Local living resource optimization and allocation method and device and terminal equipment

Abstract

The embodiment of the invention provides a local life resource optimization and allocation method and device and terminal equipment, and belongs to the technical field of data analysis. The method comprises the steps of performing clustering according to initial consumption information and a user driving path to obtain a target clustering result; obtaining related label information and corresponding related frequency information involved in the first subclass cluster in the target clustering result; determining an initial correlation index of the correlation label information to the initial consumption label according to the correlation frequency information; performing time attenuation on the initial correlation index according to the consumption time of the related label information to obtain a target correlation index; performing label screening from the initial consumption labels according to the target correlation index to obtain an initial consumption type; obtaining a related driving path of the first subclass cluster from the user driving path, and performing information adjustment on the initial consumption type according to the related driving path to obtain a target consumption type; and optimizing the initial peripheral configuration information according to the target consumption type to obtain target peripheral configuration information.

Description

Technical Field

[0001] The present invention relates to the field of data analysis technology, and in particular to a method, device and terminal device for optimizing and allocating local life resources. Background Art

[0002] In the process of collecting user consumption information based on target objects (such as gas stations), existing technologies usually rely on a single data source, such as using only gas station consumption records to analyze user consumption behavior. However, this single-source data often cannot fully cover the user's consumption scenarios, which leads to the data being incomplete and detailed, and it is difficult to accurately reflect the user's real consumption preferences and behavior patterns. In addition, due to the lack of support from multi-dimensional data, the constructed consumption portraits or identified consumption types are often too one-sided, making it difficult to deeply analyze the user's consumption habits and potential needs. This limitation further leads to the inability to accurately allocate when optimizing and allocating life resources, thereby reducing the efficiency and effectiveness of resource allocation and making it difficult to achieve optimal resource utilization. Summary of the invention

[0003] The main purpose of the embodiments of the present invention is to provide a method, device and terminal equipment for optimizing and allocating local life resources, aiming to solve the problem in the related technology that accurate allocation cannot be performed when optimizing and allocating life resources, thereby reducing the efficiency and effect of resource allocation and making it difficult to achieve optimal resource utilization.

[0004] In a first aspect, an embodiment of the present invention provides a method for optimizing and allocating local life resources, including:

[0005] Determine a target object and obtain initial surrounding configuration information corresponding to the target object;

[0006] Obtaining initial consumption information corresponding to the target user under the target object and a user travel path corresponding to the target user;

[0007] Performing user clustering on the target user according to the initial consumption information and the user's travel path to obtain a target clustering result;

[0008] Obtain relevant label information involved in each first subclass cluster in the target clustering result, and obtain relevant frequency information corresponding to the relevant label information;

[0009] Determine an initial consumption tag, and determine an initial correlation index corresponding to the initial consumption tag by the correlation tag information according to the correlation frequency information;

[0010] Obtaining the consumption time corresponding to the relevant tag information, and performing time decay on the initial relevant index according to the consumption time to obtain a target relevant index;

[0011] According to the target-related index, label screening is performed from the initial consumption labels to obtain the initial consumption type corresponding to the first sub-category cluster;

[0012] Obtaining a related driving path corresponding to the first sub-class cluster from the user driving path, and adjusting information of the initial consumption type according to the related driving path to obtain a target consumption type;

[0013] The initial peripheral configuration information is resource optimized according to the target consumption type to obtain target peripheral configuration information.

[0014] In a second aspect, an embodiment of the present invention provides a local life resource optimization and allocation device, including:

[0015] A data acquisition module, used to determine a target object and obtain initial surrounding configuration information corresponding to the target object;

[0016] An information acquisition module, used to obtain the initial consumption information corresponding to the target user under the target object and the user travel path corresponding to the target user;

[0017] A data clustering module, used to cluster the target users according to the initial consumption information and the user travel path, and obtain a target clustering result;

[0018] A data determination module, used to obtain relevant label information involved in each first subclass cluster in the target clustering result, and obtain relevant frequency information corresponding to the relevant label information;

[0019] A data analysis module, used to determine an initial consumption tag, and determine an initial correlation index corresponding to the initial consumption tag by the correlation tag information according to the correlation frequency information;

[0020] A data adjustment module, used to obtain the consumption time corresponding to the relevant tag information, and perform time decay on the initial relevant index according to the consumption time to obtain a target relevant index;

[0021] A data screening module, configured to screen the initial consumption tags according to the target-related index to obtain the initial consumption type corresponding to the first sub-class cluster;

[0022] An information adjustment module, configured to obtain a related driving path corresponding to the first sub-class cluster from the user's driving path, and adjust the information of the initial consumption type according to the related driving path to obtain a target consumption type;

[0023] The resource optimization module is used to perform resource optimization on the initial peripheral configuration information according to the target consumption type to obtain target peripheral configuration information.

[0024] In the third aspect, an embodiment of the present invention further provides a terminal device, comprising a processor, a memory, a computer program stored in the memory and executable by the processor, and a data bus for realizing connection and communication between the processor and the memory, wherein when the computer program is executed by the processor, the steps of any one of the local life resource optimization and allocation methods provided in the specification of the present invention are implemented.

[0025] The embodiments of the present invention provide a method, apparatus and terminal device for optimizing and allocating local life resources. The method includes: determining a target object and obtaining initial surrounding configuration information corresponding to the target object; obtaining initial consumption information corresponding to a target user under the target object and a user driving path corresponding to the target user, thereby clustering the target users according to the initial consumption information and the user driving path to obtain target clustering results, and then clustering the users by combining the initial consumption information and the driving path of the users, thereby more comprehensively capturing the users' consumption behaviors and travel habits and avoiding the limitations of a single data source. Then obtain the relevant label information involved in each first subclass cluster in the target clustering result, and obtain the relevant frequency information corresponding to the relevant label information, and then determine the initial consumption label, and determine the initial correlation index corresponding to the relevant label information to the initial consumption label according to the relevant frequency information; obtain the consumption time corresponding to the relevant label information, and perform time decay on the initial correlation index according to the consumption time to obtain the target correlation index, and then perform label screening from the initial consumption label according to the target correlation index to obtain the initial consumption type corresponding to the first subclass cluster, and then obtain the initial consumption type corresponding to each first subclass cluster to avoid analysis errors caused by analyzing a single user, and then obtain the relevant driving path corresponding to the first subclass cluster from the user's driving path, and adjust the initial consumption type according to the relevant driving path to obtain the target consumption type; by analyzing the user's relevant driving path, the initial consumption type can be dynamically adjusted to ensure that the identification of the consumption type is more accurate and reflects the user's real needs, and finally optimize the initial surrounding configuration information according to the target consumption type to obtain the target surrounding configuration information. Therefore, by optimizing the surrounding configuration information according to the target consumption type, personalized allocation of resources can be achieved to meet the specific needs of different user groups under the target object. Through precise resource optimization, resource waste can be avoided, resource allocation can be ensured to be more reasonable, and overall resource utilization efficiency can be improved. The optimized surrounding configuration information can better match user needs, improve the service quality of the target object, enhance user stickiness and loyalty, and thus significantly improve the accuracy of resource allocation around the target object, while enhancing the market competitiveness of the target object, reducing operating costs, and achieving data-driven scientific decision-making and continuous optimization. This all-round analysis and optimization method based on user behavior and needs provides strong support for the intelligent allocation of life resources. It also solves the problem in related technologies that it is impossible to accurately allocate life resources during optimization and allocation, thereby reducing the efficiency and effect of resource allocation and making it difficult to achieve optimal resource utilization. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0027] Figure 1 A flowchart of a method for optimizing and allocating local living resources provided by an embodiment of the present invention;

[0028] Figure 2 A schematic diagram of the module structure of a local life resource optimization and allocation device provided by an embodiment of the present invention;

[0029] Figure 3 A schematic block diagram of the structure of a terminal device provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0030] 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 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 ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0031] The flowcharts shown in the accompanying drawings are only examples and do not necessarily include all the contents and operations / steps, nor must they be executed in the order described. For example, some operations / steps may also be decomposed, combined or partially merged, so the actual execution order may change according to actual conditions.

[0032] It should be understood that the terms used in this specification of the present invention are only for the purpose of describing specific embodiments and are not intended to limit the present invention. As used in the specification of the present invention and the appended claims, unless the context clearly indicates otherwise, the singular forms "a", "an" and "the" are intended to include plural forms.

[0033] The embodiment of the present invention provides a method, apparatus and terminal device for optimizing and allocating local life resources. The method can be applied to a terminal device, which can be an electronic device such as a tablet computer, a laptop computer, a desktop computer, a personal digital assistant and a wearable device. The terminal device can be a server or a server cluster.

[0034] Some embodiments of the present invention are described in detail below in conjunction with the accompanying drawings. In the absence of conflict, the following embodiments and features in the embodiments can be combined with each other.

[0035] Please refer to Figure 1 , Figure 1 A flowchart of a method for optimizing and allocating local life resources provided by an embodiment of the present invention.

[0036] like Figure 1 As shown, the local life resource optimization and allocation method includes steps S101 to S109.

[0037] Step S101: determine a target object and obtain initial surrounding configuration information corresponding to the target object.

[0038] Exemplarily, the target object may be a user consumption place such as a gas station, a supermarket, or a restaurant, and the initial surrounding configuration information configured around the target object is determined based on the location information corresponding to the target object. The initial surrounding configuration information includes but is not limited to information on car washes, convenience stores, supermarkets, restaurants, etc. around the target object. The initial surrounding configuration information is used to characterize related supporting facilities or stores that already exist around the target object.

[0039] Step S102: Obtain the initial consumption information corresponding to the target user under the target object and the user travel path corresponding to the target user.

[0040] Exemplarily, the initial consumption record of the target user is obtained from the management system of the target object. For example, when the target object is a gas station, the initial consumption record includes the refueling time, refueling amount, refueling type, consumption amount, etc., and the membership system of the gas station is used to record the consumption behavior and preferences of the target user, including membership card number, points record, etc.

[0041] Exemplarily, a target sensor deployed from the target object and initial surrounding configuration information obtains a relevant video corresponding to the target user, thereby identifying a target license plate corresponding to the target user in the relevant video and obtaining a user driving path corresponding to the target user from the relevant video.

[0042] In some embodiments, obtaining the user driving path corresponding to the target user includes: obtaining initial vehicle image information corresponding to a target sensor related to the target object; preprocessing the initial vehicle image information to obtain corresponding target vehicle image information; performing Fourier transform on the target vehicle image information after windowing to obtain target spectrum information corresponding to the target vehicle image information; performing information detection on the target spectrum information to obtain a blur direction and a blur length corresponding to the initial vehicle image information; performing image restoration processing on the target vehicle image information according to the blur direction and the blur length to obtain repaired image information corresponding to the target vehicle image information; performing license plate information recognition on the repaired image information to obtain predicted license plate information corresponding to the initial vehicle image information; obtaining the target license plate information corresponding to the target user, and comparing the predicted license plate information with the target license plate information to obtain the user driving path corresponding to the target user.

[0043] For example, a suitable target sensor, such as a camera, is selected and installed at a key position around the target object (such as a gas station) to ensure that the image of the vehicle can be captured. The sensor collects the image of the vehicle passing by in real time to obtain the initial vehicle image information. Then, image enhancement technology (such as histogram equalization, contrast adjustment, etc.) is used to improve the clarity and contrast of the image, and remove noise in the image, such as Gaussian noise, salt and pepper noise, etc., to obtain the target vehicle image information.

[0044] Exemplarily, a suitable window function (such as a Hanning window, a Hamming window, etc.) is selected to perform windowing processing on the target vehicle image information to reduce spectrum leakage, and then a two-dimensional Fourier transform is performed on the windowed image to convert the target vehicle image information from the spatial domain to the frequency domain to obtain the target spectrum information.

[0045] For example, the target spectrum information is analyzed to identify the main features in the spectrum, such as blur direction and blur length, so as to extract the blur direction and blur length in the target spectrum information for subsequent image restoration processing. Thus, according to the blur direction and blur length, the blur kernel (such as Gaussian blur kernel, motion blur kernel, etc.) is estimated, and then the image is restored using a deblurring algorithm (such as Wiener filtering, iterative deconvolution, etc.), and a clear image of the vehicle is restored to obtain repaired image information.

[0046] Exemplarily, a license plate positioning algorithm (such as edge detection, morphological operations, etc.) is used to determine the position of the license plate, and then the license plate area is extracted from the repaired image information to ensure that the extracted license plate area is clear and complete, and then a license plate recognition algorithm (such as OCR technology, deep learning models, etc.) is used to recognize the characters on the license plate to generate predicted license plate information.

[0047] Exemplarily, the target license plate information corresponding to the target user is obtained from the gas station membership system, and then the predicted license plate information is compared with the target license plate information. When the predicted license plate information is consistent with the target license plate information, the corresponding installation position of the target sensor is obtained, thereby simulating the user driving path corresponding to the target user according to the installation position.

[0048] Step S103: clustering the target users according to the initial consumption information and the user travel paths to obtain target clustering results.

[0049] Exemplarily, user consumption characteristics are extracted from the initial consumption information, such as consumption amount, consumption frequency, consumption time, etc. User travel characteristics are extracted from the user's driving path, such as frequently visited places, driving distance, travel time, etc.

[0050] Exemplarily, a suitable clustering algorithm is selected according to the characteristics of the data, such as K-means, hierarchical clustering, DBSCAN, etc., and then the parameters of the clustering algorithm are set, such as the K value (the number of clusters) in K-means, epsilon and min_samples in DBSCAN, etc., and then information clustering is performed on user consumption characteristics and user travel characteristics to obtain target clustering results.

[0051] In some embodiments, clustering the target users according to the initial consumption information and the user driving path to obtain a target clustering result includes: determining a user departure location and a user arrival location corresponding to the target user according to the user driving path; determining a target intent label corresponding to the target user and a first label probability corresponding to the target intent label according to the user departure location and the user arrival location; performing label identification according to the initial consumption information to obtain user label information corresponding to the target user and a second label probability corresponding to the user label information; clustering the target users according to the target intent label, the first label probability, the user label information and the second label probability to obtain an initial clustering result; calculating data correlation for each second subclass cluster in the initial clustering result to obtain a data correlation value between the cluster center and the cluster data in the second subclass cluster; and adjusting the initial clustering result according to the data correlation value to obtain the target clustering result corresponding to the target user.

[0052] For example, the user's driving path is analyzed, and the corresponding position when entering the monitoring range corresponding to the target object each time is identified as the user's departure position, and the corresponding position when leaving the monitoring range corresponding to the target object each time is determined as the user's arrival position. That is, the position corresponding to when the target user first appears in the video information is determined as the user's departure position, and the position corresponding to when the target user last appears in the video information is determined as the user's arrival position.

[0053] Exemplarily, based on the user departure location and user arrival location corresponding to the target user and the member registration information corresponding to the target user, such as residential location, the intention classification model is used to infer the target user's travel intention, such as commuting, shopping, leisure, etc., so as to obtain the target intention label corresponding to the target user, and calculate the corresponding probability, that is, the first label probability, which is used to represent the confidence of the target intention label.

[0054] Exemplarily, the initial consumption information of the target user is analyzed to identify consumption patterns and preferences, such as high-frequency consumption, specific types of consumption, etc. Therefore, the target user is labeled with corresponding labels according to the corresponding consumption behavior in the initial consumption information, such as "high-spending user", "oil product preference user", etc., and the probability of each label is calculated to obtain the second label probability.

[0055] Exemplarily, features such as the target intention label, the first label probability, the user label information and the second label probability are fused to form a behavior feature vector of the target user, and then a clustering algorithm (such as K-means, hierarchical clustering, etc.) is used to cluster the user feature vector to obtain an initial clustering result.

[0056] Exemplarily, the Euclidean distance or cosine similarity is used to calculate the correlation between the cluster center in each second subclass cluster and each cluster data to obtain the data correlation value between the cluster center and the cluster data in the second subclass cluster, and then the initial clustering result is adjusted according to the data correlation, such as merging similar clusters, splitting uneven clusters, etc., and then the final clustering result, i.e., the target clustering result, is determined to ensure the accuracy and rationality of the clustering.

[0057] Specifically, by combining the user's driving path and consumption information, a more accurate user portrait is constructed to gain a deeper understanding of user behavior and preferences. Based on the target clustering results, the resource allocation around the target object (such as a gas station) is optimized to improve resource utilization efficiency.

[0058] In some embodiments, clustering the target users according to the target intent label, the first label probability, the user label information and the second label probability to obtain an initial clustering result includes: obtaining a first user and a second user from the target users, and obtaining a first intent label corresponding to the first user and a third label probability corresponding to the first intent label from the target intent label and the first label probability, and obtaining a second intent label corresponding to the second user and a fourth label probability corresponding to the second intent label; obtaining a first label information corresponding to the first user and a fifth label probability corresponding to the first label information from the user label information and the second label probability, and obtaining a first label information corresponding to the second user and a fifth label probability corresponding to the first label information The method further comprises the steps of: determining a first user and a second user by performing intersection processing on the first intention label and the second intention label to obtain first intersection information corresponding to the first user and the second user; determining a user similarity between the first user and the second user by fusing the third label probability, the fourth label probability, the fifth label probability and the sixth label probability according to the first intersection information and the second intersection information; clustering the target users according to the user similarity to obtain the initial clustering result; wherein the user similarity is obtained according to the following formula:

[0059]

[0060] in, represents the user similarity between the first user and the second user, Indicates the number of tags corresponding to the first intent tag corresponding to the first user, represents the third label probability corresponding to the t-th first intention label of the first user; represents the number of tags corresponding to the first tag information corresponding to the first user, represents the fifth label probability corresponding to the mth first label information corresponding to the first user; represents the number of tags corresponding to the second intent tag corresponding to the second user, The fourth label probability corresponding to the kth second intention label of the second user; Indicates the number of tags corresponding to the second tag information corresponding to the second user; represents the sixth label probability corresponding to the r-th second label information of the second user; represents the number of intersections corresponding to the first intersection information corresponding to the first user and the second user, represents the label probability corresponding to the first user under the dth first intersection information; represents the label probability corresponding to the second user under the dth first intersection information; represents the number of intersections corresponding to the second intersection information corresponding to the first user and the second user, represents the label probability corresponding to the first user under the sth second intersection information; represents the label probability corresponding to the second user under the sth second intersection information.

[0061] Exemplarily, two users are selected from the target users, namely the first user and the second user, and then the first intent label and the third label probability corresponding to the first user are extracted from the target intent label and the first label probability, and the second intent label and the fourth label probability corresponding to the second user are extracted. Then, the first label information corresponding to the first user and the fifth label probability are extracted from the user label information and the second label probability, and the second label information and the sixth label probability are extracted from the second user.

[0062] Exemplarily, the first intention tag of the first user and the second intention tag of the second user are compared, and their common tags are found to be determined as the first intersection information. At the same time, the first tag information of the first user and the second tag information of the second user are compared, and their common tags are found to be determined as the second intersection information.

[0063] Exemplarily, the user similarity between the first user and the second user is determined by fusing the third label probability, the fourth label probability, the fifth label probability and the sixth label probability according to the first intersection information and the second intersection information using the following formula:

[0064]

[0065] in, represents the user similarity between the first user and the second user, Indicates the number of tags corresponding to the first intent tag corresponding to the first user, represents the third label probability corresponding to the t-th first intention label of the first user; Indicates the number of tags corresponding to the first tag information corresponding to the first user, represents the fifth label probability corresponding to the mth first label information corresponding to the first user; Indicates the number of tags corresponding to the second intent tag corresponding to the second user, represents the fourth label probability corresponding to the k-th second intention label of the second user; Indicates the number of tags corresponding to the second tag information corresponding to the second user; represents the sixth label probability corresponding to the r-th second label information of the second user; represents the number of intersections corresponding to the first intersection information corresponding to the first user and the second user, represents the label probability corresponding to the first user under the d-th first intersection information; represents the label probability corresponding to the second user under the d-th first intersection information; represents the number of intersections corresponding to the second intersection information corresponding to the first user and the second user, represents the label probability corresponding to the first user under the s-th second intersection information; Represents the label probability corresponding to the second user under the sth second intersection information.

[0066] Exemplarily, the similarity between any two users is calculated for all target users according to the above formula to form a similarity matrix, and then a clustering algorithm (such as hierarchical clustering, spectral clustering, etc.) is used to cluster the users according to the similarity matrix to obtain an initial clustering result, where each cluster represents a user group with a high similarity.

[0067] In some embodiments, the data correlation calculation is performed on each second sub-cluster in the initial clustering result to obtain the data correlation value between the cluster center and the cluster data in the second sub-cluster, including: performing center point calculation on the second sub-cluster to obtain the cluster center corresponding to the second sub-cluster; obtaining the third intent label corresponding to the cluster center and the first probability information corresponding to the third intent label from the target intent label; obtaining the third label information corresponding to the cluster center and the second probability information corresponding to the third label information from the user label information; obtaining the fourth intent label corresponding to the cluster data in the second sub-cluster and the third probability information corresponding to the fourth intent label from the target intent label; obtaining the fourth label information corresponding to the cluster data in the second sub-cluster and the fourth probability information corresponding to the fourth label information from the user label information; determining the distance information between the cluster center and the cluster data according to the third intent label, the first probability information, the third label information and the second probability information and the fourth intent label, the third probability information, the fourth label information and the fourth probability information; determining the data correlation value between the cluster center and the cluster data in the second sub-cluster according to the distance information.

[0068] Exemplarily, the average value calculation (such as the centroid calculation in K-means) or other center point calculation methods are used to calculate the center points of all data points in the second sub-cluster to obtain the cluster center, and then the intent label of the cluster center, i.e., the third intent label, is extracted from the target intent label, and the first probability information corresponding to the third intent label is extracted to indicate the confidence of the intent label. The label information of the cluster center, i.e., the third label information, is extracted from the user label information, and the second probability information corresponding to the third label information is extracted to indicate the confidence of the label.

[0069] Exemplarily, the intent label of each cluster data in the second sub-cluster is extracted from the target intent label, that is, the fourth intent label. The third probability information corresponding to the fourth intent label is extracted to indicate the confidence of the intent label. The label information of each cluster data in the second sub-cluster is extracted from the user label information, that is, the fourth label information. And the fourth probability information corresponding to the fourth label information is extracted to indicate the confidence of the label.

[0070] Exemplarily, the third intent label and the fourth intent label, as well as the third label information and the fourth label information are fused to form a comprehensive feature vector, and the first probability information, the second probability information, the third probability information and the fourth probability information are fused (such as weighted summation) to obtain a comprehensive probability value. Thus, based on the fused feature vector and probability value, the distance information between the cluster center and each cluster data is calculated using Euclidean distance and cosine similarity. Finally, based on the distance information, the data correlation value between the cluster center and the cluster data is calculated. For example, the smaller the distance, the larger the correlation value; the larger the distance, the smaller the correlation value.

[0071] Step S104: obtaining relevant label information involved in each first subclass cluster in the target clustering result, and obtaining relevant frequency information corresponding to the relevant label information.

[0072] Exemplarily, relevant features of relevant users are extracted from each first sub-class cluster, including consumption behavior, intention labels, location information, etc., and then the consumption records of each relevant user in the first sub-class cluster at the target object (such as a gas station) are analyzed, including consumption frequency, consumption amount, consumption time, etc., and the intention labels of relevant users are analyzed to understand the user's travel purpose, such as commuting, traveling, shopping, etc. Therefore, the patterns and trends in the user's consumption behavior are identified based on the consumption records and travel intentions corresponding to the relevant users in the first sub-class cluster, and then the consumption types corresponding to the relevant users in the first sub-class cluster are determined based on the consumption patterns and trends, such as daily consumption, travel consumption, business consumption, etc.

[0073] Exemplarily, the number and proportion of users of consumption types corresponding to relevant users in each first sub-category cluster are counted, and then based on the statistical results, the initial consumption type of each first sub-category cluster is determined, and the consumption type with a high proportion is selected as a representative.

[0074] Exemplarily, relevant tag information involved in user data in the first sub-category cluster is obtained from the target intent tag and user tag information, and then the frequency of occurrence of each relevant tag in the first sub-category cluster, ie, relevant frequency information, is counted.

[0075] Step S105: determine an initial consumption tag, and determine an initial correlation index corresponding to the initial consumption tag of the correlation tag information according to the correlation frequency information.

[0076] Exemplarily, possible initial consumption tags (such as daily consumption, travel consumption, business consumption, etc.) are determined based on business needs and user characteristics. For each initial consumption tag, the support of the relevant tag information to the initial consumption tag is calculated using the relevant frequency information combined with the weight information of the relevant tag information in the initial consumption tag to obtain the initial relevance index.

[0077] Step S106: Obtain the consumption time corresponding to the relevant tag information, and perform time decay on the initial relevant index according to the consumption time to obtain a target relevant index.

[0078] Exemplarily, the consumption time corresponding to each relevant tag is extracted from the user data, and then the initial relevant index is subjected to time decay processing according to the consumption time. For example, the target relevant index is equal to the initial relevant index multiplied by exp(-λ *Δt), where λ is the decay coefficient and Δt is the difference between the consumption time and the earliest consumption time.

[0079] Step S107: filtering the initial consumption tags according to the target relevance index to obtain the initial consumption type corresponding to the first sub-class cluster.

[0080] Exemplarily, the target relevance indexes of all initial consumption tags are sorted, tags with higher relevance indexes are screened out, and thresholds are set according to business needs. Initial consumption tags with target relevance indexes higher than the threshold are retained, and tags with target relevance indexes lower than the threshold are removed. Then, based on the screened initial consumption tags, the initial consumption type of the first subclass cluster is determined.

[0081] In some embodiments, performing time decay on the initial correlation index according to the consumption time to obtain a target correlation index includes: obtaining the earliest behavior time and the latest behavior time from the consumption time; determining a time decay parameter corresponding to the consumption time according to the earliest behavior time and the latest behavior time; performing time decay on the initial correlation index according to the time decay parameter to obtain the target correlation index; wherein the target correlation index is obtained according to the following formula:

[0082]

[0083] in, Indicates the target relevance index corresponding to the i-th initial consumption tag of the relevant tag information; represents the consumption time corresponding to the j-th related tag information, represents the earliest behavior time, Indicates the latest behavior time. , represents the adjustment parameter determined by the initial consumption tag; Indicates the number of tags corresponding to the relevant tag information, represents the label weight corresponding to the jth related label information under the ith initial consumption label, Represents the relevant frequency information corresponding to the j-th relevant label information.

[0084] Exemplarily, a target relevance index is determined based on a mechanism of time decay and related frequency information to optimize the identification of initial consumption types, taking into account the impact of time factors on user behavior, and adjusting the relevance index through a time decay parameter so that recent consumption behaviors have a greater impact on initial consumption types, while earlier behaviors have a smaller impact.

[0085] Exemplarily, relevant tag information and its corresponding consumption time are extracted from the user data in the first subclass cluster. The earliest behavior time and the latest behavior time in the consumption time are determined. These time points are used to calculate a time decay parameter, which is used to measure the decay degree of each consumption time relative to the latest behavior time.

[0086] Exemplarily, the time decay parameter is used to adjust the initial correlation index to obtain the target correlation index. The target correlation index combines the relevant frequency information and the time decay factor to more accurately reflect the current consumption tendency of users in the first sub-class cluster. Thus, the initial consumption tags are sorted and filtered according to the target correlation index, and the initial correlation index is time decayed according to the time decay parameter to obtain the target correlation index;

[0087] Exemplarily, the target correlation index is obtained by performing time decay on the initial correlation index according to the time decay parameter using the following formula:

[0088]

[0089] in, Represents the target relevance index of the relevant tag information corresponding to the i-th initial consumption tag; Indicates the consumption time corresponding to the jth related tag information, represents the earliest behavior time, Indicates the latest behavior time. , Indicates the adjustment parameters determined by the initial consumption tag; Indicates the number of tags corresponding to the relevant tag information, Indicates the label weight corresponding to the jth related label information under the i-th initial consumption label, Represents the relevant frequency information corresponding to the j-th relevant label information.

[0090] For example, all relevant tag information and the consumption time corresponding to these tags are extracted from the user data of the first sub-class cluster. The consumption time is the time point when the user performs the behavior related to the tag. Among all the consumption times, find the earliest behavior time and the latest behavior time, and then To determine the time decay parameter, the weight of each relevant label information and related frequency information It is fused with the time decay parameter to obtain the target correlation index.

[0091] Exemplarily, the above formula can consider the influence of different related users in the first sub-class cluster at different consumption times on the calculation of the target related index, so that the introduction of the time decay parameter has a significant impact on the definition of the consumption type, thereby more accurately determining the initial consumption type of the first sub-class cluster. This makes the identification of the consumption type closer to the current state and preference of the first sub-class cluster, which helps to provide more accurate personalized services and resource optimization.

[0092] Step S108: obtaining a related driving path corresponding to the first subclass cluster from the user driving path, and adjusting information of the initial consumption type according to the related driving path to obtain a target consumption type.

[0093] Exemplarily, relevant driving paths related to the first subclass cluster are extracted from user driving data, and then relevant driving paths in the same first subclass cluster are integrated to form a typical driving path of the first subclass cluster (such as the most common path, high-frequency path, etc.).

[0094] Exemplarily, key features are extracted from typical driving paths, such as path length, type of area passed (such as commercial areas, residential areas, industrial areas, etc.), path congestion, distance between the path and the target object, etc., and then the correlation between the driving path characteristics and the initial consumption type is analyzed. For example, a path passing through a commercial area may be related to "shopping consumption", a path passing through a gas station may be related to "refueling consumption", and a path length may be related to "long-distance travel consumption". Therefore, a weight is assigned to each relevant driving path according to the importance of the path characteristics (such as path frequency, path length, matching degree between the path and the target object, etc.). According to the characteristics and weights of the relevant driving paths, the adjustment factor of the initial consumption type is calculated.

[0095] For example, if a route passes through multiple gas stations and the user travels frequently, the weight of the "gas consumption" type may be increased; if a route passes through a commercial area and the user stays for a long time, the weight of the "shopping consumption" type may be increased.

[0096] Exemplarily, the initial consumption types are weighted and adjusted according to the adjustment factor. The weight of each initial consumption type is multiplied by the adjustment factor to update the weight value of the initial consumption type. The adjusted initial consumption type weights are sorted to screen out the consumption type with the highest weight, and then the consumption type with the highest weight is selected as the target consumption type.

[0097] In some embodiments, adjusting the information of the initial consumption type according to the relevant driving path to obtain the target consumption type includes: performing path prediction according to the relevant driving path to obtain the driving destination corresponding to the relevant users in the first subclass cluster; determining the user consumption intention corresponding to the relevant users in the first subclass cluster according to the driving destination; and adjusting the information of the initial consumption type according to the user consumption intention to obtain the target consumption type.

[0098] For example, historical driving route data is used to build a prediction model, such as a machine learning-based classification model, a sequence model (such as LSTM, GRU), or a deep learning model. Features are extracted from the relevant driving routes, including starting points, waypoints, route length, driving time, frequently visited places, route frequency, etc. The extracted features are input into the prediction model to predict the user's next possible driving destination.

[0099] For example, a mapping relationship between the driving destination and the consumption type is established, such as shopping malls corresponding to shopping consumption, gas stations corresponding to refueling consumption, etc. According to the predicted destination, it is mapped to the corresponding consumption intention. At the same time, the historical consumption data of the destination is considered, such as the consumption frequency and consumption amount of the user at the location. Thus, the weight of the consumption intention is calculated according to the consumption intensity of the destination and the historical consumption behavior of the user.

[0100] Exemplarily, the initial consumption type is compared with the user's consumption intention to identify the difference and consistency between the two. According to the weight of the user's consumption intention, the adjustment factor for the initial consumption type is calculated. The initial consumption type is combined with the adjustment factor by weighted average, maximum weight selection, etc. to obtain the target consumption type.

[0101] Step S109: Optimize the resources of the initial peripheral configuration information according to the target consumption type to obtain target peripheral configuration information.

[0102] For example, the specific requirements of the target consumption type for the surrounding configuration, such as the type, quantity, layout, etc. of facilities, are determined. And according to the importance and urgency of the user's needs, the requirements of the surrounding configuration are prioritized, and then the requirements of the target consumption type are compared with the actual situation of the initial surrounding configuration information, and the existing gaps and deficiencies are identified. Then, according to the demand analysis, it is decided to add or reduce certain facilities, and optimize the location layout of the facilities to make it more in line with the user's flow and behavior habits, or to upgrade or transform the initial surrounding configuration information to better meet the user's needs, so as to obtain the target surrounding configuration information.

[0103] For example, if the target consumption type is "shopping consumption", the initial surrounding configuration information includes a small number of parking lots, several convenience stores and some dining spots. Users of the "shopping consumption" type may need more parking spaces, more shopping places, rest areas, children's play areas, etc. Through gap analysis, it can be known that the initial surrounding configuration information has insufficient parking lots to meet the demand during peak hours, a single type of shopping places, lack of diversity, and a lack of rest areas and children's play areas, which cannot meet the needs of family users. The optimization strategy is to increase the number of parking lots, expand the parking area, introduce more shopping brands, increase shopping options, set up rest areas and children's play areas, and improve the shopping experience, so as to further optimize resources and obtain the target surrounding configuration information.

[0104] See also Figure 2 , Figure 2A local life resource optimization and allocation device 200 is provided for an embodiment of the present application, and the local life resource optimization and allocation device 200 includes a data acquisition module 201, an information acquisition module 202, a data clustering module 203, a data determination module 204, a data analysis module 205, a data adjustment module 206, a data screening module 207, an information adjustment module 208, and a resource optimization module 209, wherein the data acquisition module 201 is used to determine a target object and obtain initial surrounding configuration information corresponding to the target object; the information acquisition module 202 is used to obtain initial consumption information corresponding to the target user under the target object and a user driving path corresponding to the target user; the data clustering module 203 is used to cluster the target user according to the initial consumption information and the user driving path to obtain a target clustering result; the data determination module 204 is used to obtain relevant label information involved in each first subclass cluster in the target clustering result, and Obtain relevant frequency information corresponding to the relevant tag information; a data analysis module 205 is used to determine the initial consumption tag, and determine the initial relevant index corresponding to the relevant tag information to the initial consumption tag according to the relevant frequency information; a data adjustment module 206 is used to obtain the consumption time corresponding to the relevant tag information, and perform time decay on the initial relevant index according to the consumption time to obtain the target relevant index; a data screening module 207 is used to perform tag screening from the initial consumption tag according to the target relevant index to obtain the initial consumption type corresponding to the first sub-class cluster; an information adjustment module 208 is used to obtain the relevant driving path corresponding to the first sub-class cluster from the user's driving path, and perform information adjustment on the initial consumption type according to the relevant driving path to obtain the target consumption type; a resource optimization module 209 is used to perform resource optimization on the initial surrounding configuration information according to the target consumption type to obtain the target surrounding configuration information.

[0105] In some implementations, the information acquisition module 202 performs the following steps in the process of obtaining the user driving path corresponding to the target user:

[0106] Obtaining initial vehicle image information corresponding to a target sensor associated with the target object;

[0107] Preprocessing the initial vehicle image information to obtain corresponding target vehicle image information;

[0108] Performing a windowing process on the target vehicle image information and then performing a Fourier transform to obtain target frequency spectrum information corresponding to the target vehicle image information;

[0109] Performing information detection on the target spectrum information to obtain a blur direction and a blur length corresponding to the initial vehicle image information;

[0110] Performing image restoration processing on the target vehicle image information according to the blur direction and the blur length to obtain repaired image information corresponding to the target vehicle image information;

[0111] Performing license plate information recognition on the repaired image information to obtain predicted license plate information corresponding to the initial vehicle image information;

[0112] The target license plate information corresponding to the target user is obtained, and the predicted license plate information is compared with the target license plate information to obtain the user driving path corresponding to the target user.

[0113] In some implementations, the data clustering module 203 performs, in the process of clustering the target user according to the initial consumption information and the user travel path to obtain the target clustering result:

[0114] Determine a user departure location and a user arrival location corresponding to the target user according to the user's travel path;

[0115] Determine a target intent label corresponding to the target user and a first label probability corresponding to the target intent label according to the user's departure location and the user's arrival location;

[0116] Performing tag recognition according to the initial consumption information to obtain user tag information corresponding to the target user and a second tag probability corresponding to the user tag information;

[0117] Performing user clustering on the target user according to the target intention label, the first label probability, the user label information and the second label probability to obtain an initial clustering result;

[0118] Calculating the data correlation of each second sub-cluster in the initial clustering result to obtain a data correlation value between the cluster center and the cluster data in the second sub-cluster;

[0119] The initial clustering result is adjusted according to the data correlation value to obtain the target clustering result corresponding to the target user.

[0120] In some implementations, the data clustering module 203 performs, in the process of clustering the target user according to the target intent label, the first label probability, the user label information, and the second label probability to obtain an initial clustering result:

[0121] Obtain a first user and a second user from the target user, and obtain a first intent label corresponding to the first user and a third label probability corresponding to the first intent label from the target intent label and the first label probability, and obtain a second intent label corresponding to the second user and a fourth label probability corresponding to the second intent label;

[0122] Obtain first label information corresponding to the first user and a fifth label probability corresponding to the first label information from the user label information and the second label probability, and obtain second label information corresponding to the second user and a sixth label probability corresponding to the second label information;

[0123] Performing intersection processing according to the first intent label and the second intent label to obtain first intersection information corresponding to the first user and the second user;

[0124] Performing intersection processing on the first label information and the second label information to obtain second intersection information corresponding to the first user and the second user;

[0125] Determine the user similarity between the first user and the second user by fusing the third label probability, the fourth label probability, the fifth label probability and the sixth label probability according to the first intersection information and the second intersection information;

[0126] Clustering the target users according to the user similarities to obtain the initial clustering results;

[0127] The user similarity is obtained according to the following formula:

[0128]

[0129] in, represents the user similarity between the first user and the second user, Indicates the number of tags corresponding to the first intent tag corresponding to the first user, represents the third label probability corresponding to the t-th first intention label of the first user; represents the number of tags corresponding to the first tag information corresponding to the first user, represents the fifth label probability corresponding to the mth first label information corresponding to the first user; represents the number of tags corresponding to the second intent tag corresponding to the second user, The fourth label probability corresponding to the kth second intention label of the second user; Indicates the number of tags corresponding to the second tag information corresponding to the second user; represents the sixth label probability corresponding to the r-th second label information of the second user; represents the number of intersections corresponding to the first intersection information corresponding to the first user and the second user, represents the label probability corresponding to the first user under the dth first intersection information; represents the label probability corresponding to the second user under the dth first intersection information; represents the number of intersections corresponding to the second intersection information corresponding to the first user and the second user, represents the label probability corresponding to the first user under the sth second intersection information; represents the label probability corresponding to the second user under the sth second intersection information.

[0130] In some implementations, the data clustering module 203, in the process of calculating the data correlation for each second sub-cluster in the initial clustering result to obtain the data correlation value between the cluster center and the cluster data in the second sub-cluster, executes:

[0131] Performing center point calculation on the second subclass cluster to obtain the cluster center corresponding to the second subclass cluster;

[0132] Obtaining, from the target intent label, a third intent label corresponding to the cluster center and first probability information corresponding to the third intent label;

[0133] Obtaining third label information corresponding to the cluster center and second probability information corresponding to the third label information from the user label information;

[0134] Obtaining, from the target intent label, a fourth intent label corresponding to the cluster data in the second sub-cluster and third probability information corresponding to the fourth intent label;

[0135] Obtaining, from the user label information, fourth label information corresponding to the cluster data in the second sub-cluster and fourth probability information corresponding to the fourth label information;

[0136] Determine the distance information between the cluster center and the cluster data according to the third intent label, the first probability information, the third label information and the second probability information and the fourth intent label, the third probability information, the fourth label information and the fourth probability information;

[0137] The data correlation value between the cluster center and the cluster data in the second sub-cluster is determined according to the distance information.

[0138] In some implementations, the data adjustment module 206 performs, during the process of performing time decay on the initial correlation index according to the consumption time to obtain the target correlation index:

[0139] Obtaining the earliest behavior time and the latest behavior time from the consumption time;

[0140] Determine a time decay parameter corresponding to the consumption time according to the earliest behavior time and the latest behavior time;

[0141] Performing time decay on the initial correlation index according to the time decay parameter to obtain the target correlation index;

[0142] The target correlation index is obtained according to the following formula:

[0143]

[0144] in, Indicates the target relevance index corresponding to the i-th initial consumption tag of the relevant tag information; represents the consumption time corresponding to the j-th related tag information, represents the earliest behavior time, Indicates the latest behavior time. , represents the adjustment parameter determined by the initial consumption tag; Indicates the number of tags corresponding to the relevant tag information, represents the label weight corresponding to the jth related label information under the ith initial consumption label, Represents the relevant frequency information corresponding to the j-th relevant label information.

[0145] In some implementations, during the process of adjusting the information of the initial consumption type according to the relevant driving path to obtain the target consumption type, the information adjustment module 208 executes:

[0146] Performing path prediction according to the relevant driving paths to obtain driving destinations corresponding to relevant users in the first sub-cluster;

[0147] Determining the user consumption intention corresponding to the relevant user in the first sub-cluster according to the travel destination;

[0148] The target consumption type is obtained by adjusting information of the initial consumption type according to the user's consumption intention.

[0149] In some implementations, the local life resource optimization and allocation apparatus 200 may be applied to a terminal device.

[0150] It should be noted that technical personnel in the relevant field can clearly understand that for the convenience and conciseness of description, the specific working process of the local life resource optimization and allocation device 200 described above can refer to the corresponding process in the aforementioned local life resource optimization and allocation method embodiment, and will not be repeated here.

[0151] See also Figure 3 , Figure 3 A schematic block diagram of the structure of a terminal device provided in an embodiment of the present invention.

[0152] like Figure 3 As shown, the terminal device 300 includes a processor 301 and a memory 302 , and the processor 301 and the memory 302 are connected via a bus 303 , such as an I2C (Inter-integrated Circuit) bus.

[0153] Specifically, the processor 301 is used to provide computing and control capabilities to support the operation of the entire terminal device. The processor 301 can be a central processing unit (CPU), and the processor 301 can also be other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components, etc. Among them, the general-purpose processor can be a microprocessor or the processor can also be any conventional processor, etc.

[0154] Specifically, the memory 302 may be a Flash chip, a read-only memory (ROM) disk, an optical disk, a USB flash drive, or a mobile hard disk.

[0155] Those skilled in the art will understand that Figure 3 The structure shown in the figure is only a block diagram of a partial structure related to the embodiment of the present invention, and does not constitute a limitation on the terminal device to which the embodiment of the present invention is applied. The specific server may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0156] The processor is used to run a computer program stored in the memory, and implement any one of the local life resource optimization and allocation methods provided in the embodiments of the present invention when executing the computer program.

[0157] In one embodiment, the processor is used to run a computer program stored in the memory, and implements the following steps when executing the computer program:

[0158] Determine a target object and obtain initial surrounding configuration information corresponding to the target object;

[0159] Obtaining initial consumption information corresponding to the target user under the target object and a user travel path corresponding to the target user;

[0160] Performing user clustering on the target user according to the initial consumption information and the user's travel path to obtain a target clustering result;

[0161] Obtain relevant label information involved in each first subclass cluster in the target clustering result, and obtain relevant frequency information corresponding to the relevant label information;

[0162] Determine an initial consumption tag, and determine an initial correlation index corresponding to the initial consumption tag by the correlation tag information according to the correlation frequency information;

[0163] Obtaining the consumption time corresponding to the relevant tag information, and performing time decay on the initial relevant index according to the consumption time to obtain a target relevant index;

[0164] According to the target-related index, label screening is performed from the initial consumption labels to obtain the initial consumption type corresponding to the first sub-category cluster;

[0165] Obtaining a related driving path corresponding to the first sub-class cluster from the user driving path, and adjusting information of the initial consumption type according to the related driving path to obtain a target consumption type;

[0166] The initial peripheral configuration information is resource optimized according to the target consumption type to obtain target peripheral configuration information.

[0167] In some implementations, the processor 301, in the process of obtaining the user driving path corresponding to the target user, executes:

[0168] Obtaining initial vehicle image information corresponding to a target sensor associated with the target object;

[0169] Preprocessing the initial vehicle image information to obtain corresponding target vehicle image information;

[0170] Performing a windowing process on the target vehicle image information and then performing a Fourier transform to obtain target frequency spectrum information corresponding to the target vehicle image information;

[0171] Performing information detection on the target spectrum information to obtain a blur direction and a blur length corresponding to the initial vehicle image information;

[0172] Performing image restoration processing on the target vehicle image information according to the blur direction and the blur length to obtain repaired image information corresponding to the target vehicle image information;

[0173] Performing license plate information recognition on the repaired image information to obtain predicted license plate information corresponding to the initial vehicle image information;

[0174] The target license plate information corresponding to the target user is obtained, and the predicted license plate information is compared with the target license plate information to obtain the user driving path corresponding to the target user.

[0175] In some implementations, the processor 301, in the process of clustering the target user according to the initial consumption information and the user driving path to obtain the target clustering result, executes:

[0176] Determine a user departure location and a user arrival location corresponding to the target user according to the user's travel path;

[0177] Determine a target intent label corresponding to the target user and a first label probability corresponding to the target intent label according to the user's departure location and the user's arrival location;

[0178] Performing tag recognition according to the initial consumption information to obtain user tag information corresponding to the target user and a second tag probability corresponding to the user tag information;

[0179] Performing user clustering on the target user according to the target intention label, the first label probability, the user label information and the second label probability to obtain an initial clustering result;

[0180] Calculating the data correlation of each second sub-cluster in the initial clustering result to obtain a data correlation value between the cluster center and the cluster data in the second sub-cluster;

[0181] The initial clustering result is adjusted according to the data correlation value to obtain the target clustering result corresponding to the target user.

[0182] In some implementations, the processor 301, in the process of clustering the target user according to the target intent label, the first label probability, the user label information, and the second label probability to obtain an initial clustering result, executes:

[0183] Obtain a first user and a second user from the target user, and obtain a first intent label corresponding to the first user and a third label probability corresponding to the first intent label from the target intent label and the first label probability, and obtain a second intent label corresponding to the second user and a fourth label probability corresponding to the second intent label;

[0184] Obtain first label information corresponding to the first user and a fifth label probability corresponding to the first label information from the user label information and the second label probability, and obtain second label information corresponding to the second user and a sixth label probability corresponding to the second label information;

[0185] Performing intersection processing according to the first intent label and the second intent label to obtain first intersection information corresponding to the first user and the second user;

[0186] Performing intersection processing on the first label information and the second label information to obtain second intersection information corresponding to the first user and the second user;

[0187] Determine the user similarity between the first user and the second user by fusing the third label probability, the fourth label probability, the fifth label probability and the sixth label probability according to the first intersection information and the second intersection information;

[0188] Clustering the target users according to the user similarities to obtain the initial clustering results;

[0189] The user similarity is obtained according to the following formula:

[0190]

[0191] in, represents the user similarity between the first user and the second user, Indicates the number of tags corresponding to the first intent tag corresponding to the first user, represents the third label probability corresponding to the t-th first intention label of the first user; represents the number of tags corresponding to the first tag information corresponding to the first user, represents the fifth label probability corresponding to the mth first label information corresponding to the first user; represents the number of tags corresponding to the second intent tag corresponding to the second user, The fourth label probability corresponding to the kth second intention label of the second user; Indicates the number of tags corresponding to the second tag information corresponding to the second user; represents the sixth label probability corresponding to the r-th second label information of the second user; represents the number of intersections corresponding to the first intersection information corresponding to the first user and the second user, represents the label probability corresponding to the first user under the dth first intersection information; represents the label probability corresponding to the second user under the dth first intersection information; represents the number of intersections corresponding to the second intersection information corresponding to the first user and the second user, represents the label probability corresponding to the first user under the sth second intersection information; represents the label probability corresponding to the second user under the sth second intersection information.

[0192] In some implementations, the processor 301, in the process of calculating the data correlation for each second subclass cluster in the initial clustering result to obtain the data correlation value between the cluster center and the cluster data in the second subclass cluster, executes:

[0193] Performing center point calculation on the second subclass cluster to obtain the cluster center corresponding to the second subclass cluster;

[0194] Obtaining, from the target intent label, a third intent label corresponding to the cluster center and first probability information corresponding to the third intent label;

[0195] Obtaining third label information corresponding to the cluster center and second probability information corresponding to the third label information from the user label information;

[0196] Obtaining, from the target intent label, a fourth intent label corresponding to the cluster data in the second sub-cluster and third probability information corresponding to the fourth intent label;

[0197] Obtaining, from the user label information, fourth label information corresponding to the cluster data in the second sub-cluster and fourth probability information corresponding to the fourth label information;

[0198] Determine the distance information between the cluster center and the cluster data according to the third intent label, the first probability information, the third label information and the second probability information and the fourth intent label, the third probability information, the fourth label information and the fourth probability information;

[0199] The data correlation value between the cluster center and the cluster data in the second sub-cluster is determined according to the distance information.

[0200] In some implementations, during the process of performing time decay on the initial correlation index according to the consumption time to obtain a target correlation index, the processor 301 executes:

[0201] Obtaining the earliest behavior time and the latest behavior time from the consumption time;

[0202] Determine a time decay parameter corresponding to the consumption time according to the earliest behavior time and the latest behavior time;

[0203] Performing time decay on the initial correlation index according to the time decay parameter to obtain the target correlation index;

[0204] The target correlation index is obtained according to the following formula:

[0205]

[0206] in, Indicates the target relevance index corresponding to the i-th initial consumption tag of the relevant tag information; represents the consumption time corresponding to the j-th related tag information, represents the earliest behavior time, Indicates the latest behavior time. , represents the adjustment parameter determined by the initial consumption tag; Indicates the number of tags corresponding to the relevant tag information, represents the label weight corresponding to the jth related label information under the ith initial consumption label, Represents the relevant frequency information corresponding to the j-th relevant label information.

[0207] In some implementations, the processor 301, during the process of adjusting the information of the initial consumption type according to the relevant driving path to obtain the target consumption type, executes:

[0208] Performing path prediction according to the relevant driving paths to obtain driving destinations corresponding to relevant users in the first sub-cluster;

[0209] Determining the user consumption intention corresponding to the relevant user in the first sub-cluster according to the travel destination;

[0210] The target consumption type is obtained by adjusting information of the initial consumption type according to the user's consumption intention.

[0211] It should be noted that technical personnel in the relevant field can clearly understand that for the convenience and conciseness of description, the specific working process of the terminal device described above can refer to the corresponding process in the aforementioned local life resource optimization and allocation method embodiment, and will not be repeated here.

[0212] An embodiment of the present invention also provides a storage medium for computer-readable storage, wherein the storage medium stores one or more programs, and the one or more programs can be executed by one or more processors to implement the steps of any local life resource optimization and allocation method provided in the description of the embodiment of the present invention.

[0213] The storage medium may be an internal storage unit of the terminal device described in the foregoing embodiment, such as a hard disk or memory of the terminal device. The storage medium may also be an external storage device of the terminal device, such as a plug-in hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (SecureDigital, SD) card, a flash card (Flash Card), etc., equipped on the terminal device.

[0214] It will be appreciated by those skilled in the art that all or some of the steps, systems, and functional modules / units in the methods disclosed above may be implemented as software, firmware, hardware, and appropriate combinations thereof. In a hardware embodiment, the division between the functional modules / units mentioned in the above description does not necessarily correspond to the division of physical components; for example, a physical component may have multiple functions, or a function or step may be performed by several physical components in cooperation. Some or all physical components may be implemented as software executed by a processor, such as a central processing unit, a digital signal processor, or a microprocessor, or implemented as hardware, or implemented as an integrated circuit, such as an application-specific integrated circuit. Such software may be distributed on a computer-readable medium, which may include a computer storage medium (or non-transitory medium) and a communication medium (or transient medium). As is known to those skilled in the art, the term computer storage medium includes volatile and non-volatile, removable and non-removable media implemented in any method or technology for storing information (such as computer-readable instructions, data structures, program modules, or other data). Computer storage media include, but are not limited to, RAM, ROM, EEPROM, flash memory or other memory technology, CD-ROM, digital versatile disks (DVD) or other optical disk storage, magnetic cassettes, magnetic tapes, magnetic disk storage or other magnetic storage devices, or any other medium that can be used to store the desired information and can be accessed by a computer. In addition, it is well known to those skilled in the art that communication media typically contain computer-readable instructions, data structures, program modules, or other data in a modulated data signal such as a carrier wave or other transport mechanism, and may include any information delivery media.

[0215] It should be understood that the term "and / or" used in the present specification and the appended claims refers to any combination of one or more of the associated listed items and all possible combinations, including these combinations. It should be noted that, in this article, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or system including a series of elements includes not only those elements, but also includes other elements not explicitly listed, or also includes elements inherent to such process, method, article or system. In the absence of further restrictions, an element defined by the sentence "including a..." does not exclude the presence of other identical elements in the process, method, article or system including the element.

[0216] The serial numbers of the embodiments of the present invention are only for description and do not represent the advantages and disadvantages of the embodiments. The above description is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can easily think of various equivalent modifications or replacements within the technical scope disclosed by the present invention, and these modifications or replacements should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention shall be based on the protection scope of the claims.

Claims

1. A method for optimizing and allocating local life resources, characterized in that: The method comprises: Determine a target object and obtain initial surrounding configuration information corresponding to the target object; Obtaining initial consumption information corresponding to the target user under the target object and a user travel path corresponding to the target user; Performing user clustering on the target user according to the initial consumption information and the user's travel path to obtain a target clustering result; Obtain relevant label information involved in each first subclass cluster in the target clustering result, and obtain relevant frequency information corresponding to the relevant label information; Determine an initial consumption tag, and determine an initial correlation index corresponding to the initial consumption tag by the correlation tag information according to the correlation frequency information; Obtaining the consumption time corresponding to the relevant tag information, and performing time decay on the initial relevant index according to the consumption time to obtain a target relevant index; According to the target-related index, label screening is performed from the initial consumption labels to obtain the initial consumption type corresponding to the first sub-category cluster; Obtaining a related driving path corresponding to the first sub-class cluster from the user driving path, and adjusting information of the initial consumption type according to the related driving path to obtain a target consumption type; The initial peripheral configuration information is resource optimized according to the target consumption type to obtain target peripheral configuration information.

2. The method according to claim 1, characterized in that Obtaining the user travel path corresponding to the target user includes: Obtaining initial vehicle image information corresponding to a target sensor associated with the target object; Preprocessing the initial vehicle image information to obtain corresponding target vehicle image information; Performing a windowing process on the target vehicle image information and then performing a Fourier transform to obtain target frequency spectrum information corresponding to the target vehicle image information; Performing information detection on the target spectrum information to obtain a blur direction and a blur length corresponding to the initial vehicle image information; Performing image restoration processing on the target vehicle image information according to the blur direction and the blur length to obtain repaired image information corresponding to the target vehicle image information; Performing license plate information recognition on the repaired image information to obtain predicted license plate information corresponding to the initial vehicle image information; The target license plate information corresponding to the target user is obtained, and the predicted license plate information is compared with the target license plate information to obtain the user driving path corresponding to the target user.

3. The method according to claim 1, characterized in that The step of clustering the target users according to the initial consumption information and the user travel path to obtain a target clustering result includes: Determine a user departure location and a user arrival location corresponding to the target user according to the user's travel path; Determine a target intent label corresponding to the target user and a first label probability corresponding to the target intent label according to the user's departure location and the user's arrival location; Performing tag recognition according to the initial consumption information to obtain user tag information corresponding to the target user and a second tag probability corresponding to the user tag information; Performing user clustering on the target user according to the target intention label, the first label probability, the user label information and the second label probability to obtain an initial clustering result; Calculating the data correlation of each second sub-cluster in the initial clustering result to obtain a data correlation value between the cluster center and the cluster data in the second sub-cluster; The initial clustering result is adjusted according to the data correlation value to obtain the target clustering result corresponding to the target user.

4. The method according to claim 3, characterized in that The performing user clustering on the target user according to the target intention label, the first label probability, the user label information and the second label probability to obtain an initial clustering result includes: Obtain a first user and a second user from the target user, and obtain a first intent label corresponding to the first user and a third label probability corresponding to the first intent label from the target intent label and the first label probability, and obtain a second intent label corresponding to the second user and a fourth label probability corresponding to the second intent label; Obtain first label information corresponding to the first user and a fifth label probability corresponding to the first label information from the user label information and the second label probability, and obtain second label information corresponding to the second user and a sixth label probability corresponding to the second label information; Performing intersection processing according to the first intent label and the second intent label to obtain first intersection information corresponding to the first user and the second user; Performing intersection processing on the first label information and the second label information to obtain second intersection information corresponding to the first user and the second user; Determine the user similarity between the first user and the second user by fusing the third label probability, the fourth label probability, the fifth label probability and the sixth label probability according to the first intersection information and the second intersection information; Clustering the target users according to the user similarities to obtain the initial clustering results; The user similarity is obtained according to the following formula: ； in, represents the user similarity between the first user and the second user, Indicates the number of tags corresponding to the first intent tag corresponding to the first user, represents the third label probability corresponding to the t-th first intention label of the first user; represents the number of tags corresponding to the first tag information corresponding to the first user, represents the fifth label probability corresponding to the mth first label information corresponding to the first user; represents the number of tags corresponding to the second intent tag corresponding to the second user, The fourth label probability corresponding to the kth second intention label of the second user; Indicates the number of tags corresponding to the second tag information corresponding to the second user; represents the sixth label probability corresponding to the r-th second label information of the second user; represents the number of intersections corresponding to the first intersection information corresponding to the first user and the second user, represents the label probability corresponding to the first user under the dth first intersection information; represents the label probability corresponding to the second user under the dth first intersection information; represents the number of intersections corresponding to the second intersection information corresponding to the first user and the second user, represents the label probability corresponding to the first user under the sth second intersection information; represents the label probability corresponding to the second user under the sth second intersection information.

5. The method according to claim 3, characterized in that: The step of calculating the data correlation of each second sub-cluster in the initial clustering result to obtain a data correlation value between the cluster center and the cluster data in the second sub-cluster includes: Performing center point calculation on the second subclass cluster to obtain the cluster center corresponding to the second subclass cluster; Obtaining, from the target intent label, a third intent label corresponding to the cluster center and first probability information corresponding to the third intent label; Obtaining third label information corresponding to the cluster center and second probability information corresponding to the third label information from the user label information; Obtaining, from the target intent label, a fourth intent label corresponding to the cluster data in the second sub-cluster and third probability information corresponding to the fourth intent label; Obtaining, from the user label information, fourth label information corresponding to the cluster data in the second sub-cluster and fourth probability information corresponding to the fourth label information; Determine the distance information between the cluster center and the cluster data according to the third intent label, the first probability information, the third label information and the second probability information and the fourth intent label, the third probability information, the fourth label information and the fourth probability information; The data correlation value between the cluster center and the cluster data in the second sub-cluster is determined according to the distance information.

6. The method according to claim 1, characterized in that The step of performing time decay on the initial correlation index according to the consumption time to obtain a target correlation index includes: Obtaining the earliest behavior time and the latest behavior time from the consumption time; Determine a time decay parameter corresponding to the consumption time according to the earliest behavior time and the latest behavior time; Performing time decay on the initial correlation index according to the time decay parameter to obtain the target correlation index; The target correlation index is obtained according to the following formula: ； in, Indicates the target relevance index corresponding to the i-th initial consumption tag of the relevant tag information; represents the consumption time corresponding to the j-th related tag information, represents the earliest behavior time, Indicates the latest behavior time. , represents the adjustment parameter determined by the initial consumption tag; Indicates the number of tags corresponding to the relevant tag information, represents the label weight corresponding to the jth related label information under the ith initial consumption label, Represents the relevant frequency information corresponding to the j-th relevant label information.

7. The method according to claim 1, characterized in that The step of adjusting the information of the initial consumption type according to the relevant driving path to obtain the target consumption type includes: Performing path prediction according to the relevant driving paths to obtain driving destinations corresponding to relevant users in the first sub-cluster; Determining the user consumption intention corresponding to the relevant user in the first sub-cluster according to the travel destination; The target consumption type is obtained by adjusting information of the initial consumption type according to the user's consumption intention.

8. A local life resource optimization and allocation device, characterized in that: include: A data acquisition module, used to determine a target object and obtain initial surrounding configuration information corresponding to the target object; An information acquisition module, used to obtain the initial consumption information corresponding to the target user under the target object and the user travel path corresponding to the target user; A data clustering module, used to cluster the target users according to the initial consumption information and the user travel path, and obtain a target clustering result; A data determination module, used to obtain relevant label information involved in each first subclass cluster in the target clustering result, and obtain relevant frequency information corresponding to the relevant label information; A data analysis module, used to determine an initial consumption tag, and determine an initial correlation index corresponding to the initial consumption tag by the correlation tag information according to the correlation frequency information; A data adjustment module, used to obtain the consumption time corresponding to the relevant tag information, and perform time decay on the initial relevant index according to the consumption time to obtain a target relevant index; A data screening module, configured to screen the initial consumption tags according to the target-related index to obtain the initial consumption type corresponding to the first sub-class cluster; An information adjustment module, configured to obtain a related driving path corresponding to the first sub-class cluster from the user's driving path, and adjust the initial consumption type according to the related driving path to obtain a target consumption type; The resource optimization module is used to perform resource optimization on the initial peripheral configuration information according to the target consumption type to obtain target peripheral configuration information.

9. A terminal device, characterized in that: The terminal device includes a processor and a memory; The memory is used to store computer programs; The processor is used to execute the computer program and implement the local life resource optimization and allocation method as described in any one of claims 1 to 7 when executing the computer program.

Images