Information pushing method and system based on big data matching and GPS positioning

Abstract

The invention provides an information pushing method and system based on big data matching and GPS positioning, and the method comprises the steps: receiving position data, processing the position data through a Kalman filtering algorithm, predicting a movement trend through combining a timestamp and velocity vector information, and employing a trajectory prediction algorithm, and generating a potential region of interest list; using a spatial clustering analysis algorithm to analyze historical position data of the user, identifying a behavior mode and a preference location type of the user, and constructing a user interest model according to the behavior mode and the preference location type; utilizing the user interest model to configure a dynamic adaptive geo-fencing system; sequencing the service providing points by adopting a weighted K nearest neighbor algorithm, and combining geographic position information in the service recommendation list with other sensor data by utilizing a multi-modal fusion algorithm to generate a customized information packet; according to the technical scheme provided by the invention, the accuracy of the position information is improved, the prediction precision and timeliness are enhanced, and personalized service recommendation is enhanced.

Description

Technical Field

[0001] The embodiments of the present invention relate to the field of GPS technology, and in particular to an information push method based on big data matching and GPS positioning. Background Art

[0002] With the popularization of mobile Internet and smart devices, users' demand for personalized information services is growing. In modern urban life, people rely on mobile devices such as smartphones to obtain real-time traffic information, surrounding service recommendations, and personalized advertising push, etc.

[0003] Although existing information push methods have achieved location-based service recommendations to a certain extent, they still have some significant shortcomings. First, they lack accuracy. Most existing solutions only rely on simple geographic location information for service recommendations, and fail to fully utilize users' behavior patterns and preferred location types, resulting in inaccurate recommendation results and failing to fully reflect users' actual needs. Second, they lack responsiveness. Traditional geo-fencing technology usually uses fixed boundary settings and cannot be flexibly adjusted according to users' real-time locations and behavior patterns. When users move quickly or are in complex environments, they may miss important service provision points or mistakenly include irrelevant areas, reducing the effectiveness and pertinence of service recommendations. In addition, due to limited data processing capabilities, existing systems often find it difficult to efficiently process location data with high-frequency GPS coordinate updates, especially when faced with a large number of users requesting at the same time, which is prone to delays or data loss. Furthermore, multi-source data fusion is insufficient. Existing solutions rarely effectively integrate geographic location information with other sensor data (such as images, sounds, temperature and humidity, etc.), limiting the richness and accuracy of recommended content. Finally, the degree of personalization is low. Due to the failure to deeply explore users' long-term behavior patterns and immediate needs, the recommendation results of existing systems often lack personalization and the user experience is poor. Summary of the invention

[0004] The embodiment of the present invention provides an information push method based on big data matching and GPS positioning, which is used to solve the problems of lack of accuracy, insufficient responsiveness, limited data processing capability, insufficient fusion of multi-source data and low personalization in the prior art.

[0005] In a first aspect, an embodiment of the present invention provides an information push method based on big data matching and GPS positioning, comprising:

[0006] Receiving real-time data streams from different sources, the real-time data streams containing structured data and unstructured data;

[0007] Receive in real time the location data of high-frequency GPS coordinate updates from the user's mobile device, and process the location data using a Kalman filter algorithm to obtain real-time location information;

[0008] Based on the real-time location information, combined with timestamp and velocity vector information, a trajectory prediction algorithm is applied to predict the user's movement trend and generate a list of potential areas of interest;

[0009] Using a spatial clustering analysis algorithm to analyze the potential interest area list and the user's historical location data, identify the user's behavior pattern and preferred location type, and build a user interest model based on the user's behavior pattern and preferred location type;

[0010] Utilizing the user interest model to configure a dynamic adaptive geo-fence system, the dynamic adaptive geo-fence system is used to automatically adjust the fence boundary according to the user's real-time location, behavior pattern, and surrounding environment changes to generate a set of service provision points;

[0011] A weighted K-nearest neighbor algorithm is used to sort multiple service provision points in the service provision point set to obtain a service recommendation list, and a multimodal fusion algorithm is used to combine the geographical location information in the service recommendation list with other sensor data to generate a customized information package.

[0012] Optionally, a spatial clustering analysis algorithm is used to analyze the potential interest area list and the user's historical location data to identify the user's behavior pattern and preferred location type, and a user interest model is constructed according to the user's behavior pattern and preferred location type, including:

[0013] The user's historical location data is denoised using a spatiotemporal consistency check algorithm to remove noise points caused by GPS signal drift, and missing values ​​are filled in using a time series filling algorithm to unify the timestamp format to obtain a complete and consistent user location record.

[0014] Based on the user location record and in combination with timestamp information, the user's visit behavior is extracted, the user's stay records at different points of interest as well as the arrival and departure times are identified, and the user's visit behavior sequence is generated;

[0015] According to the user access behavior sequence and the potential interest area list, a graph matching algorithm is used to convert the user access behavior sequence into a weighted directed graph, wherein nodes represent interest points, and edge weights represent access frequency and residence time. Through the weighted directed graph, the correlation between the user access behavior sequence and the potential interest area list is established, and the access probability matrix of each interest point is calculated;

[0016] By using the access probability matrix, adopting the hierarchical density peak clustering algorithm and combining the geo-fencing technology, the geographical location, access frequency and residence time are calculated, and the points of interest visited by the user are grouped. The clustering parameters are automatically adjusted through the dynamic threshold adjustment mechanism to obtain the clustering results reflecting the user's frequented places and behavior patterns;

[0017] Based on the clustering results, a hidden Markov model is applied to analyze the user's mobility state transition rules in different time periods to identify the user's behavior pattern, and semantic clustering is performed on different types of points of interest to identify the preferred location type;

[0018] Based on the user behavior patterns and preferred place types, the geographical distance, visit frequency, dwell time and thematic classification of points of interest are comprehensively calculated, a comprehensive score is calculated for each point of interest, and a multi-dimensional user interest model is constructed.

[0019] Optionally, according to the user access behavior sequence, combined with the potential interest area list, a graph matching algorithm is used to convert the user's access behavior sequence into a weighted directed graph, wherein nodes represent interest points, and edge weights represent visit frequency and stay time. Through the weighted directed graph, the correlation between the user's access behavior sequence and the potential interest area list is established, and the access probability matrix of each interest point is calculated, including:

[0020] The spatiotemporal consistency check algorithm is used to process the temporal and spatial coherence of the movement paths in the user's access behavior sequence, and an undirected graph representing the user's movement pattern is obtained, in which the nodes represent different points of interest;

[0021] According to the user's stay time and visit frequency at each point of interest, combined with the edge weights optimized by the machine learning model, the undirected graph is weighted and directed to generate a weighted directed graph reflecting the time sequence of the user's visit to the points of interest;

[0022] Based on the weighted directed graph, the user access behavior sequence is compared with the information of the points of interest in the list of potential interest areas, and a context-aware mechanism is introduced to establish the correlation between the two, so as to obtain the relevance of the adjusted points of interest;

[0023] For each point of interest, based on the relevance of the adjusted point of interest, a Bayesian network is used to perform probabilistic reasoning processing to calculate the probability of the point of interest being visited, thereby forming a visit probability matrix.

[0024] Optionally, according to the user's stay time and visit frequency at each point of interest, combined with the edge weights optimized by the machine learning model, the undirected graph is weighted and directed to generate a weighted directed graph reflecting the time sequence of the user's visit to the points of interest, including:

[0025] Optimizing the edge weights using a machine learning model trained based on a large amount of historical access data to generate optimized edge weights that can dynamically adapt to changing patterns of user behavior;

[0026] According to the optimized edge weights and the timestamp information of the user visiting the points of interest, a direction determination process is performed on the edges between each pair of points of interest in the undirected graph to obtain a direction definition reflecting the actual time sequence of the user's visits;

[0027] The undirected graph is converted using the directionality definition to generate an initial weighted directed graph;

[0028] Based on the anomaly detection algorithm, the paths in the weighted directed graph that do not conform to the user's regular behavior pattern are identified and excluded, so as to obtain an optimized weighted directed graph that can more realistically represent the user's daily mobility pattern.

[0029] Optionally, based on the real-time location information, combined with the timestamp and speed vector information, a trajectory prediction algorithm is applied to predict the user's movement trend, and a list of potential interest areas is generated, including:

[0030] The real-time location information processed by the Kalman filter algorithm is combined with the timestamp and speed vector information to reconstruct the user's historical movement path to obtain a continuous space-time trajectory. Based on the continuous space-time trajectory, a multi-step recursive neural network is used to perform short-term and medium-term prediction processing on the user's future movement path to obtain multiple predicted movement trend results;

[0031] Introducing external environment data, and using a context-aware model to adjust the movement trend result to obtain an optimized movement trend prediction;

[0032] Based on the optimized mobile trend prediction, the user's geographical location is predicted and processed in combination with the interest point database in the geographic information system to generate a preliminary list of interest areas;

[0033] Applying a spatiotemporal density clustering algorithm, based on the geographical distribution characteristics near the predicted path, the preliminary list of regions of interest is optimized to obtain a cluster-optimized list of regions of interest;

[0034] The personalized preference model is applied to filter the cluster-optimized ROI list, prioritize the POIs that meet the user's preferences, and generate a personalized list of potential ROIs.

[0035] Optionally, configuring a dynamic adaptive geo-fence system using the user interest model, wherein the dynamic adaptive geo-fence system is used to automatically adjust the fence boundary according to the user's real-time location, behavior pattern, and surrounding environment changes to generate a set of service provision points, including:

[0036] The user interest model is used to comprehensively analyze the user's real-time location information, behavior patterns, and preferred location types to determine the current user's behavior status and potential destinations;

[0037] Based on the user's real-time location information and movement trend results, combined with timestamp information and speed vector information, a spatiotemporal prediction algorithm is used to estimate the possible arrival location and obtain an estimated location set;

[0038] Combining the user's behavior patterns, preferred location types, and environmental influencing factors, the initial boundaries of the geo-fence are set and processed through a machine learning optimization algorithm to obtain the initial geo-fence boundaries;

[0039] Introducing a dynamic threshold adjustment mechanism to automatically adjust the initial geo-fence boundary based on real-time feedback from users and new location data to generate an optimized geo-fence boundary;

[0040] A context-aware recommendation system is applied to screen and sort the service provision points within the optimized geographic fence in combination with the user's real-time needs, historical preferences and actual availability of the service provision points, thereby generating an optimized set of service provision points.

[0041] Optionally, a weighted K-nearest neighbor algorithm is used to sort multiple service provision points in the service provision point set to obtain a service recommendation list, and a multimodal fusion algorithm is used to combine the geographical location information in the service recommendation list with other sensor data to generate a customized information package, including:

[0042] Using a weighted K-nearest neighbor algorithm, based on a user interest model, real-time location information, and geographic distance, visit frequency, dwell time, and rating of the service provider points, a plurality of service provider points in the service provider point set are comprehensively evaluated and ranked to obtain a personalized service recommendation list;

[0043] Introducing a context-aware module, combined with external environment data, dynamically adjusts the personalized service recommendation list to generate an optimized service recommendation list;

[0044] Based on the optimized service recommendation list, a multimodal fusion algorithm is applied to integrate data from different sources, perform multi-dimensional feature description processing on each recommended service provider, and generate service information cards;

[0045] Using natural language processing technology to perform semantic understanding and sentiment analysis on the text content in the service information card, extract key information and optimize the expression method to obtain the content display form;

[0046] The content presentation forms are integrated into a comprehensive information structure, and customized information packages are generated according to user behavior patterns.

[0047] In a second aspect, an embodiment of the present invention provides an information push system based on big data matching and GPS positioning, including:

[0048] A prediction module, for predicting the user's movement trend based on the real-time location information, in combination with the timestamp and the speed vector information, and generating a list of potential interest areas by applying a trajectory prediction algorithm;

[0049] An identification module, used to analyze the potential interest area list and the user's historical location data using a spatial clustering analysis algorithm, identify the user's behavior pattern and preferred location type, and build a user interest model based on the user's behavior pattern and preferred location type;

[0050] A configuration module, configured to configure a dynamic adaptive geo-fence system using the user interest model, wherein the dynamic adaptive geo-fence system is configured to automatically adjust the fence boundary according to the user's real-time location, behavior pattern, and surrounding environment changes to generate a set of service provision points;

[0051] A generating module is used to sort the multiple service providing points in the service providing point set by using a weighted K-nearest neighbor algorithm to obtain a service recommendation list, and to combine the geographical location information in the service recommendation list with other sensor data by using a multimodal fusion algorithm to generate a customized information package.

[0052] In a third aspect, an embodiment of the present invention provides a computing device, comprising a processor and a memory, wherein the memory stores a computer program, and the processor is configured to run the computer program to execute any one of the information push methods based on big data matching and GPS positioning as described in the first aspect.

[0053] In a fourth aspect, an embodiment of the present invention provides a computer storage medium having computer program instructions stored thereon, wherein the computer program instructions, when executed by a processor, implement an information push method based on big data matching and GPS positioning as described in any one of the first aspects.

[0054] In an embodiment of the present invention, location data updated by high-frequency GPS coordinates from a user's mobile device is received in real time, and the location data is processed using a Kalman filter algorithm to obtain real-time location information; based on the real-time location information, in combination with timestamp and velocity vector information, a trajectory prediction algorithm is applied to predict the user's movement trend and generate a list of potential interest areas; a spatial clustering analysis algorithm is used to analyze the list of potential interest areas and the user's historical location data to identify the user's behavior pattern and preferred location type, and a user interest model is constructed based on the user's behavior pattern and preferred location type; a dynamic adaptive geo-fence system is configured using the user interest model, and the dynamic adaptive geo-fence system is used to automatically adjust the fence boundary according to the user's real-time location, behavior pattern and surrounding environmental changes to generate a service provider point set; a weighted K nearest neighbor algorithm is used to sort multiple service providers in the service provider point set to obtain a service recommendation list, and a multimodal fusion algorithm is used to combine the geographic location information in the service recommendation list with other sensor data to generate a customized information package. The technical solution provided by the present invention improves the accuracy of location information, enhances the accuracy and timeliness of prediction, enhances personalized service recommendations, can dynamically respond to environmental changes, and applies multi-source data fusion to improve recommendation quality;

[0055] Furthermore, the spatiotemporal consistency check algorithm is used to denoise the user's historical location data, remove noise points caused by GPS signal drift, and fill in missing values ​​through a time series filling algorithm to unify the timestamp format and obtain complete and consistent user location records. This process ensures the quality and integrity of the data used for analysis, laying a solid foundation for subsequent analysis;

[0056] Based on the user's location records and timestamp information, the user's visit behavior is extracted, the user's stay records at different points of interest, as well as the arrival and departure times, are identified, and the user's visit behavior sequence is generated. This helps to understand the user's daily activity patterns in more detail, so as to better meet their personalized needs;

[0057] According to the user's access behavior sequence and the list of potential interest areas, the graph matching algorithm is used to transform the user's access behavior sequence into a weighted directed graph, where nodes represent points of interest and edge weights represent visit frequency and stay time. Through this conversion, the correlation between user behavior patterns and interest areas can be intuitively displayed, providing strong support for subsequent analysis.

[0058] Through a weighted directed graph, the correlation between the user's visit behavior sequence and the list of potential interest areas is established, and the visit probability matrix of each interest point is calculated. This matrix can quantify the possibility of users visiting each interest point, helping the system to more accurately predict the user's future behavior and thus optimize service recommendations;

[0059] These and other aspects of the present invention will become more apparent from the following description of the embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are 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.

[0061] Figure 1 A flowchart of an information push method based on big data matching and GPS positioning provided by an embodiment of the present invention;

[0062] Figure 2 A schematic diagram of the structure of an information push system based on big data matching and GPS positioning provided by an embodiment of the present invention;

[0063] Figure 3 A schematic diagram of the structure of a computing device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0064] In order to enable those skilled in the art to better understand the solutions of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention.

[0065] In some of the processes described in the specification and claims of the present invention and the above-mentioned figures, multiple operations that appear in a specific order are included, but it should be clearly understood that these operations may not be executed in the order in which they appear in this article or executed in parallel. The serial numbers of the operations, such as 101, 102, etc., are only used to distinguish different operations, and the serial numbers themselves do not represent any execution order. In addition, these processes may include more or fewer operations, and these operations may be executed in sequence or in parallel. It should be noted that the descriptions of "first", "second", etc. in this article are used to distinguish different messages, devices, modules, etc., do not represent the order of precedence, and do not limit the "first" and "second" to be different types.

[0066] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0067] Figure 1A flowchart of a big data processing method for implementing hybrid data analysis is provided for an embodiment of the present invention. Figure 1 As shown, the method includes:

[0068] Step 101: receiving location data updated by high-frequency GPS coordinates from a user's mobile device in real time, and processing the location data using a Kalman filter algorithm to obtain real-time location information;

[0069] In this step, high-frequency GPS coordinate update means that the mobile device obtains the user's geographic location information at a higher frequency (such as once per second or several seconds);

[0070] Kalman filter algorithm refers to a recursive optimal estimation algorithm, which is often used to estimate the state of a process from a series of incomplete and noisy measurements;

[0071] Receive real-time location data from high-frequency GPS coordinate updates in user mobile devices, and use the Kalman filter algorithm to process these location data to reduce errors caused by factors such as signal interference and multipath effects, and obtain more accurate and stable real-time location information.

[0072] Step 102: According to the real-time location information, combined with the timestamp and speed vector information, a trajectory prediction algorithm is applied to predict the user's movement trend and generate a list of potential interest areas;

[0073] In this step, the timestamp refers to data recording the exact time when an event occurs; the speed vector information refers to information including the speed and direction of the user's movement; the trajectory prediction algorithm refers to an algorithm that predicts possible future paths by analyzing historical trajectory data;

[0074] Based on the real-time location information obtained in step 101, combined with the timestamp and speed vector information, a trajectory prediction algorithm is applied, which predicts the user's movement trend, identifies the places that the user may visit in the future, and generates a list of potential areas of interest, which helps to prepare relevant service recommendations in advance.

[0075] Step 103: Analyze the potential interest area list and the user's historical location data using a spatial clustering analysis algorithm to identify the user's behavior pattern and preferred location type, and construct a user interest model based on the user's behavior pattern and preferred location type;

[0076] In this step, the spatial clustering analysis algorithm refers to an algorithm used to group points in geographic space according to certain similarity criteria; the behavior pattern refers to the behavior characteristics of users at different times and places, such as frequently visited places, activity times, etc.; the preferred place type refers to the category of places preferred by users, such as restaurants, parks, shopping malls, etc.;

[0077] A spatial clustering analysis algorithm is used to analyze the list of potential interest areas and the user's historical location data to identify the user's behavior patterns and preferred location types, and a user interest model is built based on this information. This model can reflect the user's daily activity patterns and personal preferences, providing a basis for personalized service recommendations.

[0078] Step 104: configuring a dynamic adaptive geo-fence system using the user interest model, wherein the dynamic adaptive geo-fence system is used to automatically adjust the fence boundary according to the user's real-time location, behavior pattern, and surrounding environment changes to generate a set of service provision points;

[0079] In this step, the dynamic adaptive geo-fence system refers to a virtual fence system that can automatically adjust the boundaries according to the user's real-time location, behavior pattern and changes in the surrounding environment; the service provider point set refers to a set of places or facilities that may provide services to users, such as stores, restaurants, attractions, etc.;

[0080] A dynamic adaptive geographic fencing system is configured using the user interest model. The system automatically adjusts the fence boundaries according to the user's real-time location, behavior patterns, and changes in the surrounding environment, ensuring that the service provider points within the fence are highly relevant to the user's current needs. The generated service provider point set is the basis for subsequent service recommendations, ensuring the timeliness and accuracy of the recommendation results.

[0081] Step 105: using a weighted K-nearest neighbor algorithm to sort multiple service provision points in the service provision point set to obtain a service recommendation list, and using a multimodal fusion algorithm to combine the geographical location information in the service recommendation list with other sensor data to generate a customized information package;

[0082] In this step, the dynamic adaptive geo-fence system refers to a virtual fence system that can automatically adjust the boundaries according to the user's real-time location, behavior pattern and changes in the surrounding environment; the service provider point set refers to a set of places or facilities that may provide services to users, such as stores, restaurants, attractions, etc.;

[0083] The user interest model is used to configure a dynamic adaptive geo-fence system, which automatically adjusts the fence boundary according to the user's real-time location, behavior pattern, and changes in the surrounding environment to ensure that the service provider points within the fence are highly relevant to the user's current needs. The generated service provider point set is the basis for subsequent service recommendations, ensuring the timeliness and accuracy of the recommendation results;

[0084] Assume that tourist Xiao Li is visiting a famous scenic spot. He opens the smart travel APP on his mobile phone and hopes to obtain service information such as restaurants and shops nearby. The system receives the location data of high-frequency GPS coordinate updates from Xiao Li's mobile device in real time, and uses the Kalman filter algorithm to process these location data, remove noise interference, and obtain accurate real-time location information (such as longitude and latitude). Next, the system uses the trajectory prediction algorithm to predict Xiao Li's movement trend based on the real-time location information, combined with the timestamp and speed vector information, and generates a list of potential interest areas, such as nearby major attractions, restaurants, shopping malls, etc. Subsequently, the spatial clustering analysis algorithm is used to analyze the list of potential interest areas and Xiao Li's historical location data, identify Xiao Li's behavior patterns (such as liking to visit historical and cultural attractions, preferring cafes during afternoon tea hours) and preferred location types (such as museums, art galleries, and specialty restaurants), and build a user interest model that reflects Xiao Li's interests and preferences;

[0085] Furthermore, the system uses the user interest model to configure a dynamic adaptive geographic fence system, which automatically adjusts the fence boundary according to Xiao Li's real-time location, behavior patterns, and changes in the surrounding environment (such as traffic conditions and weather conditions), ensuring that the recommended service points are in line with Xiao Li's interests and take into account the current actual situation. For example, when a large-scale event is detected nearby that causes traffic congestion, the system will adjust the fence range accordingly and recommend alternative routes or more convenient service points. Then, the system uses a weighted K-nearest neighbor algorithm to sort multiple service providers in the service provider point set to obtain a service recommendation list, such as nearby popular restaurants, exhibitions, etc. Then, the multimodal fusion algorithm is used to combine the geographic location information in the service recommendation list with other sensor data (such as weather forecasts and traffic flow) to generate a customized information package. The customized information package not only contains accurate service recommendations, but also integrates other types of rich information, such as pictures, audio introductions, etc., so that Xiao Li can fully understand the specific situation of each recommended point.

[0086] Finally, the system pushes the customized information package to Xiao Li. When Xiao Li approaches a history museum, the system pushes a detailed introduction to the museum and information about upcoming special exhibitions. When Xiao Li's lunch time is approaching, the system recommends several highly rated local restaurants, and attaches menus and user reviews. In addition, considering that it will rain that day, the system also considerately recommends some indoor activity venues. All this information is intuitively displayed to Xiao Li through the APP, helping him to better plan his next trip, significantly improving Xiao Li's travel experience and making his trip smoother and more enjoyable. At the same time, it also increases the exposure rate of the service provider and user satisfaction, achieving a win-win situation. Through this complete implementation process, it is demonstrated how the method described in claim 1 can effectively improve the accuracy, relevance and user experience of information push in practical applications to meet the diverse needs of users.

[0087] Based on this, the present invention provides a specific embodiment, wherein step 203, according to the user access behavior sequence, combined with the potential interest area list, uses a graph matching algorithm to convert the user's access behavior sequence into a weighted directed graph, wherein nodes represent interest points, and edge weights represent access frequency and residence time. Through the weighted directed graph, the correlation between the user's access behavior sequence and the potential interest area list is established, and the access probability matrix of each interest point is calculated, which specifically includes the following steps:

[0088] Step 210: De-noising the historical location data of the user using a spatiotemporal consistency check algorithm to remove noise points due to GPS signal drift, and fill in missing values ​​using a time series filling algorithm to unify the timestamp format, thereby obtaining a complete and consistent user location record;

[0089] In this step, the spatiotemporal consistency check algorithm refers to an algorithm used to detect and correct inconsistencies in time series data to ensure the consistency of time and space data; GPS signal drift refers to the GPS positioning error caused by satellite signal interference or other factors, which manifests as random fluctuations in location data; the time series filling algorithm refers to the algorithm used to fill missing values ​​in time series data to ensure data continuity and integrity; timestamp format unification refers to converting timestamps from different sources into a unified standard format to facilitate subsequent analysis;

[0090] The user's historical location data is denoised using a spatiotemporal consistency check algorithm to remove noise points caused by GPS signal drift. The missing values ​​in the historical location data are filled using a time series filling algorithm, and all timestamp formats are unified to ensure data integrity and consistency. Ultimately, a cleaned and normalized user location record is obtained, providing a reliable data foundation for subsequent steps.

[0091] Step 211: Based on the user location record and in combination with the timestamp information, the user's visit behavior is extracted, the user's stay records at different points of interest as well as the arrival time and departure time are identified, and the user's visit behavior sequence is generated;

[0092] In this step, user location records refer to the user movement trajectory data after denoising and padding; visit behavior refers to the behavior of users staying at a specific place, including arrival time and departure time; points of interest (POI) refer to places with specific meaning or attraction, such as stores, restaurants, etc.;

[0093] Based on user location records and timestamp information, we identify users’ stay records at different points of interest as well as their specific arrival and departure times, and generate a user’s visit behavior sequence, which describes the user’s activity trajectory over a period of time in detail, providing an important basis for subsequent analysis.

[0094] Step 212: Based on the user's access behavior sequence and the potential interest area list, a graph matching algorithm is used to convert the user's access behavior sequence into a weighted directed graph, wherein nodes represent interest points, and edge weights represent access frequency and residence time. Through the weighted directed graph, the correlation between the user's access behavior sequence and the potential interest area list is established, and the access probability matrix of each interest point is calculated;

[0095] In this step, the graph matching algorithm refers to an algorithm for comparing the structural similarity of two graphs, and is used here to convert the visit behavior sequence into a graph representation; the weighted directed graph refers to a graph representation method, in which nodes represent points of interest and edge weights represent visit frequency and stay time; the visit probability matrix refers to a probability distribution matrix describing the route from one point of interest to another point of interest;

[0096] According to the user's access behavior sequence and the list of potential interest areas, the graph matching algorithm is used to convert the access behavior sequence into a weighted directed graph. In this graph, the nodes represent points of interest, and the edge weights represent the user's visit frequency and residence time. Through the weighted directed graph, the correlation between the user's access behavior sequence and the list of potential interest areas is established, and the access probability matrix of each point of interest is calculated, providing a basis for subsequent clustering analysis.

[0097] Step 213: Utilizing the access probability matrix, adopting a hierarchical density peak clustering algorithm, combined with geo-fencing technology, calculating geographic location, access frequency, and dwell time, grouping the points of interest visited by the user, and automatically adjusting clustering parameters through a dynamic threshold adjustment mechanism to obtain clustering results that reflect the user's frequented locations and behavior patterns;

[0098] In this step, the hierarchical density peak clustering algorithm refers to a density-based clustering algorithm that can automatically discover natural groups in the data; the geo-fencing technology refers to defining virtual boundaries to monitor users entering or leaving a specific area; the dynamic threshold adjustment mechanism refers to automatically adjusting the algorithm parameters according to the actual situation to ensure the accuracy and adaptability of the clustering results;

[0099] The access probability matrix is ​​used, the hierarchical density peak clustering algorithm is adopted, and the geographic location, visit frequency and residence time are combined with the geographic fencing technology to calculate the geographical location, visit frequency and residence time, and the points of interest visited by users are grouped. The dynamic threshold adjustment mechanism automatically adjusts the clustering parameters to ensure that the clustering results can reflect the users' frequented places and behavior patterns, providing support for personalized service recommendations.

[0100] Step 214: Based on the clustering results, a hidden Markov model is applied to analyze the user's mobility state transition rules in different time periods to identify the user's behavior pattern, and semantic clustering is performed on different types of points of interest to identify the preferred location type;

[0101] In this step, Hidden Markov Model (HMM) refers to a statistical model used to represent a system consisting of a series of possibly unseen states with certain transition probabilities between these states; semantic clustering refers to grouping points of interest with similar semantic features into one category, such as "dining", "shopping", etc.;

[0102] Based on the clustering results, the hidden Markov model is used to analyze the user's mobility state transition rules in different time periods and identify the user's behavior patterns. At the same time, semantic clustering is performed on different types of points of interest to identify preferred place types, further refine the user interest model, and improve the accuracy of personalized recommendations.

[0103] Step 215: Based on the user behavior pattern and preferred place type, comprehensively calculate the geographical distance, visit frequency, stay time and interest point theme classification, calculate a comprehensive score for each interest point, and construct a multi-dimensional user interest model;

[0104] In this step, the comprehensive score refers to the score calculated based on multiple dimensions (such as geographical distance, visit frequency, stay time, and POI theme classification) to evaluate the user's interest in a certain POI; the multidimensional user interest model refers to a model that comprehensively considers multiple factors and is constructed to fully reflect the user's interests and preferences;

[0105] Based on user behavior patterns and preferred place types, the geographical distance, visit frequency, dwell time and thematic classification of points of interest are comprehensively calculated, a comprehensive score is calculated for each point of interest, and a multidimensional user interest model is constructed. This model not only reflects the user's geographical location preferences, but also takes into account the visit frequency, dwell time and thematic classification of points of interest, providing users with highly personalized and accurate service recommendations.

[0106] Based on this, the present invention provides a specific embodiment, wherein step 212, based on the user access behavior sequence and the potential interest area list, uses a graph matching algorithm to convert the user's access behavior sequence into a weighted directed graph, wherein nodes represent interest points, and edge weights represent access frequency and residence time. Through the weighted directed graph, the correlation between the user's access behavior sequence and the potential interest area list is established, and the access probability matrix of each interest point is calculated, which specifically includes the following steps:

[0107] Step 301: using a spatiotemporal consistency check algorithm to perform temporal and spatial coherence processing on the movement paths in the user's access behavior sequence, and obtaining an undirected graph representing the user's movement pattern, wherein nodes represent different points of interest;

[0108] In this step, the spatiotemporal consistency check algorithm refers to an algorithm used to ensure the consistency of temporal and spatial data, which can detect and correct incoherent data points; the user access behavior sequence refers to the time series data that records the user's stay time, arrival time and departure time at different points of interest; the undirected graph refers to a graphical representation method, in which nodes represent points of interest and edges have no direction, indicating the relationship between two points of interest;

[0109] The spatiotemporal consistency check algorithm is used to process the temporal and spatial coherence of the movement paths in the user's access behavior sequence to ensure that the temporal and spatial relationship between each location point is reasonable and consistent. Through this processing, an undirected graph representing the user's movement pattern is generated, in which the nodes represent different points of interest. The graph reflects the user's movement path between different points of interest, providing a structured basis for subsequent analysis.

[0110] Step 302: weighting and directing the undirected graph according to the user's stay time and visit frequency at each point of interest, combined with the edge weights optimized by the machine learning model, to generate a weighted directed graph reflecting the time sequence of the user's visit to the points of interest;

[0111] In this step, the dwell time refers to the length of time a user stays at a point of interest; the visit frequency refers to the number of times a user visits a point of interest; the edge weight optimized by the machine learning model refers to the weight value obtained through the training of the machine learning model, which is used to reflect the strength of the association between points of interest;

[0112] According to the user's residence time and visit frequency at each point of interest, combined with the edge weights optimized by the machine learning model, the undirected graph obtained in step 301 is weighted and directed. The weighted processing assigns a weight value to each edge, indicating the importance of the user's visit frequency and residence time; the directed processing determines the direction of the edge according to the user's actual movement order, and finally generates a weighted directed graph reflecting the time sequence of the user's visit to the points of interest. This process makes the user's behavior pattern more intuitively reflected in the graph structure, which is convenient for subsequent analysis and application.

[0113] Step 303: Based on the weighted directed graph, compare the user access behavior sequence with the information of the points of interest in the list of potential interest areas, introduce a context-aware mechanism, establish the correlation between the two, and obtain the relevance of the adjusted points of interest;

[0114] In this step, the weighted directed graph refers to a graphical representation method in which nodes represent points of interest, and edges are directed and weighted to represent visit frequency and dwell time; the context-aware mechanism refers to adjusting the relevance of points of interest by considering factors such as the current time, weather conditions, and public events; relevance: an indicator that measures the likelihood or importance of a user visiting a point of interest;

[0115] Based on the weighted directed graph generated in step 302, the user's access behavior sequence is compared with the information of points of interest in the list of potential interest areas, and a context-aware mechanism is introduced to consider factors such as the current time, weather conditions, and special events, and adjust the relevance of the points of interest to more accurately reflect the user's real-time needs and preferences, establish the correlation between the user's access behavior sequence and the list of potential interest areas, and obtain the relevance of the adjusted points of interest to provide a basis for subsequent probabilistic reasoning.

[0116] Step 304: for each point of interest, based on the adjusted relevance of the point of interest, a Bayesian network is used to perform probabilistic reasoning processing to calculate the probability of the point of interest being visited, thereby forming a visit probability matrix;

[0117] In this step, the Bayesian network refers to a graphical model based on probability theory, which is used to represent the dependency relationship between variables and support probabilistic reasoning; the access probability matrix refers to a probability distribution matrix describing from one point of interest to another point of interest;

[0118] For each point of interest, based on the adjusted relevance of the point of interest, the probability of the point of interest being visited is calculated using the Bayesian network to form a visit probability matrix. This matrix not only reflects the possibility of users visiting each point of interest, but also captures the transition probability between points of interest, providing an important reference for subsequent clustering analysis and service recommendation;

[0119] Let I= { i1,i2,…,i n } is the set of points of interest visited by the user, T ( i j ) For users at point of interest i j The residence time, F ( i j ) is the access frequency, R ( i j ) is the adjusted relevance (considering the context-aware mechanism), C(t) is the time decay factor, and E ( i j ) is an external environmental factor (such as weather, traffic conditions, etc.), then the access probability matrix P can be calculated by the following complex formula:

[0120]

[0121] Where I = { i1,i2,…,i n}It represents the set of points of interest visited by the user, which is extracted from the user's movement trajectory data (such as GPS records) and represents a series of places that the user has visited; ( i j ) indicates that the user is interested in j The dwell time is extracted from the user’s historical location data, usually by using time information to determine the time difference between the user entering and leaving a point of interest; ( i j ) Indicates that the user visits point of interest i j The frequency of the user's historical behavior data is counted to see how many times a certain point of interest is visited; R ( i j ) Represents the adjusted relevance, taking into account the context-aware mechanism, and constructing a user interest model through a spatial clustering analysis algorithm combined with the user's historical behavior pattern and preferred location type to obtain the relevance score;

[0122] C(t) represents the time attenuation factor, which means that as time goes by, the impact of historical access behavior on the current prediction gradually weakens, where C(t) = exp(-λt), t is the time since the last visit, and λ is the decay rate parameter, which reduces the impact of historical behavior on the current prediction and makes the prediction closer to the current behavior pattern; E ( i j ) represents the external environment influencing factor, considering the impact of the external environment (such as weather, traffic conditions, etc.) on the user's visit to the point of interest, where E ( i j ) =w1W ( i j ) +w2T r ( i j ) , W ( i j ) is the weather factor, such as sunny, rainy, etc., T r ( i j ) is a factor affecting traffic conditions, such as the degree of traffic congestion; w1 and w2 are corresponding weight coefficients, which consider the impact of external factors on user behavior, making the recommendation results more in line with the actual situation; α, β, γ, δ, η represent weight parameters, which are used to balance the importance of different factors and can be optimized through machine learning models to ensure that the final access probability matrix can accurately reflect the user's interests and behavior patterns;

[0123] The functions of the numerator and denominator are explained as follows:

[0124] exp(αT(i j )+βF(i j )+γR(i j )+δC(t)+ηE(i j ))

[0125] It is used to calculate the weighted comprehensive score of interest point ij, combining various factors through exponential function to enhance the relative importance difference between different factors;

[0126]

[0127] It is used to normalize the weighted comprehensive scores of all points of interest to ensure that the sum of the probabilities of all points of interest is 1, forming a probability distribution;

[0128] Among them, the residence time T(i j ) indicates that the longer the stay time, the greater the user's interest in the point of interest, so it is given a higher weight when calculating the visit probability; the visit frequency F(i j ) indicates that the higher the visit frequency, the more stable the user's interest in the point of interest is, which increases the possibility of its recommendation; the adjusted relevance R(i j ) represents the relevance of the interest point adjusted by the context-aware mechanism, which reflects the user's preference for the interest point and improves the accuracy of personalized recommendation; the time decay factor C(t) represents reducing the impact of historical behavior on the current prediction, making the recommendation result closer to the user's current behavior pattern; the external environment influence factor E(i j ) means taking into account the impact of external environment (such as weather, traffic conditions, etc.) on behavior, so that the recommendation results are more in line with the actual situation.

[0129] Based on this, the present invention provides a specific embodiment, wherein step 302 performs weighted and directional processing on the undirected graph according to the user's stay time and visit frequency at each point of interest, combined with the edge weights optimized by the machine learning model, to generate a weighted directed graph reflecting the time sequence of the user's visit to the points of interest, specifically including the following steps:

[0130] Step 401: Optimize the edge weights using a machine learning model trained based on a large amount of historical access data to generate optimized edge weights that can dynamically adapt to user behavior change patterns;

[0131] In this step, the machine learning model refers to an algorithm that automatically improves and adjusts parameters through training data, and can make predictions or decisions on new data; historical visit data refers to information such as the time and frequency of users' past visits to different points of interest, which is used to train the model; edge weight optimization processing refers to adjusting the weights of edges in the graph through the machine learning model to make it more accurately reflect user behavior patterns;

[0132] The edge weights in the undirected graph are optimized using a machine learning model trained based on a large amount of historical access data. The model can dynamically adapt to the changing patterns of user behavior, ensuring that the edge weights not only reflect the user's current behavior, but also predict future trends. Ultimately, it generates optimized edge weights that can dynamically adapt to the changing patterns of user behavior, providing a more accurate data foundation for subsequent steps.

[0133] Step 402: According to the optimized edge weights and the timestamp information of the user visiting the points of interest, a direction determination process is performed on the edges between each pair of points of interest in the undirected graph to obtain a direction definition reflecting the actual time sequence of the user's visits;

[0134] In this step, the timestamp information refers to recording the specific time when the user visits each point of interest; the direction determination process refers to assigning a direction to each edge in the graph, indicating the time sequence of the user moving from one point of interest to another point of interest;

[0135] According to the optimized edge weights and the timestamp information of the user's visit to the POI, the directionality of the edges between each pair of POIs in the undirected graph is determined. By analyzing the timestamp information, the actual movement path and order of the user from one POI to another is determined, and each edge is given a clear direction. The direction definition that reflects the actual time sequence of the user's visit is obtained, making the graph structure more in line with the user's mobility pattern.

[0136] Step 403: using the directionality definition to convert the undirected graph to generate an initial weighted directed graph;

[0137] In this step, the conversion process refers to the process of converting an undirected graph into a directed graph, including assigning edge directions and updating weights; the initial weighted directed graph refers to the graphical representation with directions and optimized weights after the conversion process, which preliminarily reflects the user's behavior pattern;

[0138] Using the directionality definition determined in step 402, the undirected graph is converted into a weighted directed graph, in which each edge has a clear direction and an optimized weight, to generate an initial weighted directed graph, which not only reflects the time sequence of users' visits to points of interest, but also reflects the importance of visit frequency and dwell time through edge weights, providing structured data support for further analysis.

[0139] Step 404: Based on an anomaly detection algorithm, the paths in the weighted directed graph that do not conform to the user's regular behavior pattern are identified and excluded, so as to obtain an optimized weighted directed graph that can more realistically represent the user's daily mobility pattern;

[0140] In this step, an anomaly detection algorithm refers to an algorithm for identifying data points or paths in a data set that do not conform to a regular pattern; optimizing a weighted directed graph refers to removing abnormal paths to obtain a weighted directed graph that more realistically represents the user's daily mobility pattern;

[0141] Based on the anomaly detection algorithm, the paths in the weighted directed graph generated in step 403 that do not conform to the user's regular behavior pattern are identified and excluded. Abnormal paths may include unusual movements or erroneous data points due to accidental factors. These paths will interfere with the accurate analysis of the user's behavior pattern. By removing these abnormal paths, an optimized weighted directed graph that can more realistically represent the user's daily movement pattern is obtained, ensuring the reliability and accuracy of subsequent analysis and recommendation results.

[0142] Based on this, the present invention provides a specific embodiment, wherein step 102, based on the real-time location information, combined with the timestamp and speed vector information, applies a trajectory prediction algorithm to predict the user's movement trend and generates a list of potential interest areas, specifically including the following steps:

[0143] Step 501: Reconstruct the user's historical movement path using the real-time location information processed by the Kalman filter algorithm in combination with the timestamp and velocity vector information to obtain a continuous spatiotemporal trajectory. Based on the continuous spatiotemporal trajectory, use a multi-step recursive neural network to perform short-term and medium- to long-term predictions on the user's future movement path to obtain multiple predicted movement trend results.

[0144] In this step, Kalman filter algorithm refers to a recursive optimal estimation algorithm used to estimate the state of a process from a series of incomplete and noisy measurements; real-time location information refers to the user's current location data obtained by GPS or other means; timestamp and velocity vector information refers to the time when an event occurs and the speed and direction of the user's movement; multi-step recurrent neural network (RNN) refers to a deep learning model that can process sequence data and make predictions for multiple time points in the future;

[0145] The real-time location information processed by the Kalman filter algorithm is combined with the timestamp and speed vector information to reconstruct the user's historical movement path to obtain a continuous space-time trajectory. Based on the continuous space-time trajectory, a multi-step recursive neural network (such as LSTM or GRU) is used to perform short-term and medium- to long-term predictions on the user's future movement path. This process generates multiple predicted movement trend results, covering the user's possible movement paths in different time periods, providing a rich prediction basis for subsequent steps.

[0146] Step 502: introducing external environment data, and adjusting the movement trend result using a context-aware model to obtain an optimized movement trend prediction;

[0147] In this step, external environmental data refers to external factors that affect user behavior, including traffic flow, weather conditions, public events, etc.; context-aware model refers to a model that considers various factors of the current situation and adjusts the prediction results;

[0148] External environmental data is introduced, and the mobile trend result obtained in step 501 is adjusted using the context-aware model. The context-aware model will dynamically adjust the prediction result according to factors such as the current time, weather conditions, and traffic conditions to ensure that it is more in line with the actual situation. Ultimately, an optimized mobile trend prediction is obtained, which improves the accuracy and practicality of the prediction result.

[0149] Step 503: Based on the optimized movement trend prediction, the geographical location of the user is predicted and processed in combination with the interest point database in the geographic information system to generate a preliminary interest area list;

[0150] In this step, Geographic Information System (GIS) refers to a system used to capture, store, manipulate, analyze, manage and display all types of geographic data; Point of Interest Database refers to a database containing information about various points of interest (POIs), such as restaurants, shops, attractions, etc.;

[0151] Based on the optimized mobile trend prediction and combined with the POI database in the geographic information system, the geographical locations that users may reach in the future are predicted, these geographical locations are analyzed, potential POIs are identified, and a preliminary list of areas of interest is generated. This step lays the foundation for subsequent POI screening and recommendation.

[0152] Step 504: applying a spatiotemporal density clustering algorithm to optimize the preliminary interest region list based on the geographical distribution characteristics near the predicted path, and obtaining a cluster-optimized interest region list;

[0153] In this step, the spatiotemporal density clustering algorithm refers to a density clustering algorithm based on time and space dimensions, such as DBSCAN or OPTICS, which is used to identify dense groups of data points; geographic distribution features refer to the spatial distribution of points of interest in a certain area, such as density, distance, etc.;

[0154] The spatiotemporal density clustering algorithm is applied to optimize the preliminary list of areas of interest based on the geographical distribution characteristics near the predicted path. Through cluster analysis, low-likelihood points of interest are removed and high-probability visit areas are retained, thus obtaining a cluster-optimized list of areas of interest. This process not only reduces unnecessary points of interest, but also improves the relevance and accuracy of the recommendation results.

[0155] Step 505: Apply the personalized preference model to filter the cluster-optimized interest area list, give priority to recommending interest points that meet the user's preferences, and generate a personalized potential interest area list;

[0156] In this step, the personalized preference model refers to a model trained from the user's historical behavior data through a machine learning algorithm and used to predict user preferences; the personalized potential interest area list refers to the final recommended interest point list generated after screening according to user preferences;

[0157] A personalized preference model is applied to filter the cluster-optimized list of interest areas, and points of interest that meet user preferences are recommended first. This model is based on the user's historical behavior patterns and immediate needs, ensuring that the recommended points of interest are highly personalized and closer to the user's actual needs. Ultimately, a personalized list of potential interest areas is generated to provide users with accurate and personalized service recommendations.

[0158] Based on this, the present invention provides a specific embodiment, wherein the step 105 uses a weighted K-nearest neighbor algorithm to sort multiple service provision points in the service provision point set to obtain a service recommendation list, and uses a multimodal fusion algorithm to combine the geographical location information in the service recommendation list with other sensor data to generate a customized information package, which specifically includes the following steps:

[0159] Step 701: using a weighted K-nearest neighbor algorithm, based on the user interest model, real-time location information, and the geographical distance, visit frequency, stay time, and score of the service provider point, a plurality of service providers in the service provider point set are comprehensively evaluated and sorted to obtain a personalized service recommendation list;

[0160] In this step, the weighted K nearest neighbor algorithm (WKNN) refers to a classification or regression algorithm based on distance measurement, in which the weight of each neighbor is determined by its distance; the user interest model refers to a model built based on the user's behavior pattern and preferred location type, which is used to predict the user's interests and needs; real-time location information: the user's geographic location data at the current moment; the service provider point set refers to a set of locations or facilities that may provide services to users;

[0161] The weighted K-nearest neighbor algorithm is used to comprehensively evaluate and sort multiple service providers in the service provider set based on the user interest model, real-time location information, and multiple factors such as the geographical distance, visit frequency, residence time and rating of the service provider. Through this multi-dimensional evaluation, a personalized service recommendation list is generated to ensure that the recommendation results not only meet the user's preferences, but also take into account the actual geographical location and the characteristics of the service provider.

[0162] Step 702: Introduce a context-aware module, combine external environment data, dynamically adjust the personalized service recommendation list, and generate an optimized service recommendation list;

[0163] In this step, the context-aware module refers to a system component that can recognize and respond to environmental changes, and can adjust the recommendation results by considering external environmental data; external environmental data refers to external factors that affect user behavior, such as traffic flow, weather conditions, and public events;

[0164] A context-aware module is introduced to dynamically adjust the personalized service recommendation list generated in step 701 in combination with external environmental data (such as traffic conditions, weather conditions, public events, etc.). The context-aware module will automatically adjust the content and order of the recommendation list according to the current time, location and environmental changes to ensure that the recommendation results are more in line with the user's immediate needs and changes in the surrounding environment.

[0165] Finally, an optimized service recommendation list is generated, which improves the timeliness and relevance of the recommendation results.

[0166] Step 703: Based on the optimized service recommendation list, a multimodal fusion algorithm is applied to integrate data from different sources, perform multi-dimensional feature description processing on each recommended service provider, and generate a service information card;

[0167] In this step, the multimodal fusion algorithm refers to the algorithm that integrates information from different sensors or data sources to improve the quality of decision-making; the multidimensional feature description processing refers to describing the characteristics of the object from multiple angles, such as geographic location, image, sound, temperature and humidity, etc.;

[0168] Based on the optimized service recommendation list, a multimodal fusion algorithm is applied to integrate data from different sources (such as geographic location information, image recognition results, sound analysis, temperature and humidity sensor data), and perform multi-dimensional feature description processing on each recommended service provider. This process generates detailed service information cards. Each card contains rich information about the service provider, allowing users to fully understand the specific situation of each recommended point and make more informed choices.

[0169] Step 704: using natural language processing technology to perform semantic understanding and sentiment analysis on the text content in the service information card, extract key information and optimize the expression method to obtain a content display form;

[0170] In this step, natural language processing (NLP) technology refers to the technology involving the interaction between computers and human language, which is used for understanding and generating text content; semantic understanding refers to parsing the meaning of text content and extracting key information; sentiment analysis refers to evaluating the emotional tendency expressed in the text, such as positive, negative or neutral;

[0171] Natural language processing technology is used to perform semantic understanding and sentiment analysis on the text content in service information cards, extract key information, and optimize the expression method, making the information display more concise and clear, and easy for users to understand. The optimized content display format not only improves the user experience, but also enhances the effective transmission of information.

[0172] Step 705: Integrate the content display forms into a comprehensive information structure, and generate a customized information package according to the user's behavior pattern;

[0173] In this step, the comprehensive information structure: organize different types of information into a whole to ensure the coherence and completeness of the information.

[0174] Customized information package: A comprehensive information service that combines multiple types of information according to the user's needs and preferences;

[0175] Integrate the content display format obtained in step 704 into a comprehensive information structure to ensure that all recommended content is logically clear and coherent. Generate a customized information package based on the user's behavior patterns and historical preferences. This information package not only contains accurate service recommendations, but also integrates other types of rich information to provide users with a comprehensive and personalized service experience. The customized information package finally generated is designed to provide highly personalized, information-rich and easy-to-understand content, significantly improving user satisfaction and service quality.

[0176] Figure 2 A schematic diagram of a big data processing system for implementing hybrid data analysis is provided for an embodiment of the present invention. Figure 2 As shown, the system includes:

[0177] The receiving module 21 is used to receive the location data updated by the high-frequency GPS coordinates from the user's mobile device in real time, and process the location data using the Kalman filter algorithm to obtain real-time location information;

[0178] A prediction module 22, configured to predict the user's movement trend based on the real-time location information, in combination with the timestamp and the speed vector information, and generate a list of potential interest areas by applying a trajectory prediction algorithm;

[0179] An identification module 23 is used to analyze the potential interest area list and the user's historical location data using a spatial clustering analysis algorithm to identify the user's behavior pattern and preferred location type, and to construct a user interest model based on the user's behavior pattern and preferred location type;

[0180] A configuration module 24, configured to configure a dynamic adaptive geo-fence system using the user interest model, wherein the dynamic adaptive geo-fence system is configured to automatically adjust the fence boundary according to the user's real-time location, behavior pattern, and surrounding environment changes to generate a set of service provision points;

[0181] The generating module 25 is used to sort the multiple service providing points in the service providing point set by using a weighted K-nearest neighbor algorithm to obtain a service recommendation list, and combine the geographical location information in the service recommendation list with other sensor data by using a multimodal fusion algorithm to generate a customized information package.

[0182] Figure 2 The information push system based on big data matching and GPS positioning can be executed Figure 1 The implementation principle and technical effect of the information push method based on big data matching and GPS positioning described in the illustrated embodiment will not be repeated. The specific way in which each module and unit performs operations in the information push system based on big data matching and GPS positioning in the above embodiment has been described in detail in the embodiment of the method, and will not be elaborated here.

[0183] In one possible design, Figure 2 The information push system based on big data matching and GPS positioning in the embodiment shown can be implemented as a computing device, such as Figure 3 As shown, the computing device may include a storage component 31 and a processing component 32;

[0184] The storage component 31 stores one or more computer instructions, wherein the one or more computer instructions are called and executed by the processing component 32 .

[0185] The processing component 32 is used to: receive in real time the location data of high-frequency GPS coordinate updates from the user's mobile device, and process the location data using a Kalman filter algorithm to obtain real-time location information; based on the real-time location information, combined with timestamp and velocity vector information, apply a trajectory prediction algorithm to predict the user's movement trend and generate a list of potential interest areas; use a spatial clustering analysis algorithm to analyze the list of potential interest areas and the user's historical location data to identify the user's behavior pattern and preferred location type, and build a user interest model based on the user's behavior pattern and preferred location type; use the user interest model to configure a dynamic adaptive geographic fence system, the dynamic adaptive geographic fence system is used to automatically adjust the fence boundary according to the user's real-time location, behavior pattern and changes in the surrounding environment to generate a set of service provider points; use a weighted K nearest neighbor algorithm to sort multiple service provider points in the set of service provider points to obtain a service recommendation list, and use a multimodal fusion algorithm to combine the geographic location information in the service recommendation list with other sensor data to generate a customized information package.

[0186] The processing component 32 may include one or more processors to execute computer instructions to complete all or part of the steps in the above method. Of course, the processing component may also be implemented by one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors or other electronic components to perform the above method.

[0187] The storage component 31 is configured to store various types of data to support operations at the terminal. The storage component can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk or optical disk.

[0188] Of course, the computing device may also include other components, such as input / output interfaces, display components, communication components, etc.

[0189] The input / output interface provides an interface between the processing component and the peripheral interface module, which may be an output device, an input device, etc.

[0190] The communication component is configured to facilitate, among other things, wired or wireless communications between the computing device and other devices.

[0191] Among them, the computing device can be a physical device or an elastic computing host provided by a cloud computing platform, etc. In this case, the computing device can refer to a cloud server, and the above-mentioned processing components, storage components, etc. can be basic server resources rented or purchased from the cloud computing platform.

[0192] The embodiment of the present invention further provides a computer storage medium storing a computer program, which can achieve the above-mentioned Figure 1 The illustrated embodiment is an information push method based on big data matching and GPS positioning.

[0193] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0194] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative labor.

[0195] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0196] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. An information push method based on big data matching and GPS positioning, characterized in that: include: Receive in real time the location data of high-frequency GPS coordinate updates from the user's mobile device, and process the location data using a Kalman filter algorithm to obtain real-time location information; Based on the real-time location information, combined with timestamp and velocity vector information, a trajectory prediction algorithm is applied to predict the user's movement trend and generate a list of potential areas of interest; Using a spatial clustering analysis algorithm to analyze the potential interest area list and the user's historical location data, identify the user's behavior pattern and preferred location type, and build a user interest model based on the user's behavior pattern and preferred location type; Utilizing the user interest model to configure a dynamic adaptive geo-fence system, the dynamic adaptive geo-fence system is used to automatically adjust the fence boundary according to the user's real-time location, behavior pattern, and surrounding environment changes to generate a set of service provision points; A weighted K-nearest neighbor algorithm is used to sort multiple service provision points in the service provision point set to obtain a service recommendation list, and a multimodal fusion algorithm is used to combine the geographical location information in the service recommendation list with other sensor data to generate a customized information package.

2. The method according to claim 1, characterized in that The potential interest area list and the user's historical location data are analyzed using a spatial clustering analysis algorithm to identify the user's behavior pattern and preferred location type, and a user interest model is constructed according to the user's behavior pattern and preferred location type, including: The user's historical location data is denoised using a spatiotemporal consistency check algorithm to remove noise points caused by GPS signal drift, and missing values ​​are filled in using a time series filling algorithm to unify the timestamp format to obtain a complete and consistent user location record. Based on the user location record and in combination with timestamp information, the user's visit behavior is extracted, the user's stay records at different points of interest as well as the arrival and departure times are identified, and the user's visit behavior sequence is generated; According to the user access behavior sequence and the potential interest area list, a graph matching algorithm is used to convert the user access behavior sequence into a weighted directed graph, wherein nodes represent interest points, and edge weights represent access frequency and residence time. Through the weighted directed graph, the correlation between the user access behavior sequence and the potential interest area list is established, and the access probability matrix of each interest point is calculated; By using the access probability matrix, adopting the hierarchical density peak clustering algorithm and combining the geo-fencing technology, the geographical location, access frequency and residence time are calculated, and the points of interest visited by the user are grouped. The clustering parameters are automatically adjusted through the dynamic threshold adjustment mechanism to obtain the clustering results reflecting the user's frequented places and behavior patterns; Based on the clustering results, a hidden Markov model is applied to analyze the user's mobility state transition rules in different time periods to identify the user's behavior pattern, and semantic clustering is performed on different types of points of interest to identify the preferred location type; Based on the user behavior patterns and preferred place types, the geographical distance, visit frequency, dwell time and thematic classification of points of interest are comprehensively calculated, a comprehensive score is calculated for each point of interest, and a multi-dimensional user interest model is constructed.

3. The method according to claim 2, characterized in that According to the user access behavior sequence, combined with the potential interest area list, the user access behavior sequence is converted into a weighted directed graph using a graph matching algorithm, wherein nodes represent interest points, and edge weights represent access frequency and residence time. Through the weighted directed graph, the correlation between the user access behavior sequence and the potential interest area list is established, and the access probability matrix of each interest point is calculated, including: The spatiotemporal consistency check algorithm is used to process the temporal and spatial coherence of the movement paths in the user's access behavior sequence, and an undirected graph representing the user's movement pattern is obtained, in which the nodes represent different points of interest; According to the user's stay time and visit frequency at each point of interest, combined with the edge weights optimized by the machine learning model, the undirected graph is weighted and directed to generate a weighted directed graph reflecting the time sequence of the user's visit to the points of interest; Based on the weighted directed graph, the user access behavior sequence is compared with the information of the points of interest in the list of potential interest areas, and a context-aware mechanism is introduced to establish the correlation between the two, so as to obtain the relevance of the adjusted points of interest; For each point of interest, based on the relevance of the adjusted point of interest, a Bayesian network is used to perform probabilistic reasoning processing to calculate the probability of the point of interest being visited, thereby forming a visit probability matrix.

4. The method according to claim 3, characterized in that According to the user's stay time and visit frequency at each point of interest, combined with the edge weights optimized by the machine learning model, the undirected graph is weighted and directed to generate a weighted directed graph reflecting the time sequence of the user's visit to the points of interest, including: Optimizing the edge weights using a machine learning model trained based on a large amount of historical access data to generate optimized edge weights that can dynamically adapt to changing patterns of user behavior; According to the optimized edge weights and the timestamp information of the user visiting the points of interest, a direction determination process is performed on the edges between each pair of points of interest in the undirected graph to obtain a direction definition reflecting the actual time sequence of the user's visits; The undirected graph is converted using the directionality definition to generate an initial weighted directed graph; Based on the anomaly detection algorithm, the paths in the weighted directed graph that do not conform to the user's regular behavior pattern are identified and excluded, so as to obtain an optimized weighted directed graph that can more realistically represent the user's daily mobility pattern.

5. The method according to claim 1, characterized in that According to the real-time location information, combined with the timestamp and speed vector information, a trajectory prediction algorithm is applied to predict the user's movement trend and generate a list of potential interest areas, including: The real-time location information processed by the Kalman filter algorithm is combined with the timestamp and speed vector information to reconstruct the user's historical movement path to obtain a continuous space-time trajectory. Based on the continuous space-time trajectory, a multi-step recursive neural network is used to perform short-term and medium-term prediction processing on the user's future movement path to obtain multiple predicted movement trend results; Introducing external environment data, and using a context-aware model to adjust the movement trend result to obtain an optimized movement trend prediction; Based on the optimized mobile trend prediction, the user's geographical location is predicted and processed in combination with the interest point database in the geographic information system to generate a preliminary list of interest areas; Applying a spatiotemporal density clustering algorithm, based on the geographical distribution characteristics near the predicted path, the preliminary list of regions of interest is optimized to obtain a cluster-optimized list of regions of interest; The personalized preference model is applied to filter the cluster-optimized ROI list, prioritize the POIs that meet the user's preferences, and generate a personalized list of potential ROIs.

6. The method according to claim 1, characterized in that The user interest model is used to configure a dynamic adaptive geo-fence system, wherein the dynamic adaptive geo-fence system is used to automatically adjust the fence boundary according to the user's real-time location, behavior pattern, and surrounding environment changes to generate a set of service provision points, including: The user interest model is used to comprehensively analyze the user's real-time location information, behavior patterns, and preferred location types to determine the current user's behavior status and potential destinations; Based on the user's real-time location information and movement trend results, combined with timestamp information and speed vector information, a spatiotemporal prediction algorithm is used to estimate the possible arrival location and obtain an estimated location set; Combining the user's behavior patterns, preferred location types, and environmental influencing factors, the initial boundaries of the geo-fence are set and processed through a machine learning optimization algorithm to obtain the initial geo-fence boundaries; Introducing a dynamic threshold adjustment mechanism to automatically adjust the initial geo-fence boundary based on real-time feedback from users and new location data to generate an optimized geo-fence boundary; A context-aware recommendation system is applied to screen and sort the service provision points within the optimized geographic fence in combination with the user's real-time needs, historical preferences and actual availability of the service provision points, thereby generating an optimized set of service provision points.

7. The method according to claim 1, characterized in that A weighted K-nearest neighbor algorithm is used to sort multiple service provision points in the service provision point set to obtain a service recommendation list, and a multimodal fusion algorithm is used to combine the geographical location information in the service recommendation list with other sensor data to generate a customized information package, including: Using a weighted K-nearest neighbor algorithm, based on the user interest model, real-time location information, and the geographical distance, visit frequency, stay time, and score of the service provider point, a plurality of service providers in the service provider point set are comprehensively evaluated and ranked to obtain a personalized service recommendation list; Introducing a context-aware module, combined with external environment data, dynamically adjusts the personalized service recommendation list to generate an optimized service recommendation list; Based on the optimized service recommendation list, a multimodal fusion algorithm is applied to integrate data from different sources, perform multi-dimensional feature description processing on each recommended service provider, and generate service information cards; Using natural language processing technology to perform semantic understanding and sentiment analysis on the text content in the service information card, extract key information and optimize the expression method to obtain the content display form; The content presentation forms are integrated into a comprehensive information structure, and customized information packages are generated according to user behavior patterns.

8. An information push system based on big data matching and GPS positioning, characterized in that: include: A receiving module is used to receive the location data updated by the high-frequency GPS coordinates from the user's mobile device in real time, and process the location data using a Kalman filter algorithm to obtain real-time location information; A prediction module, for predicting the user's movement trend based on the real-time location information, in combination with the timestamp and the speed vector information, and generating a list of potential interest areas by applying a trajectory prediction algorithm; An identification module, used to analyze the potential interest area list and the user's historical location data using a spatial clustering analysis algorithm, identify the user's behavior pattern and preferred location type, and build a user interest model based on the user's behavior pattern and preferred location type; A configuration module, configured to configure a dynamic adaptive geo-fence system using the user interest model, wherein the dynamic adaptive geo-fence system is configured to automatically adjust the fence boundary according to the user's real-time location, behavior pattern, and surrounding environment changes to generate a set of service provision points; The generating module is used to sort the multiple service providing points in the service providing point set by using a weighted K-nearest neighbor algorithm to obtain a service recommendation list, and combine the geographical location information in the service recommendation list with other sensor data by using a multimodal fusion algorithm to generate a customized information package.

9. A computing device, characterized in that It comprises a processing component and a storage component; the storage component stores one or more computer instructions; the one or more computer instructions are used to be called and executed by the processing component to implement an information push method based on big data matching and GPS positioning as described in any one of claims 1 to 7.

10. A computer storage medium, characterized in that: A computer program is stored, and when the computer program is executed by a computer, an information push method based on big data matching and GPS positioning as described in any one of claims 1 to 7 is implemented.

Images