Method and system for analyzing user consumption behaviors in different scenarios within a commercial space

Abstract

The present invention relates to the technical field of spatial behavior analysis. The present invention discloses a method and system for analyzing user consumption behaviors in different scenarios within a commercial space, including obtaining user behavior data through wireless positioning devices and distributed environmental sensors; performing spatio-temporal alignment on multi-source heterogeneous data, and constructing a three-dimensional trajectory model by integrating commodity granularity data of a transaction system; identifying the residence characteristics and movement patterns of users based on the three-dimensional trajectory model, and generating a dynamic heat map by associating commodity distributions; adjusting the guidance strategy and environmental parameters according to the density gradient of the heat map and the user attribute level; integrating payment feature data and user behavior data to update user tags, and generating a cross-region association strategy. This method provides a comprehensive, flexible and efficient method for analyzing user consumption behaviors in a commercial space, solves the problem of insufficient understanding of user dynamic behaviors in the prior art, and thus promotes the process of commercial intelligence.

Description

Technical Field

[0001] The present invention relates to the technical field of spatial behavior analysis, and particularly to a method and system for analyzing user consumption behaviors in different scenarios within a commercial space. Background Art

[0002] With the continuous development of commercial spaces and the increasing complexity of user behaviors, enterprises need more accurate and efficient methods for analyzing user consumption behaviors in order to formulate more scientific marketing strategies and enhance the customer experience. In recent years, with the rapid progress of Internet of Things (IoT), big data, and cloud computing technologies, significant achievements have been made in the field of user behavior analysis. User behavior data capture systems within commercial spaces generally adopt technical means such as wireless sensor networks, mobile device positioning, and visual monitoring to collect user consumption behavior data in real time and analyze user behavior patterns and preferences based on data analysis models.

[0003] In particular, cloud computing-based analysis systems have made it possible to extract, process, and store information from user behavior data more efficiently. At the same time, by using advanced algorithms such as machine learning and deep learning to build user models, business analysts can identify user residence characteristics, movement patterns, and generate visual dynamic heatmaps by analyzing this data. However, there are still some deficiencies in the existing technologies, which provide insufficient exact basis for business decisions.

[0004] The existing technologies have limitations in the diversity and heterogeneity of data sources. Data sources within commercial spaces are often scattered, involving various devices such as positioning devices, environmental sensors, and transaction systems, lacking a unified standard and an effective spatio-temporal alignment mechanism, resulting in difficulties in data integration and affecting the accuracy of analysis results. In addition, existing systems often rely on static models and are difficult to adapt to the rapid changes in user behaviors in real time, presenting challenges in maintaining timeliness and accuracy.

[0005] Most existing technologies still adopt traditional pattern recognition methods, failing to make full use of adaptive learning mechanisms and being less sensitive to the continuous changes in user behaviors, resulting in a lag in model updates, which makes user portraits and behavior analysis often unable to reflect the real needs of users. In addition, the process of generating heatmaps lacks the integration of user feedback, making it difficult to mobilize the participation of users and affecting the effectiveness of decision-making. Summary of the Invention

[0006] In view of the above existing problems, the present invention is proposed.

[0007] Therefore, the present invention provides a method for analyzing user consumption behaviors in different scenarios within a commercial space, which can solve problems such as data silos, insufficient dynamic adaptability, and lack of user feedback mechanism in the prior art. By comprehensively integrating multi-source data, enhancing real-time response capabilities, and realizing intelligent management within the commercial space.

[0008] To solve the above technical problems, the present invention provides the following technical solutions. A method for analyzing user consumption behaviors in different scenarios within a commercial space includes: obtaining user behavior data through a wireless positioning device and a distributed environmental sensor; performing spatio-temporal alignment on multi-source heterogeneous data, and fusing commodity granularity data of a transaction system to construct a three-dimensional trajectory model; identifying the residence characteristics and movement patterns of users based on the three-dimensional trajectory model, and associating commodity distributions to generate a dynamic heat map; adjusting the guidance strategy and environmental parameters according to the density gradient of the heat map and the user attribute level; integrating payment feature data and user behavior data to update user tags and generate a cross-regional association strategy.

[0009] As a preferred solution of the method for analyzing user consumption behaviors in different scenarios within a commercial space according to the present invention, wherein: the wireless positioning device includes positioning the movement trajectory of a user within the commercial space by receiving signals from a mobile device carried by the user;

[0010] The distributed environmental sensor includes an RFID reader and a pressure sensing device deployed in a functional area, and is used for recording the attribute information and interaction frequency of an item being operated.

[0011] As a preferred solution of the method for analyzing user consumption behaviors in different scenarios within a commercial space according to the present invention, wherein: the spatio-temporal alignment of the multi-source heterogeneous data includes constructing a central API gateway, receiving and routing data requests from the wireless positioning device and the distributed environmental sensor, using a mapping table to convert different data structures of each device into a unified format and transmitting them to a data processing platform;

[0012] Introducing an adaptive heterogeneous data parser to automatically identify and process data from different sources according to data characteristics, and removing noise data and redundant information;

[0013] Assigning a unique timestamp and geographic location identifier to each piece of data, establishing a connection through the user's activity trajectory and dynamic environmental data stored in the database, and calculating the similarity of the timestamp and location identifier in space and time to dynamically associate the data.

[0014] As a preferred solution of the method for analyzing user consumption behaviors in different scenarios within a commercial space according to the present invention, wherein: the construction of the three-dimensional trajectory model includes the construction of the three-dimensional trajectory model using collaborative processing based on cloud computing technology to construct a multi-level cloud computing architecture;

[0015] Edge computing nodes are set in the user access layer. Edge devices are set to collect and preliminarily process user data. Inside each edge node, a lightweight data processing module is developed to perform real-time analysis on user behavior data. The cleaned data is transmitted to the central cloud node through a secure encrypted channel for further analysis;

[0016] Create a RESTful API to enable communication between edge nodes and the cloud, for two-way data transfer and command issuance.

[0017] As a preferred solution of the method for analyzing user consumption behavior in different scenarios within a commercial space according to the present invention, wherein: the identification of the residence characteristics and movement patterns of users includes extracting behavioral characteristics such as residence time, frequency, and movement speed from user behavior data, and constructing a user activity model;

[0018] Apply the behavior cloning method in reinforcement learning to simulate the typical movement patterns of users in the commercial space. Through a real-time data feedback mechanism, continuously monitor changes in user behavior data, adopt stream data analysis technology, capture user behavior trends in real time, and adjust the parameters of the user activity model;

[0019] Through the incremental learning method, without retraining the user activity model, only adjust the parameters based on new data, so that the user activity model always reflects the latest user behavior characteristics.

[0020] As a preferred solution of the method for analyzing user consumption behavior in different scenarios within a commercial space according to the present invention, wherein: the generation of a dynamic heat map includes converting user behavior data into a heat map format and distinguishing the heat levels in different time periods;

[0021] Capture the non-linear relationships in user behavior data, identify the behavioral characteristics of user activities, and map the behavioral characteristics to the heat map space through quadratic or cubic polynomial function mapping. Divide the user hot spots into high, medium, and low heat zones, with each heat zone corresponding to a different color. Prioritize planning a detour path for high-consumption-level users to a low-density waiting area;

[0022] Integrate a simple user feedback survey function, combine user feedback with heat map data, use a simple weighted model to calculate the impact of user satisfaction on heat data, and dynamically update the heat map according to the results.

[0023] As a preferred solution of the method for analyzing user consumption behavior in different scenarios within a commercial space according to the present invention, wherein: the generation of cross-regional association strategies includes generating a behavior sequence containing timestamps, coordinates, and residence durations through the collected user movement trajectories; align the payment moment with the residence events in the user trajectory to bind the same user entity;

[0024] Build a multi-dimensional fusion index to associate user behaviors and consumption characteristics into a unified data object. When a user stays in any functional area for a long time without consumption but generates an average order value higher than the mean in the remaining areas, it is marked as a potential associated behavior.

[0025] Construct a multi-level label system, including a basic attribute layer, a behavior derivative layer, and a deep interaction type, and use an incremental learning model to dynamically correct label weights: when the behavior of labeled users deviates from the historical pattern, reduce the confidence level of the current label's association with the category.

[0026] When a user leaves the current area, match the optimal associated area according to the label: for "price-sensitive" users, preferentially push the navigation of adjacent areas with promotional activities; for "experience-oriented" users, recommend the path of the display area equipped with interactive devices.

[0027] When a group of labeled users gather in any functional area, automatically trigger a collaborative strategy to adjust the commodity display density in adjacent areas and highlight the associated categories.

[0028] Deploy a module for behavior feedback analysis, use the changes in user behavior after the execution of the strategy as an optimization signal, and automatically downgrade the priority of the current strategy when any associated strategy fails to achieve the expected goal.

[0029] As a preferred solution of a user consumption behavior analysis system for different scenarios in a commercial space according to the present invention, it includes: a data collection module, a data processing module, a user behavior analysis module, a dynamic heat map generation module, and a strategy generation and execution module;

[0030] The data collection module, through a wireless positioning device and distributed environmental sensors, real-time obtains the behavior data of users, is responsible for receiving the device signals carried by users, environmental sensor data, and data from RFID readers, and records the movement trajectories, interaction frequencies, and relevant attribute information of users in the commercial space;

[0031] The data processing module performs spatio-temporal alignment on the collected multi-source heterogeneous data, and performs data cleaning and conversion to construct a structured data set for analysis.

[0032] The user behavior analysis module, based on the cleaned user behavior data, extracts the residence characteristics and movement patterns of users, and optimizes the user activity model using an adaptive learning mechanism.

[0033] The dynamic heat map generation module converts the processed user behavior data into a heat map format, generates a user activity heat map for different time periods, and displays the areas concerned by users for merchants.

[0034] The strategy generation and execution module generates targeted business strategies based on the analysis results, including cross-regional association strategies, and adjusts product display and user guidance in real time.

[0035] A computer device includes a memory and a processor. The memory stores a computer program, and when the processor executes the computer program, it implements the steps of a method for analyzing user consumption behaviors in different scenarios within a commercial space.

[0036] A computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, it implements the steps of a method for analyzing user consumption behaviors in different scenarios within a commercial space.

[0037] Advantages of the present invention: Through comprehensive analysis of multi-source data within a commercial space, the real-time understanding and response capabilities of user consumption behaviors are improved. By integrating user behavior data from wireless positioning technology and distributed environmental sensors, the present invention constructs a new user activity model and three-dimensional trajectory model, thereby being able to more accurately identify the residence characteristics and movement patterns of users.

[0038] It realizes comprehensive capture and in-depth analysis of user behaviors. Through the generation of dynamic heat maps, merchants can clearly identify the distribution of products preferred by users and consumption trends, thereby optimizing product layouts and guidance strategies to enhance the user experience. At the same time, a user feedback survey function and a dynamic update mechanism are introduced, enabling the system to adjust behavior analysis strategies according to real-time data and user satisfaction, integrating the subjective experience of users and objective data, and providing more targeted decision-making basis for merchants. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for description in the embodiments will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present invention, and those of ordinary skill in the art can obtain other drawings without creative efforts based on these drawings.

[0040] Figure 1 It is a schematic flowchart of a method for analyzing user consumption behaviors in different scenarios within a commercial space provided by an embodiment of the present invention.

[0041] Figure 2 It is a schematic diagram of the working modules of a system for analyzing user consumption behaviors in different scenarios within a commercial space provided by an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0042] To make the above objects, features, and advantages of the present invention more apparent and understandable, the following provides a detailed description of the specific embodiments of the present invention in conjunction with the accompanying drawings of the specification. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the scope of protection of the present invention.

[0043] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those described herein. Those skilled in the art can make similar generalizations without departing from the spirit of the present invention. Therefore, the present invention is not limited by the specific embodiments disclosed below.

[0044] Secondly, the so-called "one embodiment" or "embodiment" herein refers to specific features, structures, or characteristics that may be included in at least one implementation manner of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it an embodiment that is separate or selectively exclusive of other embodiments.

[0045] The present invention is described in detail in conjunction with schematic diagrams. When detailing the embodiments of the present invention, for ease of explanation, the cross-sectional views showing the device structure are enlarged locally out of the general scale, and the schematic diagrams are only examples and should not limit the scope of protection of the present invention herein. In addition, in actual production, three-dimensional spatial dimensions of length, width, and depth should be included.

[0046] At the same time, in the description of the present invention, it should be noted that the orientation or positional relationship indicated by terms such as "upper, lower, inner, and outer" is based on the orientation or positional relationship shown in the drawings, and is only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and thus cannot be construed as a limitation on the present invention. In addition, the terms "first, second, or third" are only used for descriptive purposes and cannot be construed as indicating or implying relative importance.

[0047] Unless otherwise clearly defined and limited in the present invention, the terms "installation, connection, and coupling" should be understood in a broad sense. For example, it may be a fixed connection, a detachable connection, or an integral connection; it may also be a mechanical connection, an electrical connection, or a direct connection, or may be indirectly connected through an intermediate medium, or may be the communication inside two elements. For those of ordinary skill in the art, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0048] Example 1, referring to Figure 1 , which is the first embodiment of the present invention. This embodiment provides a method for analyzing user consumption behaviors in different scenarios within a commercial space, including:

[0049] S1: Obtain user behavior data through a wireless positioning device and distributed environmental sensors.

[0050] Furthermore, the wireless positioning device includes: locating the movement trajectory of a user within a commercial space by receiving signals from a mobile device carried by the user; establishing Bluetooth and Wi-Fi network base stations, and the device records the environmental signal strength;

[0051] After the user enters the commercial space, the positioning device pairs with the user's mobile phone via Bluetooth and simultaneously obtains the Wi-Fi signal strength. Through multi-path propagation and signal backpropagation technologies, the precise location of the Bluetooth signal is fused with the Wi-Fi information to calculate the user's real-time location.

[0052] The distributed environmental sensors include RFID readers and pressure sensing devices deployed within functional areas for recording the attribute information and interaction frequency of the items being operated on.

[0053] S2: Perform spatio-temporal alignment on multi-source heterogeneous data and fuse the commodity granularity data of the transaction system to construct a three-dimensional trajectory model.

[0054] Furthermore, construct a central API gateway to receive and route data requests from the wireless positioning device and distributed environmental sensors. Using a mapping table, convert the different data structures of each device into a unified format and transmit them to the data processing platform;

[0055] Introduce an adaptive heterogeneous data parser to automatically identify and process data from different sources according to data characteristics, and remove noise data and redundant information;

[0056] Assign a unique timestamp and geographic location identifier to each piece of data, and establish a connection between the user's activity trajectory and the dynamic environmental data stored in the database. Spatio-temporal alignment adopts an improved sliding window protocol: use the RFID trigger event as the reference timestamp; match the skeleton pose of visual data within a ±300ms window; perform multi-path effect compensation on the Wi-Fi signal strength and then interpolate for positioning; calculate the distribution of untraded item attributes by comparing the difference between the RFID activation times and POS transaction records.

[0057] Establish a spatial index based on location (commercial area, floor), establish a time index based on time period (hour, day), and associate time and space information through two-way association pointers, so that when accessing one index, the associated data of the other dimension can be quickly obtained, improving the data query efficiency.

[0058] It should be noted that the construction of the three-dimensional trajectory model adopts collaborative processing based on cloud computing technology to construct a multi-level cloud computing architecture;

[0059] Edge computing nodes are set at the user access layer. By setting up edge devices, user data is collected and preliminarily processed, the original data stream provided by mobile devices is processed, and the processing throughput is designed to be 100,000 items per second. The data stream includes the user's location information, timestamp, and additional environmental data (such as temperature, humidity, etc.).

[0060] Within each edge node, a lightweight data processing module is developed for lightweight data preprocessing, including data cleaning, noise filtering, etc., to perform real-time analysis on user behavior data. The cleaned data is transmitted to the central cloud node through a secure encrypted channel for further analysis;

[0061] Furthermore, when performing user behavior analysis at the central node, the cubic spline interpolation method is used to smooth the moving speed, and the user speed data is smoothed with high precision to eliminate noise and obtain more realistic speed characteristics;

[0062] Spark Structured Streaming is used for stream data processing. The sliding window size is set to 60 seconds for real-time aggregation analysis. This process includes the aggregation statistics of user behavior, such as calculating the average residence time and moving speed of each user in the recent 60 seconds.

[0063] Furthermore, a multi-dimensional feature vector is established. By analyzing the product categories associated with each residence event, corresponding features are constructed for each product to improve accuracy. A daily new user behavior data trigger mechanism is set up, and on this basis, the three-dimensional trajectory model is fine-tuned.

[0064] The real-time data is dynamically sharded according to the time range, user identity, or geographical location. Each shard contains the user activity data within the time period; a RESTful API is created to enable communication between the edge node and the cloud for two-way data transfer and command issuance.

[0065] S3: Based on the three-dimensional trajectory model, identify the residence characteristics and movement patterns of users, associate the product distribution to generate a dynamic heat map; adjust the guidance strategy and environmental parameters according to the density gradient of the heat map and the user attribute level.

[0066] Furthermore, extract the behavior characteristics of residence time, frequency, and moving speed from the user behavior data, and use a deep neural network (DNN) to fit the user's movement trajectory and construct a user activity model;

[0067] It should be noted that the input layer of the user activity model receives various status information, including but not limited to:

[0068] The user location information, that is, the current coordinates (x, y), can be the real-time coordinate position of the user within the commercial space.

[0069] A timestamp, used to capture the time characteristics of user behavior (such as hour, weekend / weekday).

[0070] User historical behavior, including the user's last few behaviors (such as the time and location of the last stay), to help the user activity model capture the user's habits.

[0071] The hidden layer of the user activity model uses 3 - 5 layers of fully connected hidden layers, with each layer containing multiple neurons:

[0072] The ReLU (Rectified Linear Unit) activation function is used to introduce non - linearity and accelerate convergence; Dropout is adopted to reduce the risk of overfitting and improve generalization ability.

[0073] The output layer is used to predict the user's next behavior:

[0074] When making a rough behavior prediction, it is regarded as a binary classification problem and outputs two nodes, representing the probabilities of "moving" and "staying" respectively. Among them, the F1 - score of the move / stay classification ≥ 0.92 to ensure the reliability of the coarse - grained behavior judgment.

[0075] When making an accurate prediction of the next location, a regression layer is created to output the user's next location coordinates (x', y'), providing the specific location of the user's expected behavior.

[0076] Furthermore, the mean Euclidean distance error (MAE) between the predicted location and the true coordinates, when meters, is determined to converge. Calculation method:

[0077] ;

[0078] Among them, is the total number of samples, 、 are the horizontal and vertical coordinate values of the predicted location respectively, 、 are the horizontal and vertical coordinate values of the true location respectively.

[0079] Apply the behavior cloning method in reinforcement learning to simulate the typical movement patterns of users in commercial spaces. Classify the user's trajectory into different activity stages (such as browsing, purchasing, resting). Through a real - time data feedback mechanism, continuously monitor the changes in user behavior data, adopt stream data analysis techniques, capture the user behavior trends in real - time, and adjust the parameters of the user activity model;

[0080] Through the incremental learning method, without retraining the user activity model, only adjust the parameters based on the new added data, so that the user activity model always reflects the latest user behavior characteristics.

[0081] It should be noted that the error gradient of the newly added data to the current model is calculated using a regularization loss function (such as L2 regularization), and the formula is:

[0082] ;

[0083] where is the error gradient, α is the dynamic learning rate (adaptively adjusted according to the data distribution change), λ is the regularization coefficient (to prevent overfitting), is the gradient of the loss function of the new data, is the input data;

[0084] The parameters of the underlying feature extraction layer of the model (the first 3 fully connected networks) are fixed, and only the parameters of the top-level behavior prediction layer are fine-tuned to ensure that the basic feature expression ability is not damaged. The confidence level is measured by the Softmax probability value. When the prediction confidence level of the newly added data is lower than 0.7, local parameter update is triggered; otherwise, the original parameters are retained. Since the edge nodes upload a data shard every 60 seconds, the central node updates in mini-batches, and the single-parameter adjustment amplitude is limited by the dynamic learning rate α to avoid overall model oscillation.

[0085] Furthermore, the user behavior data is converted into a heatmap format to distinguish the heat in different time periods;

[0086] Capture the non-linear relationships in the user behavior data, identify the behavioral characteristics of user activities, and map the behavioral characteristics to the heatmap space through quadratic or cubic polynomial function mapping. The user hotspots are divided into high, medium, and low heat zones, and each heat zone corresponds to a different color. Priority is given to planning a detour path for high-consumption-level users to the low-density waiting area.

[0087] Integrate a simple user feedback survey function, combine the user feedback with the heatmap data, use a simple weighted model to calculate the impact of user satisfaction on the heat data, and dynamically update the heatmap according to the results.

[0088] S4: Integrate the payment feature data and the user behavior data to update the user tags and generate a cross-regional association strategy.

[0089] Furthermore, through the collected user movement trajectories, generate a behavior sequence including timestamps, coordinates, and stay durations; align the payment moment with the stay events in the user trajectory to bind the same user entity;

[0090] Establish a multi-dimensional fusion index to associate user behavior and consumption characteristics into a unified data object. When a user stays in any functional area for a long time but does not consume, and generates a per-customer price higher than the average in the remaining areas, it is marked as a potential association behavior;

[0091] Build a multi-level label system, including a basic attribute layer, a behavior derivative layer, and a deep interaction type, and use an incremental learning model to dynamically correct label weights: when the behavior of the labeled user deviates from the historical pattern, reduce the confidence level of the current label's association with the category;

[0092] When the user leaves the current area, match the optimal associated area according to the label: for "price-sensitive" users, preferentially push the navigation of adjacent areas with promotional activities; for "experience-oriented" users, recommend the path of the display area equipped with interactive devices;

[0093] When any functional area gathers labeled users, automatically trigger a collaborative strategy to adjust the commodity display density in adjacent areas and highlight the associated categories;

[0094] Deploy a module for behavior feedback analysis, and use the changes in user behavior after the execution of the strategy as an optimization signal. When any associated strategy fails to achieve the expected goal, automatically downgrade the priority of the current strategy.

[0095] Example 2, the second example of the present invention, which is different from the previous example:

[0096] If the said function is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of the present invention, in essence, or the part that makes contributions to the prior art, or a part of this 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 for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. And the aforementioned storage medium includes: USB flash drives, mobile hard disks, read-only memories (ROM, Read-Only Memory), random access memories (RAM, Random Access Memory), magnetic disks, or optical discs and other various media that can store program codes.

[0097] The logic and / or steps represented in the flowchart or described in other ways herein, for example, can be considered as a definite sequence list of executable instructions for implementing logical functions, and can be specifically implemented in any computer-readable medium for use by an instruction execution system, apparatus, or device (such as a computer-based system, a system including a processor, or other systems that can fetch instructions from the instruction execution system, apparatus, or device and execute the instructions), or in combination with these instruction execution systems, apparatus, or devices. For the purposes of this specification, a "computer-readable medium" can be any device that can contain, store, communicate, propagate, or transmit a program for use by or in connection with an instruction execution system, apparatus, or device.

[0098] More specific examples (non-exhaustive list) of computer-readable media include the following: electrical connections (electronic devices) having one or more wirings, portable computer disk cartridges (magnetic devices), random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber devices, and portable compact disc read-only memory (CDROM). Additionally, the computer-readable media can even be paper or other suitable media on which the program can be printed, because the program can be obtained electronically, for example, by optically scanning the paper or other media, followed by editing, interpretation, or otherwise processing as appropriate, and then stored in a computer memory.

[0099] It should be understood that various parts of the present invention can be implemented by hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, any one or a combination of the following techniques well known in the art can be used: discrete logic circuits having logic gate circuits for implementing logical functions on data signals, application-specific integrated circuits having appropriate combinational logic gate circuits, programmable gate arrays (PGA), field programmable gate arrays (FPGA), etc.

[0100] Example 3, referring to Figure 2 , is an embodiment of the present invention, which provides a user consumption behavior analysis system for different scenarios in a commercial space, including a data acquisition module, a data processing module, a user behavior analysis module, a dynamic heat map generation module, and a policy generation and execution module;

[0101] The data acquisition module, through a wireless positioning device and a distributed environmental sensor, obtains the behavior data of users in real time, is responsible for receiving the signals of the devices carried by users, the data of environmental sensors, and the data of RFID readers, and records the movement trajectories, interaction frequencies, and related attribute information of users in the commercial space;

[0102] The data processing module aligns the multi-source heterogeneous data collected in space and time, and performs data cleaning and conversion to construct a structured data set available for analysis;

[0103] The user behavior analysis module, based on the cleaned user behavior data, extracts the residence characteristics and movement patterns of users, and optimizes the user activity model by using an adaptive learning mechanism;

[0104] The dynamic heat map generation module converts the processed user behavior data into a heat map format, generates user activity heat maps for different time periods, and shows the areas of user attention to merchants;

[0105] The strategy generation and execution module generates targeted business strategies based on the analysis results, including cross-regional association strategies, and adjusts product display and user guidance in real time.

[0106] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit them. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the technical solutions of the present invention, and they should all be covered within the scope of the claims of the present invention.

Claims

1. A method for analyzing user consumption behaviors in different scenarios within a commercial space, characterized in that: including obtaining user behavior data through a wireless positioning device and distributed environmental sensors performing spatio-temporal alignment on multi-source heterogeneous data, establishing a connection between the user's activity trajectory and the dynamic environmental data stored in the database, calculating the similarity of timestamps and location identifiers in space and time, dynamically associating the data, and constructing a three-dimensional trajectory model identifying the residence characteristics and movement patterns of the user based on the three-dimensional trajectory model, and generating a dynamic heat map by associating commodity distributions the identifying the residence characteristics and movement patterns of the user includes extracting the behavior characteristics of residence time, frequency, and movement speed from the user behavior data and constructing a user activity model the input layer of the user activity model receives various state information, including user location information, that is, the current coordinates (x, y), which are the real-time coordinate positions of the user in the commercial space a timestamp, which is used to capture the time characteristics of user behavior the user's historical behavior, which helps the user activity model capture the user's habits the hidden layer of the user activity model uses a fully connected hidden layer with 3-5 layers, and each layer contains multiple neurons using the ReLU activation function to introduce non-linearity and accelerate convergence; adopting Dropout to reduce the risk of overfitting and improve the generalization ability the output layer is used to predict the user's next behavior when making a rough behavior prediction, it outputs two nodes as a binary classification problem, representing the probabilities of "moving" and "staying" respectively, where the F1-score of the moving / staying classification ≥ 0.92 when making an accurate prediction of the next location, a regression layer is created to output the next location coordinates (x', y') of the user, providing the specific location of the user's expected behavior The mean Euclidean distance error MAE between the predicted position and the true coordinates, when meters, it is determined to converge; calculation method: ； Among them, is the total number of samples, , are the horizontal and vertical coordinate values of the predicted position respectively, , are the horizontal and vertical coordinate values of the true position respectively; applying the behavior cloning method in reinforcement learning, simulating the typical movement patterns of users in the commercial space, continuously monitoring the changes in user behavior data through a real-time data feedback mechanism, adopting stream data analysis technology, capturing the user behavior trend in real time, and adjusting the parameters of the user activity model through the incremental learning method, without retraining the user activity model, only adjusting the parameters based on the new data, so that the user activity model always reflects the latest user behavior characteristics using a regularization loss function to calculate the error gradient of the new data for the current model, and the formula is ； wherein, is the error gradient, α is the dynamic learning rate, and λ is the regularization coefficient, is the gradient of the new data loss function, is the input data; the parameters of the first 3 layers of the fully connected network in the underlying feature extraction layer of the model are fixed, and only the parameters of the top-level behavior prediction layer are adjusted; the confidence level is measured by the Softmax probability value. When the prediction confidence level of the new data, it triggers local parameter update; otherwise, the original parameters are retained. Since the edge nodes upload a data shard every minute, the central node updates in micro-batches, and the single-parameter adjustment amplitude is limited by the dynamic learning rate α the generating the dynamic heat map includes converting the user behavior data into a heat map format and distinguishing the heat of different time periods capturing the non-linear relationship in the user behavior data, identifying the behavior characteristics of user activities, and mapping the behavior characteristics to the heat map space through a quadratic or cubic polynomial function mapping. The user hot spots are divided into high, medium, and low heat zones, and each heat zone corresponds to a different color. Priority is given to planning a detour path for high-consumption-level users to the low-density waiting area Adjust the guidance strategy according to the density gradient of the heat map and the user attribute level, combine user feedback with heat map data, calculate the impact of user satisfaction on heat data, and dynamically update the heat map according to the results; Integrate payment feature data and user behavior data to update user tags, generate cross-regional association strategies, and use the changes in user behavior after the execution of the strategies as optimization signals. When any association strategy fails to meet the expected goal, automatically downgrade the priority of the current strategy.

2. The method for analyzing user consumption behaviors in different scenarios within a commercial space according to claim 1, wherein: The wireless positioning device includes positioning the moving trajectory of a user in a commercial space by receiving the signal of a mobile device carried by the user; The distributed environmental sensor includes an RFID reader and a pressure sensing device deployed in a functional area.

3. The method for analyzing user consumption behaviors in different scenarios within a commercial space according to claim 2, wherein: The spatio-temporal alignment of multi-source heterogeneous data includes constructing a central API gateway, receiving and routing data requests from the wireless positioning device and the distributed environmental sensor, and using a mapping table to convert different data structures of each device into a unified format and transmit them to the data processing platform; Adopt an adaptive heterogeneous data parser to automatically identify and process data from different sources according to data characteristics, and remove noise data and redundant information; Assign a unique timestamp and geographical location identifier to each piece of data, establish a connection between the user's activity trajectory and the dynamic environmental data stored in the database, calculate the similarity of the timestamp and location identifier in space and time, and dynamically associate the data.

4. The method for analyzing user consumption behaviors in different scenarios within a commercial space according to claim 3, wherein: The construction of the three-dimensional trajectory model includes that the construction of the three-dimensional trajectory model adopts collaborative processing based on cloud computing technology to construct a multi-level cloud computing architecture; Edge computing nodes are set at the user access layer. User behavior data is collected and preliminarily processed by setting up edge devices. Inside each edge node, a lightweight data processing module is developed to perform real-time analysis on user behavior data. The cleaned data is transmitted to the central cloud node through a secure encrypted channel for further analysis; Create a RESTful API to enable communication between edge nodes and the cloud, and perform two-way data transfer and command issuance.

5. The method for analyzing user consumption behaviors in different scenarios within a commercial space according to claim 4, wherein: The generation of cross-regional association strategies includes generating a behavior sequence containing timestamps, coordinates, and residence duration through the collected user movement trajectory; aligning the payment moment with the residence event in the user trajectory and binding the same user entity; Establish a multi-dimensional fusion index to associate user behavior and consumption characteristics into a unified data object. When a user stays in any functional area for a long time but does not consume, and generates a customer unit price higher than the average in the remaining areas, it is marked as a potential association behavior; Construct a multi-level label system. The multi-level label system includes a basic attribute layer, a behavior derivative layer, and a deep interaction type. An incremental learning model is used to dynamically correct the label weights: when the behavior of the labeled user deviates from the historical pattern, reduce the association confidence of the current label and category; When the user leaves the current area, match the optimal associated area according to the label; when any functional area gathers labeled users, automatically trigger a collaborative strategy to adjust the commodity display density in adjacent areas and highlight the associated categories.

6. A system adopting a method for analyzing user consumption behaviors in different scenarios within a commercial space as described in any one of claims 1 to 5, characterized in that: Include a data collection module, a data processing module, a user behavior analysis module, a dynamic heat map generation module, and a strategy generation and execution module; The data acquisition module receives device signals carried by users, environmental sensor data, and data from RFID readers, and records the movement trajectories, interaction frequencies, and relevant attribute information of users in the commercial space; The data processing module performs spatio-temporal alignment on the collected multi-source heterogeneous data, and performs data cleaning and transformation; The user behavior analysis module extracts the residence characteristics and movement patterns of users, and optimizes the user activity model; The dynamic heat map generation module converts the processed user behavior data into a heat map; The policy generation and execution module generates cross-regional association policies based on the analysis results.

7. A computer device, comprising a memory and a processor, the memory storing a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method described in any one of claims 1 to 5.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the method described in any one of claims 1 to 5.

Images