme equipped with AI digital human in the embodiment of the present invention includes: embedding AI digital human in the digital signage, the AI ​​digital human interacting with the user through voice, text or gesture 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 selecting an adaptive algorithm based on the chart, analyzing the data characteristics in real time to dynamically adjust the chart layout, style and interaction mode; 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; obtaining the user's feedback information Q2 for 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. The present invention has the following beneficial effects:

[0078] a AI digital human is deeply embedded in the data analysis system:

[0079] Through advanced 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 emotional computing:

[0081] The system analyzes user emotions in real time and dynamically adjusts the interactive content and display methods to make the user experience more humane 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 predictions for future trends, 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. Users can freely choose the most suitable interaction method, which greatly improves the flexibility and experience of user operation. BRIEF DESCRIPTION OF THE DRAWINGS

[0086] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0087] Figure 1 A schematic diagram of a flow chart of an intelligent data processing method equipped with an AI digital human provided by an embodiment of the present invention;

[0088] Figure 2 A schematic flow chart of another intelligent data processing method equipped with an AI digital human provided by an embodiment of the present invention;

[0089] Figure 3 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 by an embodiment of the present invention. DETAILED DESCRIPTION

[0091] The embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0092] The following describes the embodiments of the present disclosure through specific examples, and those skilled in the art can easily understand other advantages and effects of the present disclosure from the contents disclosed in this specification. Obviously, the described embodiments are only a part of the embodiments of the present disclosure, rather than all of the embodiments. The present disclosure can also be implemented or applied through other different specific embodiments, and the details in this specification can also be modified or changed in various ways based on different viewpoints and applications without departing from the spirit of the present disclosure. It should be noted that the following embodiments and features in the embodiments can be combined with each other without conflict. Based on the embodiments in the present disclosure, all other embodiments obtained by ordinary technicians in the field without making creative work are within the scope of protection of the present disclosure.

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

[0094] It should also be noted that the illustrations provided in the following embodiments are only schematic illustrations of the basic concept of the present disclosure. The drawings only show components related to the present disclosure rather than being drawn according to the number, shape and size of components in actual implementation. In actual implementation, the type, quantity and proportion of each component may be changed arbitrarily, and the component layout may also be more complicated.

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

[0096] The embodiment of the present disclosure provides an intelligent data processing method equipped with an AI digital human. The intelligent data processing method equipped with an AI 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, and the computing device can be integrated in a server, terminal device, etc.

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

[0098] S101, embedding an AI digital human in a digital signage, wherein the AI ​​digital human interacts with a user through voice, text or gesture to obtain user demand information Q1.

[0099] AI digital human is the core module for the system to interact with users, communicating with users through voice, expression and natural language understanding. It can not only answer users' questions, but also actively guide users to conduct complex data analysis. The core technologies of this module include the following parts:

[0100] Expression and speech synthesis technology:

[0101] The present invention uses a multimodal fusion model based on a deep neural network (DNN) to generate the expressions and voices of AI digital humans. The model is implemented through the following steps:

[0102] Expression synthesis: AI digital human analyzes the user's emotional state through deep convolutional neural network (DCNN) and automatically generates corresponding facial expressions. DCNN adjusts the facial expression of AI digital human in real time according to the input emotional labels (such as "happy", "confused", "anxious", etc.), generates rich expression details, and ensures the naturalness of interaction.

[0103] Speech synthesis: A speech synthesis system based on a generative adversarial network (GAN) is used to convert text content into speech output. The system uses the generative network of GAN to generate natural speech waveforms, while using the discriminative network to ensure the synchronization and naturalness of speech and expression. Compared with traditional text-to-speech (TTS) systems, GAN synthesized speech is more natural and fluent.

[0104] Synchronous generation of lip movements and speech: Using self-supervised learning technology, the lip movements of AI digital people 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, AI digital humans not only respond passively to user requests, but can also actively interpret and recommend data. This process is based on reinforcement learning (RL), and the specific details are as follows:

[0107] User intention prediction: The system builds a Bayesian network based on the user's interaction history and current operation context to predict the data or operation that the user may be interested in next. For example, when a user is viewing the interaction data of a post, the system can infer based on the Bayesian network that the user may also be interested in the geographical distribution or time trend of the interaction data, and actively recommend relevant data.

[0108] Interaction strategy optimization: In order to enhance the rationality of AI digital human's proactive recommendation data and user satisfaction, the system uses a reinforcement learning algorithm based on the deep Q network (DQN) to optimize the interaction strategy through continuous iteration and learning of user feedback. DQN uses each user's interactive behavior as a learning signal, gradually optimizes the recommended content and interaction methods, and makes AI digital human more in line with user needs.

[0109] S102, based on the user's demand information Q1, a multi-dimensional data visualization chart T is generated, and an adaptive algorithm is selected based on the chart to dynamically adjust the chart layout, style and interaction method by analyzing data characteristics in real time.

[0110] The data visualization generation module is responsible for generating suitable multi-dimensional visualization charts based on user needs and the interactive guidance of AI digital people. This module not only supports static data display, but also has the functions of dynamic adaptation and scene perception.

[0111] Traditional visualization systems can only display fixed-form charts and are difficult to flexibly respond 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 best visualization form based on the complexity, dimension and user preference of the data. For example, when the dimension of the data set increases, the model automatically switches from a simple bar chart to a more suitable multi-dimensional heat map or scatter plot. The training data of the model includes the visualization type selected by the user in history, the characteristic dimension of the data and the display effect.

[0113] Adaptive visualization framework: Based on the chart optimization adaptive algorithm, the engine dynamically adjusts the chart layout, style and interaction mode by real-time analysis of data characteristics (such as time span, geographical distribution, value changes, etc.). For example, the system can dynamically adjust chart parameters such as time axis and color gradient according to the trend and density of the data, so that the data display effect is always optimal.

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

[0115] Emotional perception: Use convolutional neural networks (CNN) combined with long short-term memory networks (LSTM) to analyze the user's expressions, voices, and behaviors. By capturing the user's facial expressions and voice intonation in real time through cameras and microphones, the system can identify the user's emotional state (such as anxiety, confusion, satisfaction, etc.) and provide a reference for the visualization process.

[0116] Emotion-driven data simplification: Based on the results of sentiment analysis, the system will automatically simplify the chart and highlight the most core data when the user feels stressed or confused. For example, when it detects that the user is confused or doubtful, the system will automatically switch to simplified mode, hiding unnecessary details and only showing key information to reduce the user's cognitive burden.

[0117] S103, 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.

[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 self-learning prediction systems. The specific details are as follows:

[0119] 4.3.1 Personalized data recommendation:

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

[0121] User portrait modeling: Based on the user's historical interaction data and operation behavior, the system builds a user portrait through a multi-layer perceptron (MLP) neural network. The model learns the user's operation habits and preferences from multiple dimensions such as the user's interaction frequency, the indicators they pay attention to, and the preference chart.

[0122] Collaborative filtering recommendation: Using the content-based collaborative filtering algorithm, the system generates a personalized data recommendation list by comparing the user profile with the operation history of other similar users. This technology can actively recommend dimensions or analysis angles that users may be interested in, thereby enhancing the user experience.

[0123] 4.3.2 Self-learning prediction model:

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

[0125] Data preprocessing: Clean and extract features from historical data to ensure that the model can correctly identify trends and cyclical changes in the data. Use a time series feature enhancement algorithm to add key features in the time series (such as cyclical changes, seasonal fluctuations, etc.) to the model input.

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

[0127] S104, obtaining user feedback information Q2 regarding 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.

[0128] Feedback information Q2 from users on the data visualization chart T, historical analysis data D1, and future prediction data D2 can be collected through the user interface. The feedback information may include text comments, ratings, click behaviors, and the like.

[0129] Use natural language processing technology to analyze the sentiment polarity of user feedback and determine the user's satisfaction and opinions on the data. Based on the sentiment analysis results of user feedback and the user's behavior data in the system, determine the user's social media account information M. The system can identify and verify the user's social media account information through the user's feedback content, interaction frequency, and behavior pattern.

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

[0131] The user's social media account information M can be extracted from the feedback information Q2 provided by the user, and the authenticity of the social media account information M can be verified using a variety of verification methods (such as account activity records, number of fans, authenticity of published content, etc.).

[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 published content, number of followed users, and number of likes.

[0133] If the user's trust score is high, it means that the account information is authentic and reliable, and the AI ​​digital human can generate expressions that express satisfaction and trust, such as smiles, nods, etc. If the user's trust score is low, it means that there may be doubts about the account information, and the AI ​​digital human can generate expressions that express caution and suspicion, such as frowns, shakes of the head, etc.

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

[0135] S1051, collect multi-dimensional behavior data DY of users from the social media platform, perform data cleaning, data deduplication and data standardization on the collected behavior data, and obtain standard data DB.

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

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

[0138] aAccount information characteristics:

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

[0140] Growth rate of followers: The growth rate of followers of fake accounts may be abnormal, such as gaining a large number of followers in a short period of time or keeping the number of followers unchanged for a long time.

[0141] Avatar setting: Fake accounts may not set an avatar or use a default avatar.

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

[0143] bBehavioral pattern characteristics:

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

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

[0146] Abnormal interactions: behaviors such as frequently forwarding a specific content, excessive number of likes, or an abnormal number of meaningless comments.

[0147] c Social network characteristics:

[0148] Friends / following user networks: Fake accounts often follow a large number of other fake accounts, forming an unnatural network structure.

[0149] Interaction depth: The frequency and depth of interaction between fake accounts and other users are shallow, and the interaction is often one-way.

[0150] dContent features:

[0151] Text features: Natural language processing (NLP) technology is used to analyze the complexity and diversity of user-posted content. The content of fake accounts is usually simple, repetitive, and has little emotional fluctuation.

[0152] Sentiment analysis: Conduct sentiment analysis on the content of posts. The content of fake accounts often lacks natural emotional expression.

[0153] S1053, calculate the entropy H(DB) of the data set DB, and the conditional entropies H(DB|Zh), H(DB|Xw), H(DB|Sw), H(DB|Nt) of the features Zh, Xw, Sw, and Nt, calculate 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), H(DB|Nt), and determine the root node of the decision tree based on the information gain IG(DB,X).

[0154] The decision tree algorithm can be used to train the model and identify fake accounts:

[0155] aFeature selection and construction:

[0156] The multi-dimensional features extracted by the feature extraction module are used as input data to construct feature vectors. In order to improve the effect of the model, the system further selects and constructs these features.

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

[0158]

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

[0160] According to 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) according to the entropy of the subset and the size of the subset:

[0161]

[0162] According to 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) according to the entropy of the subset and the size of the subset:

[0163]

[0164] According to the different values ​​v3 of the feature Sw, the data set 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) according to the entropy of the subset and the size of the subset:

[0165]

[0166] According to the different values ​​v4 of feature Nt, the data set DB is divided into multiple subsets Dv4, and the entropy H(Dv4) is calculated for each subset Dv4. The conditional entropy H(DB|Nt) is calculated based on the entropy of the subset and the size of the subset:

[0167]

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

[0169] According to a specific implementation method of an embodiment of the present invention, the entropy H(DB) of the data set DB and the conditional entropies H(DB|Zh), H(DB|Xw), H(DB|Sw), H(DB|Nt) of the features Zh, Xw, Sw, and Nt are calculated, and the information gain IG(DB,X) of the feature vector X is calculated based on the entropy H(DB) and the conditional entropies H(DB|Zh), H(DB|Xw), H(DB|Sw), and H(DB|Nt), and 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 main effect information gain :

[0176]

[0177] Calculate feature interaction effect information gain :

[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. is the weight 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, 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 characteristic 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 sub-nodes Jz and decision rules 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, thereby forming a decision tree DT=Fj(Dz,Jz,Rj,Fz), where Fj is the decision tree processing function.

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

[0184] Constructing the root node: The system first selects the feature with the largest information gain as the root node, and divides the data set 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, generating new nodes until a stopping condition is reached (such as the data in the node meets the purity requirement, or the maximum depth limit of the tree).

[0186] Forming decision rules: Each path of 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 and removes redundant nodes. Pruning can be based on the minimum description length (MDL) principle or the cross-validation method.

[0188] S1055, after receiving the behavior data Dn of the new user, perform feature extraction on the behavior data Dn to generate a feature vector Xn corresponding to the behavior data Dn, and based on the feature vector Xn, the decision tree DT and the preset classification function Fc, generate a prediction result y=Fc(DT,Xn) corresponding to the behavior data Dn of the new user. 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 is completed, 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: The new user's behavior data is processed by the feature extraction module to generate the corresponding feature vector.

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

[0192] Result output: The output result of the leaf node 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 the user's 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, the AI ​​digital human is embedded in the digital signage, and the AI ​​digital human interacts with the user through voice, text or gesture to obtain the user's demand information Q1, including:

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

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

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

[0198] Based on the text Qt, text input Qr and interactive command Qm, the user's demand information Q1 is generated.

[0199] According to a specific implementation of an embodiment of the present invention, based on the user's demand information Q1, a multi-dimensional data visualization chart T is generated, and an adaptive algorithm is selected based on the chart to dynamically adjust the chart layout, style and interaction mode by analyzing data characteristics in real time, including:

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

[0201] Get data from the specified data source, and pre-process the acquired data to obtain chart data;

[0202] The chart data is converted into a multi-dimensional data visualization chart T using a data visualization library.

[0203] According to a specific implementation of an embodiment of the present invention, based on the user's demand information Q1, a multi-dimensional data visualization chart T is generated, and an adaptive algorithm is selected based on the chart to dynamically adjust the chart layout, style and interaction mode by analyzing data characteristics in real time, and further includes:

[0204] Analyze data characteristics in real time, including data volume, data distribution characteristics, data change trends, and identify key features of data;

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

[0206] Dynamically adjust the layout of the chart based on the amount of data and chart type.

[0207] According to a specific implementation of an embodiment of the present invention, the method of analyzing the user's interaction history Jh by 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:

[0208] The state update formula for building a deep learning model is:

[0209]

[0210] is the output of the forget gate, It is a Sigmoid activation function with an output range between 0 and 1, which is used to control the degree of information retention. is the weight matrix of the forget gate, is the hidden state at the previous moment, is the input at the current moment, is the bias vector of the forget gate;

[0211]

[0212] is the output of the input gate, is the weight matrix of the input gate, is the bias vector of the input gate;

[0213]

[0214] is the candidate cell state, which is the new information calculated by the current input and the hidden state of the previous moment. is the hyperbolic tangent activation function, and the output range is between -1 and 1. is the weight matrix of the candidate cell state, is the bias vector of the candidate cell state;

[0215]

[0216] is the current state of the memory cell, is an element-by-element multiplication operation, is the state of the memory cell at the previous moment;

[0217]

[0218] is the output of the output gate, is the weight matrix of the output gate, is the bias vector of the output gate;

[0219]

[0220] is the hidden state at the current moment, which is used to pass to the next time step. Memory cell state The value after the hyperbolic tangent activation function;

[0221] Use an additive model to decompose the time series into trend, seasonal, and random components:

[0222]

[0223] is the original time series data, is the trend component, which indicates the long-term trend of the data. is the seasonal component, indicating the periodic fluctuation of the data. is the random component, which represents the random noise or unpredictable part of the data.

[0224] According to a specific implementation of an embodiment of the present invention, the step of obtaining user feedback information Q2 regarding the data visualization chart T, the historical analysis data D1, and the future prediction data D2, and determining the user's social media account information M based on the feedback information Q2, includes:

[0225] Use the logistic regression model to analyze the sentiment polarity of user feedback and calculate the probability P(y=1|x) that the user feedback is positive:

[0226]

[0227] is the Sigmoid activation function, w is the weight vector, x is the feature vector, and b is the bias term;

[0228] Based on the sentiment analysis results of the user feedback information Q2 and combined with the user's behavior data in the system, the user's social media account information M is determined.

[0229] According to a specific implementation of an embodiment of the present invention, the authenticity verification of the user's social media account information M is performed, and based on the authenticity verification result, different types of AI digital human expressions are generated, including:

[0230] Calculate the trust score S of the user’s social media account information M:

[0231]

[0232] S is the trust score, α, β, γ, δ, ϵ are feature weights, A is the age of the user account, C is the number of connections of the user account, P is the number of content published by the user account, F is the number of users followed by the user account, and L is the number of likes of the user account;

[0233] Based on the trust score, the authenticity of the user's social media account information M is determined.

[0234] According to a specific implementation of an embodiment of the present invention, the authenticity verification of the user's social media account information M and the generation of different types of AI digital human expressions based on the authenticity verification result also include:

[0235] Use a multi-layer perceptron model to predict the probability of expression selection:

[0236]

[0237] y is the output vector, which represents the selection probability of each expression. Each element represents the probability of selecting the i-th expression, is the Softmax function, which is used to convert the output vector into a probability distribution. The Softmax function is defined as:

[0238]

[0239] Where z is the input vector and k is the number of output categories;

[0240] is the weight matrix of the second layer, which is used to transform the output of the hidden layer into the final output layer. is the activation function:

[0241]

[0242] is the weight matrix of the first layer, which is used to convert the input feature vector x into the input of the hidden layer. x is the input feature vector, which contains features extracted from user feedback and social media account information. is the bias vector of the first layer, which is used to adjust the input of the hidden layer. is the bias vector of the second layer, which is used to adjust the input of the output layer.

[0243] Corresponding to the above method embodiment, see Figure 3 The embodiment of the present invention further discloses an intelligent data processing device 30 equipped with an AI digital human, comprising:

[0244] Acquisition module 301, embedding an AI digital human in the digital signage, wherein the AI ​​digital human interacts with the user through voice, text or gesture to acquire the user's demand information Q1;

[0245] The generation module 302 generates a multi-dimensional data visualization chart T based on the user's demand information Q1, and selects an adaptive algorithm based on the chart to dynamically adjust the chart layout, style and interaction mode by analyzing data characteristics in real time;

[0246] An analysis module 303 analyzes the user's interaction history Jh using a deep learning model, and generates historical analysis data D1 and future prediction data D2 for the user based on the interaction history;

[0247] The determination module 304 obtains the user's feedback information Q2 regarding the data visualization chart T, the historical analysis data D1, and the future prediction data D2, and determines the user's social media account information M based on the feedback information Q2;

[0248] 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 authenticity verification results.

[0249] See also Figure 4 The embodiment of the present invention further provides an electronic device 60, the electronic device comprising:

[0250] at least one processor; and,

[0251] a memory communicatively connected to the at least one processor; wherein,

[0252] The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the intelligent data processing method equipped with an AI digital human in the aforementioned method embodiment.

[0253] An embodiment of the present invention further provides a non-transitory computer-readable storage medium, which stores computer instructions, and the computer instructions are used to enable the computer to execute the aforementioned method embodiment.

[0254] An embodiment of the present invention also provides a computer program product, which includes a computer program stored on a non-transitory computer-readable storage medium. The computer program includes program instructions. When the program instructions are executed by a computer, the computer executes the intelligent data processing method equipped with an AI digital human in the aforementioned method embodiment.

[0255] Reference below Figure 4, which shows a schematic diagram of the structure of an electronic device 60 suitable for implementing the embodiment of the present disclosure. The electronic device in the embodiment of the present disclosure may include, but is not limited to, mobile terminals such as mobile phones, notebook computers, digital broadcast receivers, PDAs (personal digital assistants), PADs (tablet computers), PMPs (portable multimedia players), vehicle terminals (such as vehicle navigation terminals), etc., and fixed terminals such as digital TVs, desktop computers, etc. Figure 4 The electronic device shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present disclosure.

[0256] like Figure 4 As shown, the electronic device 60 may include a processing device (e.g., a central processing unit, a graphics processing unit, etc.) 601, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 602 or a program loaded from a storage device 608 into a random access memory (RAM) 603. In the RAM 603, various programs and data required for the operation of the electronic device 60 are also stored. The processing device 601, the ROM 602, and the RAM 603 are connected to each other via a bus 604. An input / output (I / O) interface 605 is also connected to the bus 604.

[0257] Typically, the following devices may be connected to the I / O interface 605: an input device 606 including, for example, a touch screen, a touch pad, a keyboard, a mouse, an image sensor, a microphone, a gage, a gyroscope, etc.; an output device 607 including, for example, a liquid crystal display (LCD), a speaker, a vibrator, etc.; a storage device 608 including, for example, a magnetic tape, a hard disk, etc.; and a communication device 609. The communication device 609 may allow the electronic device 60 to communicate with other devices wirelessly or by wire to exchange data. Although Figure 4 The electronic device 60 is shown with various devices, but it should be understood that it is not required to implement or have all the devices shown. More or fewer devices may be implemented or have instead.

[0258] In particular, according to an embodiment of the present disclosure, the process described above with reference to the flowchart can be implemented as a computer software program. For example, an embodiment of the present disclosure includes a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program includes a program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from a network through 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, the above-mentioned functions defined in the method of the embodiment of the present disclosure are executed.

[0259] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any changes or substitutions that can be easily thought of by a person skilled in the art within the technical scope disclosed by the present invention should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.

Claims

1. An intelligent data processing method equipped with AI digital human, characterized in that: include: An AI digital human is embedded in the digital signage, and the AI ​​digital human interacts with the user through voice, text or gesture to obtain the user's demand information Q1; Based on the user's demand information Q1, a multi-dimensional data visualization chart T is generated, and an adaptive algorithm is selected based on the chart to dynamically adjust the chart layout, style and interaction method by analyzing the data characteristics in real time; Analyze the user's interaction history Jh using a deep learning model, and generate historical analysis data D1 and future prediction data D2 for the user based on the interaction history; Obtain user feedback information Q2 regarding the data visualization chart T, the historical analysis data D1, and the 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 different types of AI digital human expressions are generated based on the results of the authenticity verification.

2. The method according to claim 1, characterized in that: The AI ​​digital human is embedded in the digital signage, and the AI ​​digital human interacts with the user through voice, text or gesture to obtain the user's demand information Q1, including: Use the speech recognition API to convert user voice input into text Qt; Use natural language processing technology to parse the user's text input Qr; Capture the user's body movements and map the user's gestures to specific interaction commands Qm; Based on the text Qt, text input Qr and interactive command Qm, the user's demand information Q1 is generated.

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, and an adaptive algorithm is selected based on the chart to dynamically adjust the chart layout, style and interaction mode by analyzing data characteristics in real time, including: Analyze the user's demand information Q1, and determine the data source, time range, data type, and chart type corresponding to the demand information Q1; Get data from the specified data source, and pre-process the acquired data to obtain chart data; The chart data is converted into a multi-dimensional data visualization chart T using a data visualization library.

4. The method according to claim 3, characterized in that The method generates a multi-dimensional data visualization chart T based on the user's demand information Q1, selects an adaptive algorithm based on the chart, and dynamically adjusts the chart layout, style and interaction mode by analyzing data characteristics in real time, and also includes: Analyze data characteristics in real time, including data volume, data distribution characteristics, data change trends, and identify key features of data; Automatically select the corresponding chart type based on data characteristics; Dynamically adjust the layout of the chart based on the amount of data and chart type.

5. The method according to claim 4, characterized in that The method of analyzing the user's interaction history Jh by 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: The state update formula for building a deep learning model is: is the output of the forget gate, It is a Sigmoid activation function with an output range between 0 and 1, which is used to control the degree of information retention. is the weight matrix of the forget gate, is the hidden state at the previous moment, is the input at the current moment, is the bias vector of the forget gate; is the output of the input gate, is the weight matrix of the input gate, is the bias vector of the input gate; is the candidate cell state, which is the new information calculated by the current input and the hidden state of the previous moment. is the hyperbolic tangent activation function, and the output range is between -1 and 1. is the weight matrix of the candidate cell state, is the bias vector of the candidate cell state; is the current state of the memory cell, is an element-by-element multiplication operation, is the state of the memory cell at the previous moment; is the output of the output gate, is the weight matrix of the output gate, is the bias vector of the output gate; is the hidden state at the current moment, which is used to pass to the next time step. Memory cell state The value after the hyperbolic tangent activation function; Use an additive model to decompose the time series into trend, seasonal, and random components: is the original time series data, is the trend component, which indicates the long-term trend of the data. is the seasonal component, indicating the periodic fluctuation of the data. is the random component, which represents the random noise or unpredictable part of the data.

6. The method according to claim 5, characterized in that The obtaining of user feedback information Q2 regarding the data visualization chart T, the historical analysis data D1, and the future prediction data D2, and determining the user's social media account information M based on the feedback information Q2, includes: Use the logistic regression model to analyze the sentiment polarity of user feedback and calculate the probability P(y=1|x) that the user feedback is positive: 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 the user feedback information Q2 and combined with the user's behavior data in the system, the user's social media account information M is determined.

7. The method according to claim 6, characterized in that The authenticity verification of the user's social media account information M is performed, and based on the authenticity verification result, different types of AI digital human expressions are generated, including: Calculate the trust score S of the user’s social media account information M: S is the trust score, α, β, γ, δ, ϵ are feature weights, A is the age of the user account, C is the number of connections of the user account, P is the number of content published by the user account, F is the number of users followed by the user account, and L is the number of likes of the user account; Based on the trust score, the authenticity of the user's social media account information M is determined.

8. The method according to claim 7, characterized in that The authenticity verification of the user's social media account information M and the generation of different types of AI digital human expressions based on the authenticity verification result also include: Use a multi-layer perceptron model to predict the probability of expression selection: y is the output vector, which represents the selection probability of each expression. Each element represents the probability of selecting the i-th expression, is the Softmax function, which is used to convert the output vector into a probability distribution. The Softmax function is defined as: Where z is the input vector and k is the number of output categories; is the weight matrix of the second layer, which is used to transform the output of the hidden layer into the final output layer. is the activation function: is the weight matrix of the first layer, which is used to convert the input feature vector x into the input of the hidden layer. x is the input feature vector, which contains features extracted from user feedback and social media account information. is the bias vector of the first layer, which is used to adjust the input of the hidden layer. is the bias vector of the second layer, which is used to adjust the input of the output layer.

9. An intelligent data processing device equipped with an AI digital human, characterized in that: include: An acquisition module embeds an AI digital human in the digital signage, and the AI ​​digital human interacts with the user through voice, text or gesture to acquire the user's demand information Q1; The generation module generates a multi-dimensional data visualization chart T based on the user's demand information Q1, and selects an adaptive algorithm based on the chart to dynamically adjust the chart layout, style and interaction method by analyzing data characteristics in real time; An analysis module, which 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; A determination module obtains user feedback information Q2 regarding the data visualization chart T, the historical analysis data D1, and the future prediction data D2, and determines the user's social media account information M based on the feedback information Q2; 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 results of the authenticity verification.

10. An electronic device, characterized in that: The electronic device comprises: 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, and the instructions are executed by the at least one processor so that the at least one processor can execute the intelligent data processing method equipped with an AI digital human as described in any one of claims 1 to 8.

Images