User behavior prediction method and system based on AI and big data, and storage medium

Abstract

The embodiment of the invention discloses a user behavior prediction method and system based on AI and big data and a medium. The method comprises the following steps: defining general indexes; obtaining interaction behavior data fed back by the user, and performing analysis based on the general indexes by adopting a big data technology to obtain target index data; and inputting the target index data into the AI large language model for user behavior prediction to obtain an output result. Through a set of universal index definition, big data and AI model capability can be conveniently connected in series at low cost, user behaviors are modeled, corresponding behavior data are obtained, next interactive operation of the user is predicted, meanwhile, the interactive operation is fed back to the system, and the user experience is improved. And an operator can be helped to know the current state of the user more clearly and intuitively. According to the method and the system, an enterprise system can obtain more real-time, more personalized and more accurate user behavior analysis and prediction capability by low-cost investment in the aspect of whatsp marketing.

Description

Technical Field

[0001] This invention relates to the field of data statistical analysis technology, specifically to a user behavior prediction method, system, and storage medium based on AI and big data. Background Technology

[0002] With the widespread use of social media, businesses are increasingly relying on instant messaging platforms such as WhatsApp for customer communication and marketing campaigns. These campaigns generate a large amount of user behavior data, and analyzing this data has significant commercial value.

[0003] The existing user behavior analysis methods and their shortcomings are as follows:

[0004] 1. User behavior analysis based on software tools; software tools often lack real-time and personalization, making it difficult to effectively predict users' specific needs and behavioral tendencies, or requiring a large number of marketing and operations personnel with professional skills to connect with them, resulting in high costs.

[0005] 2. User behavior analysis based on big data technology; Although big data technology can process massive amounts of information, it still has shortcomings in real-time performance and accuracy, especially in the field of instant messaging, and it is difficult to integrate with traditional applications.

[0006] 3. User behavior analysis based on large language models; the AI ​​capabilities provided by large language models often lack applications for specific domains and needs, and their value needs to be enhanced through more convenient focusing. Summary of the Invention

[0007] In view of the technical deficiencies mentioned in the background art, the purpose of this invention is to provide a user behavior prediction method, system and storage medium based on AI and big data, aiming to improve the real-time, personalized and accurate analysis of user behavior in instant messaging or marketing message platforms.

[0008] To achieve the above objectives, in a first aspect, embodiments of the present invention provide a user behavior prediction method based on AI and big data, comprising:

[0009] Define general metrics; based on the definition of these general metrics, message delivery, big data output, and AI output can be linked together;

[0010] Obtain user feedback interaction behavior data, use big data technology and analyze the interaction behavior data based on the general indicators to obtain target indicator data;

[0011] The target indicator data is input into an AI large language model to predict user behavior and obtain the output results. The output results include personalized content, the type of message to be sent, and the possible user actions corresponding to the message content.

[0012] As a specific implementation of this application, the target indicator data is obtained as follows:

[0013] Establish a database based on the defined general indicators;

[0014] A big data processing framework is deployed to clearly analyze the historical data and output target indicator data.

[0015] As a preferred implementation of this application, after obtaining the output result, the method further includes:

[0016] The output results are formatted and then entered into the candidate pool; the candidate pool can display the prediction effects of different message types and contents for operators to select.

[0017] Generate target messages of the corresponding type based on the operators' selections;

[0018] The target message is sent to the user via the message delivery interface.

[0019] As a preferred implementation of this application, after the target message is sent to the user through the message delivery interface, the method further includes:

[0020] Record the user's response behavior and timeliness to the target message, and feed the response behavior and timeliness back to the big data processing framework and AI big prediction model.

[0021] Secondly, embodiments of the present invention also provide a user behavior prediction system based on AI and big data, including a processor, an input device, an output device, and a memory, wherein the processor, input device, output device, and memory are interconnected, wherein the memory is used to store a computer program, the computer program includes program instructions, and the processor is configured to call the program instructions to execute the method described in the first aspect.

[0022] Thirdly, embodiments of the present invention also provide a computer-readable storage medium storing a computer program, the computer program including program instructions, which, when executed by a processor, cause the processor to perform the method described in the first aspect.

[0023] This invention also provides another user behavior prediction system based on AI and big data, including:

[0024] The metrics definition module is used to define data metrics for individual users and the types of behaviors that interact with users;

[0025] The database construction and update module is used to build a database based on the data indicators, clean and analyze historical data using big data technology, receive the latest user behavior data in real time, and dynamically update the indicator data in the database.

[0026] The user behavior prediction module is used to input indicator data from the database into the AI ​​large language model to predict user behavior and obtain output results. The output results include personalized content, the type of message to be sent, and the possible user actions corresponding to the message content.

[0027] As a preferred implementation of this application, the user behavior prediction system further includes:

[0028] The prediction results display module is used to display the output results to the operations personnel through the decision support system;

[0029] The operations decision support module is used to provide decision-making suggestions and optimization solutions to operations personnel based on the output results and business indicators.

[0030] The interface adaptation module is used to adapt user behavior types and message interfaces, and convert the output results into a message to be sent to the user.

[0031] The feedback data collection module is used to collect feedback data and feed it back to the AI ​​large language model; the feedback data includes the user's response behavior and timeliness to the sent message;

[0032] The model optimization and adjustment module is used to retrain and adjust the parameters of the AI ​​large language model based on the feedback data using machine learning algorithms.

[0033] Compared with existing technologies, the embodiments of the present invention can conveniently and cost-effectively connect big data and AI model capabilities through a set of generalized indicator definitions, model user behavior, obtain corresponding behavioral data, predict the user's next interaction, and at the same time feed back the interaction to the system to achieve a positive cycle capability, helping operators to understand the user's current state more clearly and intuitively.

[0034] Furthermore, based on behavioral prediction, targeted marketing strategies and recommended content are generated to help business operators make more effective business decisions, thereby improving operational efficiency and reducing operating costs.

[0035] On the other hand, by defining behavior types in a standardized and structured way, the AI ​​output can be easily applied to WhatsApp, forming a combination with and empowering WhatsApp, without having to build a complicated message delivery interface, thus improving the real-time performance of message delivery. Attached Figure Description

[0036] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the accompanying drawings used in the description of the specific embodiments or the prior art will be briefly introduced below.

[0037] Figure 1 This is a flowchart of the user behavior prediction method based on AI and big data provided in the embodiments of the present invention;

[0038] Figure 2 yes Figure 1 Another flowchart of the method shown;

[0039] Figure 3 This is a structural diagram of the user behavior prediction system based on AI and big data provided in the first embodiment of the present invention;

[0040] Figure 4 This is a structural diagram of the user behavior prediction system based on AI and big data provided in the second embodiment of the present invention. Detailed Implementation

[0041] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0042] It should be understood that, when used in this specification and the appended claims, the terms "comprising" and "including" indicate the presence of the described features, integrals, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components and / or collections thereof.

[0043] This invention aims to improve the real-time performance, personalization, and accuracy of user behavior analysis in instant messaging or marketing messaging platforms. Leveraging big data and large language model capabilities, it integrates data with AI capabilities, limiting the application scenarios for AI through preset contextual ranges, thereby enhancing the personalization and relevance of message content. Furthermore, by utilizing key data provided by big data, it further supplies AI with high-value keywords, compensating for the shortcomings in real-time user behavior analysis and improving the accuracy of AI in predicting the behavior of specific individual users.

[0044] In summary, this invention enables enterprise systems to achieve more real-time, personalized, and accurate user behavior analysis and prediction capabilities with a lower cost for WhatsApp marketing applications.

[0045] Please refer to Figure 1 and Figure 2 The user behavior prediction method based on AI and big data provided in this embodiment of the invention may include the following steps:

[0046] S1 defines general metrics.

[0047] In this embodiment, the core capabilities of the system are defined and broken down from different dimensions. These definitions are used to redirect data flow and connect the three capabilities: WhatsApp message delivery, key big data output, and AI capability output, thereby achieving the desired effect. Specifically, the relevant core definitions are as follows:

[0048] (1) Define data metrics for individual users: These metrics are mainly a "summary" of users' historical behavior data, and the objects of the metrics are individual users; they are the output of big data capabilities and also the input data of AI capabilities. Through the definition of these metrics, user behavior, big data, and AI capabilities are linked together. Currently defined metrics include: the timeliness of the user's previous message response, the relevance of the content of the user's previous message response, the timeliness of the user's historical responses, the timeliness of all users' historical responses, the user's historical purchase records and service evaluations, and the historical service evaluations of all users.

[0049] (2) Define the types of user interaction behaviors: These behaviors mainly refer to the types of behaviors that WhatsApp can provide. By establishing this behavior type dimension, different behavior types can be vertically aggregated for analysis, so that different behavior types have different meanings. This behavior type is also a dimension of AI prediction output. Currently, the definition includes: button clicks, rich text content, category products, and complaints and suggestions.

[0050] (3) Define the predicted output: The predicted output is the final output of the entire system. It is mainly aimed at the current user, the current scenario, and the user's possible operations. It includes: personalized content, the type of message to be sent, and the user's possible operations corresponding to the message content.

[0051] S2, obtain user feedback interaction behavior data, use big data technology and analyze the interaction behavior data based on the general indicators to obtain target indicator data.

[0052] In practice, a corresponding database is established based on the definition of the indicators, historical data is cleaned and analyzed through big data analytics, and real-time user behavior data is received to update and output the corresponding indicator data.

[0053] For example, the first step is to model the metric data and build an efficient columnar database using ClickHouse. Tables are then created to process user behavior data, supporting rapid querying and analysis. Next, big data processing frameworks such as Hadoop and Flink are deployed to clean and analyze historical data and update user behavior data in real time, ensuring the accuracy and timeliness of the metric data.

[0054] S3 inputs the target indicator data into the AI ​​large language model to predict user behavior and obtain the output results.

[0055] In practice, the user's metric data results are used as user features and submitted to the AI ​​model. The AI ​​then outputs multiple results, including the current behavior type and corresponding content, as well as the behavioral actions the user will take based on that behavior type and content.

[0056] For example, large language models such as ChatGPT and Kimi can be used to directly leverage their powerful content understanding and generation capabilities to predict behavior types and possible actions based on user characteristics.

[0057] S4 formats the output and then enters the candidate pool.

[0058] In practice, the structural data obtained in the previous step is used as a candidate pool for the current operations staff to understand and select. The decision support system is used to display the predictive effects of different message types and content to assist the operations staff in making decisions.

[0059] The candidate pool can be a set of structured JSON structures, which are then parsed to present reference options to operations personnel.

[0060] S5 generates target messages of the corresponding type based on the operator's selection.

[0061] S6 sends the target message to the user through the message delivery interface.

[0062] In practice, a WhatsApp message delivery interface is developed, which uses the HTTP protocol to connect with WhatsApp's API. The interface parameters receive the data body of the structure from the previous step. Once the operations staff makes a selection, the interface will generate a message of the corresponding type and deliver the message to the user through the WhatsApp API interface.

[0063] S7 records user response behavior and timeliness to target messages, and feeds the response behavior and timeliness back to the big data processing framework and AI big prediction model.

[0064] In practice, the system records the user's response to the sent messages and the timeliness of the response, and then re-flows these responses to the big data processing frameworks Hadoop and Flink. At the same time, these data are fed back to the LLM model to verify and adjust the accuracy of the predictions.

[0065] Compared with existing technologies, the embodiments of the present invention can conveniently and cost-effectively connect big data and AI model capabilities through a set of generalized indicator definitions, model user behavior, obtain corresponding behavioral data, predict the user's next interaction, and at the same time feed back the interaction to the system to achieve a positive cycle capability, helping operators to understand the user's current state more clearly and intuitively.

[0066] Furthermore, based on behavioral prediction, targeted marketing strategies and recommended content are generated to help business operators make more effective business decisions, thereby improving operational efficiency and reducing operating costs.

[0067] On the other hand, by defining behavior types in a standardized and structured way, the AI ​​output can be easily applied to WhatsApp, forming a combination with and empowering WhatsApp, without having to build a complicated message delivery interface, thus improving the real-time performance of message delivery.

[0068] Based on the same inventive concept, embodiments of the present invention provide a user behavior prediction system based on AI and big data, such as... Figure 3 As shown, the system includes an indicator definition module, a database construction and update module, a user behavior prediction module, a prediction result display module, an operational decision support module, an interface adaptation module, a feedback data collection module, and a model optimization and adjustment module. Each module will be described in detail below.

[0069] 1. Indicator Definition Module

[0070] This module is responsible for precisely defining data metrics for individual users, covering dimensions such as the timeliness of the user's previous message response, content relevance, historical response timeliness, purchase history, and service evaluations. Through these metrics, the system can comprehensively summarize users' historical behavioral data, laying the foundation for subsequent big data analysis and AI capability input. The generalized design of the metric definitions allows the system to flexibly adapt to different business scenarios and user groups, achieving accurate capture and analysis of user behavior.

[0071] Furthermore, this module is responsible for standardizing the definition of user interaction behaviors, including button clicks, rich text content, product categories, complaints, and suggestions. By establishing a behavior type dimension, the system can vertically aggregate and analyze different behavior types to uncover the user intent and business value behind them. This standardized definition not only provides an important dimension for AI prediction output but also lays the foundation for subsequent message delivery and user interaction.

[0072] 2. Database Construction and Update Module

[0073] Based on the defined metrics, this module establishes a corresponding database structure to store and manage user data. Utilizing big data analytics capabilities, it cleans and analyzes historical data to ensure accuracy and completeness. Simultaneously, it receives the latest user behavior data in real time, dynamically updating the metric data in the database to provide real-time and reliable data support for the system, ensuring the timeliness of predictions and decision-making.

[0074] 3. User Behavior Prediction Module

[0075] This module is responsible for inputting indicator data from the database into the AI ​​large language model to predict user behavior and obtain output results. These output results include, but are not limited to, personalized content, the type of message to be sent, and the user's possible actions corresponding to the message content.

[0076] 4. Interface adapter module

[0077] This module is responsible for adapting behavior types to the WhatsApp messaging interface, efficiently converting AI-generated predictions into WhatsApp messages. By simplifying the message delivery interface construction process, the system can quickly deliver predictions to users via WhatsApp, improving the real-time nature and accuracy of message delivery. This empowerment approach expands WhatsApp's application value in enterprise marketing and user interaction, enabling businesses to more easily utilize WhatsApp for their business activities.

[0078] 5. Prediction Result Display Module

[0079] This module, through a decision support system, presents AI prediction results to operations personnel in an intuitive and easy-to-understand manner. The system provides comparisons of prediction effectiveness for different message types and content, including potential user behaviors and response times. This visualization enables operations personnel to quickly understand the prediction results, assess the potential effects of different strategies, and thus make more accurate and efficient operational decisions.

[0080] 6. Operational Decision Support Module

[0081] By combining forecast results with business objectives, this module provides operations personnel with decision-making suggestions and optimization solutions. Based on user behavior data and business rules, the system can intelligently recommend the best message type, content, and delivery timing, helping operations personnel develop personalized marketing strategies and user interaction plans. Through this decision support, businesses can improve operational efficiency, reduce operating costs, and achieve continuous business optimization and growth.

[0082] 7. Feedback Data Collection Module

[0083] This module is responsible for collecting user response behavior and timeliness data to sent messages. When a user receives a message and takes an action, the system records this data in real time, including message open rate, click rate, and reply content. This feedback data is fed into the LLM model to verify and adjust the accuracy of predictions, enabling the system to continuously optimize the prediction model and improve prediction accuracy.

[0084] 8. Model Optimization and Adjustment Module

[0085] Based on the collected feedback data, this module optimizes and adjusts the LLM model. The system uses machine learning algorithms to retrain the model and adjust its parameters to adapt to changes in user behavior and updated business needs. Through this positive feedback loop of prediction, the system can achieve self-iteration and continuous improvement, providing users with more accurate and personalized services, while helping operations personnel better understand user status and make more effective business decisions.

[0086] It should be noted that the specific workflow of this embodiment is described in the foregoing method embodiment section, and will not be repeated here.

[0087] Optionally, another embodiment of the present invention also provides a user behavior prediction system based on AI and big data. For example... Figure 4As shown, the user behavior prediction system may include one or more processors 101, one or more input devices 102, one or more output devices 103, and a memory 104. The processors 101, input devices 102, output devices 103, and memory 104 are interconnected via a bus 105. The memory 104 stores a computer program, which includes program instructions. The processor 101 is configured to invoke the program instructions to execute the method described in the above-described method embodiment.

[0088] It should be understood that, in this embodiment of the invention, the processor 101 may be a central processing unit (CPU), but it may also be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or any conventional processor.

[0089] Input device 102 may include a keyboard, etc., and output device 103 may include a display (LCD, etc.), a speaker, etc.

[0090] The memory 104 may include read-only memory and random access memory, and provides instructions and data to the processor 101. A portion of the memory 104 may also include non-volatile random access memory. For example, the memory 104 may also store device type information.

[0091] In specific implementations, the processor 101, input device 102, and output device 103 described in the embodiments of the present invention can execute the implementation methods described in the embodiments of the user behavior prediction method based on AI and big data provided in the embodiments of the present invention, which will not be repeated here.

[0092] Accordingly, embodiments of the present invention provide a computer-readable storage medium storing a computer program, the computer program including program instructions, which, when executed by a processor, implement the above-described user behavior prediction method based on AI and big data.

[0093] The computer-readable storage medium can be an internal storage unit of the system described in any of the foregoing embodiments, such as the system's hard disk or memory. The computer-readable storage medium can also be an external storage device of the system, such as a plug-in hard disk, Smart Media Card (SMC), Secure Digital (SD) card, or Flash Card. Furthermore, the computer-readable storage medium can include both internal storage units and external storage devices. The computer-readable storage medium is used to store the computer program and other programs and data required by the system. The computer-readable storage medium can also be used to temporarily store data that has been output or will be output.

[0094] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components and steps of the various examples have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0095] In the several embodiments provided in this application, it should be understood that the disclosed apparatus and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative. For instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. In addition, the mutual coupling or direct coupling or communication connection shown or discussed may be indirect coupling or communication connection through some interfaces, devices or units, or may be electrical, mechanical or other forms of connection.

[0096] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of the embodiments of the present invention, depending on actual needs.

[0097] Furthermore, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0098] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0099] 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 person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in the present invention, and these modifications or substitutions should all be covered 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 user behavior prediction method based on AI and big data, characterized in that, include: Define general metrics; based on the definition of these general metrics, message delivery, big data output, and AI output can be linked together; Obtain user feedback interaction behavior data, use big data technology and analyze the interaction behavior data based on the general indicators to obtain target indicator data; The target indicator data is input into an AI large language model to predict user behavior and obtain the output results. The output results include personalized content, the type of message to be sent, and the possible user actions corresponding to the message content.

2. The user behavior prediction method as described in claim 1, characterized in that, The general metrics include individual user metrics, user interaction behavior type metrics, and AI prediction output metrics; the individual user metrics include the timeliness of the user's previous message response, the relevance of the content of the user's previous message response, the user's historical response timeliness, the historical response timeliness of all users, the user's historical purchase records and service evaluations, and the historical service evaluations of all users.

3. The user behavior prediction method as described in claim 1, characterized in that, The interactive behavior data includes historical data and real-time behavior data; the target indicator data is specifically obtained as follows: Establish a database based on the defined general indicators; A big data processing framework is deployed to clearly analyze the historical data and output target indicator data.

4. The user behavior prediction method as described in claim 1, characterized in that, After obtaining the output result, the method further includes: The output results are formatted and then entered into the candidate pool; the candidate pool can display the prediction effects of different message types and contents for operators to select. Generate target messages of the corresponding type based on the operators' selections; The target message is sent to the user via the message delivery interface.

5. The user behavior prediction method as described in claim 4, characterized in that, After the target message is sent to the user via the message delivery interface, the method further includes: Record the user's response behavior and timeliness to the target message, and feed the response behavior and timeliness back to the big data processing framework and AI big prediction model.

6. A user behavior prediction system based on AI and big data, characterized in that, The system includes a processor, an input device, an output device, and a memory, which are interconnected. The memory is used to store a computer program, which includes program instructions. The processor is configured to invoke the program instructions to perform the method as described in any one of claims 1-5.

7. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program, the computer program including program instructions that, when executed by a processor, cause the processor to perform the method as described in any one of claims 1-5.

8. A user behavior prediction system based on AI and big data, characterized in that, include: The metrics definition module is used to define data metrics for individual users and the types of behaviors that interact with users; The database construction and update module is used to build a database based on the data indicators, clean and analyze historical data using big data technology, receive the latest user behavior data in real time, and dynamically update the indicator data in the database. The user behavior prediction module is used to input indicator data from the database into the AI ​​large language model to predict user behavior and obtain output results. The output results include personalized content, the type of message to be sent, and the possible user actions corresponding to the message content.

9. The user behavior prediction system as described in claim 8, characterized in that, Also includes: The prediction results display module is used to display the output results to the operations personnel through the decision support system; The operations decision support module is used to provide operations personnel with decision-making suggestions and optimization solutions based on the output results and business indicators. The interface adaptation module is used to adapt user behavior types and message interfaces, and convert the output results into a message to be sent to the user.

10. The user behavior prediction system as described in claim 9, characterized in that, Also includes: The feedback data collection module is used to collect feedback data and feed it back to the AI ​​large language model; the feedback data includes the user's response behavior and timeliness to the sent message; The model optimization and adjustment module is used to retrain and adjust the parameters of the AI ​​large language model based on the feedback data using machine learning algorithms.

Images