Information push methods, information push devices and information push equipment

Abstract

This application discloses an information push method, an information push device, and an information push equipment, belonging to the field of information processing technology. The method includes: in response to obtaining information that a target vehicle has entered a parking lot, determining the target vehicle's parking intention based on the target user profile of the target vehicle; determining target merchants within the geographical area of ​​the corresponding business district of the parking lot based on the parking intention; and sending push information about the target merchants to the terminal device corresponding to the target vehicle. This application can improve the matching degree between push information and user needs.

Description

Technical Field

[0001] This application belongs to the field of information processing technology, and in particular relates to an information push method, information push device and information push equipment. Background Technology

[0002] With the deepening of smart city construction, the synergistic optimization of urban parking resources and commercial services has become an important direction for improving urban operational efficiency and user experience. Parking lots are no longer just infrastructure for vehicle parking, but are gradually evolving into key nodes connecting car owners' consumption behavior with the surrounding commercial ecosystem. How to effectively integrate parking scenarios with surrounding business district services, and push accurate and timely commercial information to car owners while they are parking, has become an important application scenario in the field of smart parking.

[0003] However, existing information push methods typically push the same business district information to all users in the parking lot, resulting in a low degree of matching between the pushed information and user needs. Summary of the Invention

[0004] This application provides an information push method, an information push device, and an information push equipment, which can improve the matching degree between pushed information and user needs.

[0005] In a first aspect, embodiments of this application provide an information push method, the method comprising: In response to the acquisition of the target vehicle entering the parking lot, the parking intention of the target vehicle is determined based on the target user profile of the target vehicle; Based on the parking intention, identify the target merchants within the geographical area of ​​the corresponding business district of the parking lot; Send push notifications from the target merchant to the terminal device corresponding to the target vehicle.

[0006] Secondly, embodiments of this application provide an information push device, the device comprising: The determination module is used to determine the parking intention of a target vehicle in response to the acquisition of a target vehicle entering the parking lot, based on the target user profile of the target vehicle. The matching module is used to determine the target merchants that match the parking intention within the geographical area of ​​the business district corresponding to the parking lot; The push module is used to send push information from the target merchant to the terminal device corresponding to the target vehicle.

[0007] Thirdly, embodiments of this application provide an information push device, including: a processor and a memory storing computer program instructions; When the processor executes computer program instructions, it implements the information push method as described in the first aspect.

[0008] Fourthly, embodiments of this application provide a computer-readable storage medium storing computer program instructions, which, when executed by a processor, implement the information push method as described in the first aspect.

[0009] Fifthly, embodiments of this application provide a computer program product in which instructions, when executed by a vehicle's processor, cause the vehicle to perform the information push method as described in the first aspect.

[0010] In this embodiment, in response to the acquisition of a target vehicle entering a parking lot, the parking intention of the target vehicle is determined based on the target user profile of the target vehicle to gain a deeper understanding of the user's potential purpose for parking. Then, based on the parking intention, target merchants matching the parking intention are determined within the geographical area of ​​the business district corresponding to the parking lot, and push information of the target merchants is sent to the terminal device corresponding to the target vehicle. This allows the merchants corresponding to the push information received by each vehicle user to be determined based on the vehicle's user profile. In this way, the personalization and accuracy of information push can be achieved, and the matching degree between push information and user needs can be improved. Attached Figure Description

[0011] To more clearly illustrate the technical solutions of the embodiments of this application, the accompanying drawings used in the embodiments of this application will be briefly introduced below. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0012] Figure 1 This is a schematic diagram of the structure of a system framework applicable to the embodiments of this application; Figure 2 This is one of the flowcharts illustrating the information push method provided in the embodiments of this application; Figure 3 This is an interactive diagram of a system for which the information push method provided in this application embodiment can be applied; Figure 4 This is a second flowchart illustrating the information push method provided in the embodiments of this application; Figure 5 This is a schematic diagram of the structure of the information push device provided in the embodiments of this application; Figure 6 This is a schematic diagram of the structure of the information push device provided in the embodiments of this application. Detailed Implementation

[0013] The features and exemplary embodiments of various aspects of this application will be described in detail below. To make the objectives, technical solutions, and advantages of this application clearer, the application will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are only intended to explain this application and not to limit it. For those skilled in the art, this application can be implemented without some of these specific details. The following description of the embodiments is merely to provide a better understanding of this application by illustrating examples.

[0014] It should be noted that, in this document, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising..." does not exclude the presence of additional identical elements in the process, method, article, or apparatus that includes the element.

[0015] In all specific embodiments of this application, when processing data related to user identity or characteristics, such as user information, user behavior data, user historical data, and user location information, user permission or consent is obtained first. Furthermore, the collection, use, and processing of this data comply with relevant laws, regulations, and standards. Additionally, when embodiments of this application require access to sensitive personal information, separate permission or consent from the user is obtained through pop-ups or redirects to confirmation pages. Only after obtaining the user's separate permission or consent is the necessary user-related data required for the proper functioning of these embodiments obtained.

[0016] To address the problems of the prior art, embodiments of this application provide an information push method, an information push device, and an information push equipment. The information push method provided in this application embodiment will be described first below.

[0017] Figure 1 This is a schematic diagram of a system framework applicable to embodiments of this application, including: user terminal, parking system, profile service, business engine, and merchant system.

[0018] The user terminal can be an in-vehicle device or a mobile terminal of the user driving the vehicle. The mobile terminal can be a smartphone, tablet, laptop, desktop computer, etc. The user terminal can be used to receive push information and collect feedback information from the user regarding the push information.

[0019] The parking system can be used to collect signals of vehicles entering the parking lot, as well as information such as the vehicle's parking trajectory and parking time. The parking system can also be used to receive push information determined by the business engine and push the information to the user terminal.

[0020] The user profile service can be used to store a profile library. It should be noted that this application embodiment does not limit the method of obtaining user profiles in the preset profile library. User profiles in the preset profile library can be constructed, but are not limited to, the following methods: they can be constructed based on the target vehicle's user's historical consumption preferences, historical parking behavior, vehicle attribute information, and basic user information. The profile dimensions of each user profile in the profile library may differ, and the profile dimensions can be determined based on the number of features in the user profile.

[0021] A business engine can be used to determine push notifications.

[0022] The merchant system can be used to provide merchants' marketing information.

[0023] Figure 2 This illustration shows one of the flowcharts of an information push method provided in an embodiment of this application. The information push method provided in this embodiment is applied to an information push device. Specifically, the information push device can be an electronic device. The method can be executed by the information push device itself, or by components of the information push device, such as its processor, chip, or chip system. It can also be implemented by logic modules or software that implement all or part of the functions of the information push device. In practical applications, the information push device can be a business engine, server, service platform, cloud, distributed system, Internet of Things (IoT), or vehicle network system, etc.

[0024] like Figure 2 As shown, the information push method of this application embodiment may include the following steps 101-103.

[0025] Step 101: In response to the acquisition of the target vehicle entering the parking lot, determine the parking intention of the target vehicle based on the target user profile of the target vehicle.

[0026] In this step, the acquisition of the target vehicle entering the parking lot can be achieved by recognizing the license plate number through the parking lot entrance camera in the parking lot system and sending a signal of the vehicle entering the parking lot to the information push device through the parking lot system. Alternatively, the vehicle-mounted device in the user terminal can send a signal of the vehicle entering the parking lot to the information push device when it detects that the vehicle has entered a specific geographical location, such as within the coordinate range of the parking lot.

[0027] This application does not limit the method of obtaining the target user profile. In some embodiments, the target user profile can be obtained by querying, or more specifically, by querying from a preset profile database. For details on the implementation, please refer to the relevant descriptions below, which will not be repeated here.

[0028] In other embodiments, the target user profile can be obtained through generation, as detailed in the following descriptions, which will not be repeated here. Furthermore, the generated target user profile can be stored in a preset profile library.

[0029] The target user profile can include features such as the user's consumption preferences, parking behavior, and frequently visited merchants associated with the target vehicle. In some implementations, the user profile can be updated based on the user's feedback on the push notifications or new parking events of the vehicle. This update can modify the presentation of features in the user profile or the number of features in the user profile.

[0030] Furthermore, the parking intention of a target vehicle can be determined based on the target user profile. In some implementations, the target user profile can be input into a pre-trained neural network model, such as a Markov model or a decision tree model, and the neural network model outputs the parking intention. In other implementations, the user profile can have a pre-defined matching relationship between target features and parking intentions. Different manifestations of the target features will match different parking intentions, and the parking intention can be determined based on the manifestation of the target features in the target user profile. The target features can be any feature in the user profile, such as consumption preference features.

[0031] Step 102: Based on the parking intention, identify the target merchants within the geographical area of ​​the business district corresponding to the parking lot.

[0032] In this step, the geographical scope of the business district corresponding to the parking lot can be a pre-defined fixed range, such as an area with a radius of 2 kilometers centered on the parking lot; it can also be determined based on the actual relationship between the parking lot and the surrounding business district, such as the parking lot being a supporting facility of a certain business district; it can also be adjusted according to the actual application scenario, without specific limitations here.

[0033] In some implementations, the geographical area of ​​a business district may include multiple Points of Interest (POIs). POIs can be independent businesses or facilities with consumption or service functions, such as shopping malls, restaurants, cinemas, supermarkets, and entertainment venues. Each POI corresponds to a physical business, and each POI's data may include fields such as the business's name, category label, geographical location, and business hours.

[0034] In some implementations, the geographical extent of the business district can be determined based on the walkable radius R_walkable, that is, the geographical extent of the business district is the area within the radius of the parking lot. The radius R_walkable can be calculated using formula (1): R_walkable = v_avg * T_max_walk(1) In the formula, v_avg is the average walking speed, and T_max_walk is the maximum acceptable walking time.

[0035] It should be noted that v_avg and T_max_walk can be set according to the actual situation, and no specific restrictions are made here.

[0036] For example, v_avg can be 1.2 m / s, T_max_walk can be 10 minutes, then R_walkable is 720 meters.

[0037] Furthermore, parking intentions can be matched with category tags of businesses within the geographical area of ​​the business district to determine target businesses. For example, if a user's parking intention is to dine, businesses within the geographical area categorized as "food" are identified as target businesses. In some implementations, parking intentions can first be matched with tags of businesses within the geographical area of ​​the business district to determine candidate businesses, and then the target business can be determined from the candidate businesses based on information such as price, popularity, and reviews.

[0038] Step 103: Send push information of the target merchant to the terminal device corresponding to the target vehicle.

[0039] In this step, the terminal device corresponding to the target vehicle can be an in-vehicle terminal or a mobile device of the user driving the target vehicle; no specific limitation is made here. Information can be pushed to the terminal device via SMS, application notifications, pop-up windows on the vehicle's infotainment screen, or voice announcements, etc.

[0040] The push notification can include the name and location of the target business, the walking time from the current parking lot location to the target business, and exclusive coupons. For example, the push notification could be: "Welcome to the parking lot. We recommend you visit the Fun Kids Playground on the third floor. There are plenty of seats available, so you don't have to wait. We also have a 20% discount coupon for our Western-style family restaurant. Click to navigate there."

[0041] In this embodiment, in response to the acquisition of a target vehicle entering the parking lot, the parking intention of the target vehicle is determined based on the target user profile of the target vehicle to gain a deeper understanding of the user's potential purpose for parking. Then, based on the parking intention, target merchants matching the parking intention are identified within the geographical area of ​​the business district corresponding to the parking lot, and push information of the target merchants is sent to the terminal device corresponding to the target vehicle, thereby realizing the personalization and accuracy of information push and improving the matching degree between push information and user needs.

[0042] In one embodiment of this application, the target user profile can be obtained through the following steps: Based on the target vehicle's identifier, query the preset profile database; the preset profile database stores the vehicle identifiers of each vehicle and the user profiles corresponding to the vehicle identifiers. In response to finding the target vehicle identifier in the preset profile database, the user profile corresponding to the target vehicle identifier in the preset profile database is determined as the target user profile; In response to the failure to find the target vehicle identifier in the preset profile database, a target user profile is generated based on the target vehicle's vehicle information and its parking information in the parking lot.

[0043] In this embodiment, a profile library can be pre-built, which stores the vehicle identifiers of each vehicle and the user profiles corresponding to the vehicle identifiers.

[0044] User profiles in the profile library can be constructed by obtaining multi-source heterogeneous data of each vehicle from multiple heterogeneous data sources, and then cleaning, normalizing and structuring the multi-source heterogeneous data.

[0045] In some implementations, the multi-source heterogeneous data sources may include parking management systems, license plate recognition systems, IoT sensor networks, third-party commercial platforms, and user terminal mini-programs. In some implementations, the parking management system may obtain vehicle entry and exit records, the license plate recognition system may identify vehicle brand, model, and color, the IoT sensor network may provide vehicle micro-trajectories generated from parking space occupancy status, the third-party commercial platform may obtain real-time information on nearby businesses, and the user terminal mini-program may record clicks, favorites, and in-store verification behaviors.

[0046] Furthermore, the acquired multi-source heterogeneous data can be preprocessed to remove outliers and standardize the format of data from different sources. In some implementations, the multi-source heterogeneous data can be cleaned to remove outliers, such as transit vehicles with a stop time of less than 1 minute or incorrectly identified license plates. The multi-source heterogeneous data can also be normalized, that is, the data format from different sources can be unified. A unified user ID index is recommended; for example, the license plate can be used as the primary key for indexing. Spatiotemporal alignment of the multi-source heterogeneous data can also be performed, associating parking time with commercial activities around the parking lot in both time and space.

[0047] It should be noted that the methods for preprocessing the acquired multi-source heterogeneous data can be combined according to the actual situation, and no specific limitations are made here.

[0048] In some implementations, a dynamically updated, multi-dimensional user profile can be built for each vehicle based on the preprocessed multi-source heterogeneous data of each vehicle.

[0049] For example, the structure of a user profile User_Profile can be User_Profile = {UID, Static_Attrs, Dynamic_Behaviors, Preference_Vector, Life_Cycle_Stage}.

[0050] UID can be a vehicle identifier, which can be a license plate number or other unique identifier of the vehicle.

[0051] Static_Attrs can be static attributes that can include vehicle brand level, such as economy, comfort, and luxury models, as well as the vehicle's license plate location. Brand level can be determined through a predefined brand-price mapping table.

[0052] Dynamic_Behaviors can be a dynamic behavior matrix used to record the characteristics of a user's most recent N, such as 100, parking behaviors. Each parking behavior characteristic can be recorded as [t_start, t_end, duration, lot_type, poi_category, is_weekend, payment_amount], where t_start represents the time of entering the parking lot, t_end represents the time of exiting the parking lot, duration represents the parking duration, lot_type represents the parking lot type code (e.g., 1-residential area, 2-office building, 3-commercial area, 4-hospital, 5-transportation hub), poi_category represents the main PPOI types around the parking lot, which can be determined through clustering algorithms, is_weekend indicates whether the parking behavior occurred on a weekend, and payment_amount represents the amount of money spent during this parking period.

[0053] Preference_Vector can be a consumption preference vector, which can be the preference scores for various business interests inferred from the user's historical consumption behavior. For example, Preference_Vector can be represented as a 10-dimensional vector: Preference_Vector = [Dining 0.8, Shopping 0.6, Entertainment 0.3, Parenting 0.1, Education 0.0, Sports 0.2, Car Services 0.5, Hotels 0.1, Finance 0.0, Medical 0.1]. Furthermore, when new consumption behavior of the vehicle is received, the consumption preference vector can be updated using the following formula (2): P_new(c) = P_old(c) * (1 - α) + α * Reward(c) (2) In the formula, c represents a preference category, α represents the learning rate (0 < α < 1), and Reward(c) represents the feedback value of this behavior to category c. If a restaurant coupon is redeemed after parking, then Reward(restaurant) = 1; if a coupon is pushed but not redeemed, then Reward(restaurant) = 0 or a small negative value.

[0054] Life_Cycle_Stage can represent the life cycle stage. It can be inferred based on the frequency and pattern of a vehicle's parking activities in the city. For example, by calculating the average number of parkings in the last 30 days, users of vehicles with an average number of parkings > 15 can be identified as high-frequency active users, users of vehicles with an average number of parkings greater than 10 but less than or equal to 10 can be identified as commuter users, and users of vehicles with an average number of parkings less than or equal to 5 can be identified as low-frequency users, and so on.

[0055] Furthermore, upon detecting that a target vehicle has entered a parking lot, the target vehicle identifier can be used to query the vehicle profile database. The target vehicle identifier can be a license plate number or other unique identifier for the target vehicle.

[0056] If the target vehicle identifier is found in the preset user profile database, the user profile corresponding to the target vehicle identifier in the database is determined as the target user profile. If the target vehicle identifier is not found in the preset user profile database, a target user profile is generated based on the vehicle information of the target vehicle and its parking information in the parking lot. The vehicle information may include the vehicle brand and class, license plate location, etc., and the parking information may include the type of parking lot, parking time, and parking duration, etc.

[0057] In some implementations, target user profiles can be determined using clustering algorithms based on vehicle and parking information. It should be noted that user profiles obtained through different acquisition methods share the same feature composition.

[0058] In this embodiment, target user profiles are obtained by querying a preset profile database and generating them in real time when no profile is queried. This not only allows for the rapid retrieval of existing accurate user profiles to improve the efficiency of parking intent recognition, but also covers new users, i.e. vehicles not yet entered into the profile database, thus avoiding the problem of not being able to achieve accurate push due to the inability to find user profiles.

[0059] In one embodiment of this application, different intent recognition models can be selected to recognize the parking intent of a vehicle based on the method of obtaining the target user profile. Specifically, step 101 may include the following: Based on the method of obtaining the target user profile, the target intent recognition model is determined; where the acquisition method is query acquisition, the target intent recognition model is a Markov model; where the acquisition method is generation acquisition, the target intent recognition model is a decision tree model. The target user profile is input into the target intent recognition model, which then outputs the target vehicle's parking intent.

[0060] In this embodiment, the target intent recognition model can be a Markov model or a decision tree model.

[0061] In some implementations, the input to the Markov model can be the contextual information of the current parking event and the target user profile. The contextual information may include the current parking time, current parking date, whether it is a holiday, parking lot type, and the distribution of POIs around the parking lot.

[0062] A Markov Model (HMM) can be constructed through the following steps. In some implementations, the true intention to park is considered as a hidden state S∈{going to work, going home, shopping, dining, entertainment, seeking medical treatment, picking up / dropping off, other}, and the observed parking behavior O (such as {parking lot type, time period, duration}) is generated from these hidden states. The HMM model λ = (A, B, π) is defined as follows: The state transition probability matrix A: a_ij = P(S_t+1 = j | S_t = i). This represents the transition probability between parking intentions. For example, the probability of transitioning from leaving get off work (S_t = leaving get off work) to going home (S_t+1 = going home) is high. This matrix can be obtained through statistical learning from long-term behavioral data of all vehicles.

[0063] The observation probability matrix B: b_j(k) = P(O_t = k | S_t = j). This represents the probability of generating a specific observation k under a specific intention j. For example, under the intention to dine, the probability of generating the observation that the parking lot type is a commercial area, the time period is 18:00-20:00, and the stay duration is 1.5 hours is very high.

[0064] Initial state probability π: π_i = P(S_1 = i). The probability distribution of the vehicle's parking intention when it first stops.

[0065] When a vehicle initiates a new parking event, the resulting observation sequence O_t can be a single observation. Using the Viterbi algorithm or the forward algorithm, the hidden state S_t that maximizes P(S_t | O_1...O_t, λ) under the current model λ and the historical observation sequence is calculated. This hidden state is the inferred parking intention Intent. Intent = argmax_{S} P(S | O_t, User_Profile, λ) In some implementations, the decision logic of a decision tree model can be: text IF Parking type == 'Hospital' AND Stay duration > 2h THEN Intent = 'Medical Treatment' ELSE IF Parking type == 'Office building' AND Time period in [08:00-10:00] AND Duration of stay > 6h THEN Intent = 'Going to work' ELSE IF Parking type == 'Residential' AND Time period in [18:00-21:00] THENIntent = 'Returning Home' ELSE IF Parking type == 'Business area' AND Time period in [11:00-13:00] AND Duration of stay in [1h, 2h] THEN Intent = 'Dining' ELSE IF Parking type == 'Business area' AND Time period in [18:00-21:00] AND Duration of stay in [1.5h, 3h] THEN Intent = 'Dining' ELSE IF Parking type == 'Business area' AND Time period in [13:00-17:00] AND Duration of stay in [1h, 3h] THEN Intent = 'Shopping' ELSE ... (default) = 'Other'.

[0066] Furthermore, a correspondence between acquisition methods and intent recognition models can be established in advance, and the target intent recognition model can be determined based on the acquisition method of the user profile.

[0067] If the acquisition method is query-based, meaning the target user profile is obtained from a pre-defined profile database, the target intent recognition model can be determined to be a Markov model. If the acquisition method is generation-based, meaning no vehicle matches the profile database, the target intent recognition model can be determined to be a decision tree model.

[0068] In other implementations, a correspondence between the user profile dimensions and the intent recognition model can be established in advance, and the profile dimensions can be determined based on the quantitative features in the profile.

[0069] For user profiles with a profile dimension greater than or equal to the preset profile dimension threshold, the target user profile contains a rich dynamic behavior matrix and a consumption preference vector that is continuously iterated based on historical redemption behavior. These data have significant temporal state transition characteristics, and the target intent recognition model can be determined to be a Markov model.

[0070] For user profiles with a profile dimension smaller than the preset profile dimension threshold, the user profile may only contain static attribute features such as vehicle brand and grade, lacking historical time-series behavior. For such high-dimensional and sparse discrete feature scenarios, the target intent recognition model can be determined as a decision tree model. The ability of the decision tree model to quickly classify based on a small number of strongly correlated static rules can be used, for example, by combining the parking lot type with the current time to directly derive branch conclusions for inference.

[0071] It should be noted that the preset profile dimension thresholds can be set according to actual conditions, and specific examples are not provided here. For different user profiles in the preset profile library, their profile dimensions can be the same or different.

[0072] Furthermore, after determining the target intent recognition model, the target user profile is input into the target intent recognition model, and the target intent recognition model outputs the parking intent of the target vehicle.

[0073] In this embodiment, different intent recognition models are adaptively selected based on the method of obtaining user profiles: for profiles with rich historical sequence information obtained from the profile database, a Markov model is used to capture the temporal transition pattern of user parking behavior to achieve high-precision intent inference; for profiles with fewer feature dimensions and lack of historical data generated temporarily, a decision tree model is used for interpretable classification with low overfitting risk to ensure stable output in data-sparse scenarios, thereby improving the accuracy and robustness of parking intent recognition.

[0074] In one embodiment of this application, candidate merchants can be matched according to the parking intention, and then the candidate merchants can be further filtered to obtain the target merchant. Specifically, step 102 may include the following: Based on the parking intention, candidate merchants are identified within the geographical area of ​​the business district corresponding to the parking lot; Based on the target user profile and the marketing information of each candidate merchant, determine the matching degree between each candidate merchant and the target vehicle; Target merchants are determined based on the degree of matching.

[0075] In this embodiment, all candidate merchants within the geographical area of ​​the business district corresponding to the parking lot can be identified as target merchants.

[0076] In some embodiments, a search can be performed within the geographical area of ​​the business district corresponding to the parking lot to determine candidate merchants that match the parking intention. For example, if the parking intention is to dine at a restaurant, all restaurants within the business district that belong to the food and beverage category can be selected as candidate merchants by comparing their category tags.

[0077] Furthermore, the matching degree between each candidate merchant and the target vehicle can be determined based on the target user profile and the marketing information of each candidate merchant. In some implementations, the matching degree of each candidate merchant can be obtained by matching relevant features in the target user profile, such as consumption preferences, consumption budget, and travel habits, with the marketing information of the candidate merchants. The marketing information may include the merchant's specific business category, real-time customer traffic, user reviews, and current promotional offers. For example, the matching degree can be calculated using a weighted summation method. For instance, the weight for a user profile's preference for hot pot is 0.4, and the score for the candidate merchant's main business being hot pot is 1. The weight multiplied by the score is included in the total matching degree. The weight for the consumption budget and merchant price matching dimension is 0.3, and the weight for user review fit is 0.3. The overall matching degree is then calculated.

[0078] In some implementations, candidate merchants with a matching degree greater than a preset threshold can be identified as target merchants. The preset threshold can be a pre-set fixed value, such as 0.8; or it can be determined based on the matching degree of each candidate merchant, for example, selecting the three candidate merchants with the highest matching degree as target merchants. No specific limitation is made here.

[0079] In some implementations, the candidate merchants can be sorted according to their matching degree, and the merchants with the highest matching degree among the top preset number can be selected. The preset number can be set according to the actual situation and is not specifically limited here. For example, the preset number can be set to the top 10% of merchants or the top three merchants.

[0080] In this embodiment, candidate merchants are determined based on user profiles. Then, the matching degree is calculated by combining the user profiles with the marketing information of each merchant, and a matching degree threshold is set for filtering. This makes the target merchants pushed in the end not only match the general category of parking intention, but also achieve a fine match between the user's personal preferences and the merchant's current promotional activities, which significantly improves the personalization of recommendations and conversion effect.

[0081] In one embodiment of this application, the marketing information includes business category, promotional information, and popularity information. Determining the matching degree between each candidate merchant and the target vehicle based on the target user profile and the marketing information of each candidate merchant may include the following: The correlation analysis between the consumption preference characteristics in the target user profile and the business categories of each candidate merchant is conducted to determine the preference matching score between each candidate merchant and the target user profile. Based on the promotional information, popularity information, and preference matching scores of each candidate merchant, the matching degree between each candidate merchant and the target vehicle is determined.

[0082] In this embodiment, the correlation analysis between the consumption preference features in the target user profile and the business categories of each merchant can be achieved by calculating the similarity between the consumption preference features in the target user profile and the business categories of each merchant. The similarity value can be used as the preference matching score between each candidate merchant and the target user profile.

[0083] Promotional information can include the merchant's real-time bidding information, discount details, spending threshold reduction activities, and free gifts. Popularity information can include the merchant's real-time traffic, user reviews and ratings, and recent exposure.

[0084] In some embodiments, promotional information and merchant popularity information may be normalized to convert them into a format that can be used to calculate preference matching scores.

[0085] Furthermore, the matching degree between each candidate merchant and the target vehicle can be determined by the following formula (3): Score_i= w1 *CosineSimilarity(User_Preference_Vector, M_i.category_vector) + w2 * σ(Bid_Price_i) + w3 * M_i.popularity + w4 * M_i.promotion_strength (3) In the formula, Score_i represents the matching degree, CosineSimilarity represents the preference matching score, σ(Bid_Price_i) represents the normalization of the real-time bidding information of merchant M_i so that its value range is between [0,1], M_i.popularity represents the merchant's popularity information, and M_i.promotion_strength represents the quantitative value of the merchant's current discount strength, such as the discount rate of "30 off for every 100 spent".

[0086] It should be noted that w1, w2, w3, and w4 represent learnable weight parameters, which can be adjusted according to the actual situation and updated based on subsequent user feedback.

[0087] In this embodiment, the matching degree of each candidate merchant is calculated based on multi-dimensional marketing information, which significantly improves the accuracy and real-time performance of personalized recommendations in parking lot scenarios.

[0088] In one embodiment of this application, before sending the target merchant's push information to the terminal device corresponding to the target vehicle, the information push method provided in this embodiment further includes: Based on the target user profile, determine the price sensitivity of the target vehicle's users; Based on the target merchant's basic discount range and price sensitivity, determine the personalized discount value for the target vehicle; Based on personalized discount values, push notifications are generated.

[0089] In this embodiment, price sensitivity can be determined based on information such as the frequency of users using vouchers in the target user profile, their reaction to discount rates, and average spending amount. It can also be determined based on information such as the brand and class of vehicles in the target user profile, or it can be obtained directly from the price-sensitive tags in the target user profile. No specific limitations are made here.

[0090] The basic discount can be preset by the merchant, such as an 8.8% discount on all items or a 30% discount for purchases over 200. This basic discount is then personalized based on the user's price sensitivity, generating a personalized discount value. In some implementations, higher price sensitivity results in a larger discount, while lower price sensitivity results in a smaller discount, i.e., only the basic discount or no additional discount is offered.

[0091] In some implementations, for users whose user profiles contain historical consumption information, the user's loyalty to the target merchant, or the loyalty to the target merchant or this type of merchant, can be calculated based on the number of times the user has visited or redeemed their accounts.

[0092] Furthermore, personalized discount values ​​are determined based on loyalty and price sensitivity.

[0093] For example, the personalized discount value can be determined by the following formula (4): V_coupon = f(User_Profile, Merchant_Profit_Model)(4) In the formula, V_coupon represents the personalized discount value, User_Profile represents the user profile, and Merchant_Profit_Model represents the merchant profit model. V_coupon should not be fixed, but should be a function of the user profile and the merchant profit.

[0094] Furthermore, the personalized discount value can also be determined using the following simplified formula (5): V_coupon = Base_Discount*(1+β1*Sensitivity_Price +β2 * Loyalty_Score) (5) In the formula, Base_Discount represents the merchant's basic discount rate, Sensitivity_Price represents the user's price sensitivity, Loyalty_Score represents the user's loyalty to the merchant or the type of merchant, and β1 and β2 represent weighting coefficients.

[0095] By determining personalized discount values ​​based on price sensitivity and loyalty, price-sensitive users may receive higher-value coupons, while loyal users may receive more distinctive, non-price-based benefits such as priority queuing or free side dishes, thereby improving user experience and conversion rates.

[0096] In some implementations, push notifications can be generated based on personalized offers combined with the user's price sensitivity. For example, for users with high price sensitivity, the emphasis can be placed on the discount and cost-effectiveness; for users with low price sensitivity, the focus can be on the merchant's quality and service advantages, while simultaneously highlighting personalized offers, thus balancing practicality and user experience.

[0097] In this embodiment, by extracting price sensitivity from user profiles and dynamically calculating personalized discount values ​​in conjunction with the merchant's basic discount range, larger discounts are offered to price-sensitive users to promote conversion, while moderate discounts are offered to price-insensitive users to ensure merchant profits. This achieves precise allocation of discount resources and further improves the effectiveness of push notifications and overall revenue.

[0098] In one embodiment of this application, a push strategy for the push information can also be determined based on the user type of the target vehicle. Specifically, step 103 may include the following: In response to the target vehicle parking more times than the preset parking number threshold, the user type corresponding to the target vehicle is determined based on the target vehicle's historical consumption information within the geographical area of ​​the parking lot's business district. A push strategy corresponding to user type is adopted to send push information of the target merchant to the terminal device corresponding to the target vehicle.

[0099] In this embodiment, the number of times the target vehicle parks in the parking lot can be obtained through the parking management system. The statistical period for the number of parking times can be set according to the actual application scenario, such as the past 30 days or the past 90 days. The preset parking time threshold can be adjusted according to the usage scenario of the parking lot. For example, the preset threshold for a commercial parking lot is 5 times, and the preset threshold for an office area parking lot is 10 times.

[0100] If the number of parking attempts is less than or equal to the parking attempt threshold, the user type of the target vehicle can be identified as a low-frequency user or a new user, and the default push strategy can be directly adopted, such as pushing information when the vehicle is parked.

[0101] If the number of parking attempts exceeds the parking attempt threshold, the user type can be further determined based on historical consumption information, such as consumption frequency and recent consumption time.

[0102] Furthermore, push strategies can be determined based on user types. Push strategies can include the frequency of pushes, the timing of pushes, and increasing discounts based on the content determined by user profiles.

[0103] In some implementations, user types may include high-frequency active users, commuting users, low-frequency inactive users, and users at risk of churn.

[0104] For example, users with a cumulative spending amount greater than 2,000 yuan and a high frequency of recent purchases (e.g., a purchase record after parking exceeding 70%) can be identified as high-frequency active users. The push notification strategy for high-frequency active users can be to send push notifications at a moderate frequency, such as every time the car parks.

[0105] Users with low purchase frequency (e.g., less than 30% of them have a purchase record after parking) and whose parking times are concentrated during weekday morning / evening rush hours or short lunch breaks can be identified as commuter users. The corresponding push notification strategy for commuter users can be to send push notifications during weekday morning / evening rush hours.

[0106] Users who have had no recent spending history (e.g., a week) or whose spending is significantly below average can be identified as low-frequency, inactive users. The corresponding push strategy for low-frequency, inactive users can be to increase incentives and discounts, based on the content determined according to their user profile.

[0107] Users with high parking frequency can be categorized by frequency. For example, users whose parking frequency has decreased by more than 50% month-on-month in the past month, and whose most recent parking occurred more than a month ago without any purchases, are identified as users at risk of churn. The corresponding push notification strategy for users at risk of churn can be reactivation offers. In addition to determining the push content based on user profiles, this strategy can include adding exclusive return offers and other methods to reactivate users.

[0108] In this embodiment, the push strategy is accurately matched based on the user's actual parking and consumption information, thereby effectively improving the conversion rate of different types of users.

[0109] In one embodiment of this application, after step 103, the information push method provided in this embodiment may further include the following: In response to the feedback event from the user of the target vehicle regarding the push notification, the information used to determine the push notification is updated using the update strategy corresponding to the feedback event.

[0110] In this embodiment, feedback events generated by the user in the target vehicle in response to the pushed information can be continuously received through a user terminal application or an in-vehicle terminal. Feedback events can take various forms, such as a click event where the user triggers a click to view merchant details, a save event where a coupon is saved to the wallet, a real in-store redemption event recorded based on location or Bluetooth beacon data, or an ignored event where the user does not interact with the information within a preset time window after it is sent.

[0111] Furthermore, in response to the different feedback events received, an update strategy corresponding to the type of feedback event can be adopted to accurately backtrack and correct the information used to determine the current push notification. The information used to determine the push notification can be at least one of the following: target user profile, parameters of the intent inference model, and weights of each dimension in the matching degree calculation.

[0112] In some implementations, the consumption preference vector in the user profile can be updated based on the redemption behavior.

[0113] New observation data, such as parking behavior after redemption, can be used as weak labels to fine-tune the observation probability matrix of the HMM. For example, if it is inferred that the user dined and was pushed a Sichuan restaurant coupon and redeemed it, the weight of the observation-state pair (O_t, S_t=dining) can be increased.

[0114] Push notifications and conversion results can be used as feedback signals to adjust the weights w1, w2, w3, and w4 in the matching formula in real time through online learning algorithms, such as online gradient descent. The optimization goal is to maximize a certain business metric, such as click-through rate or conversion rate. For example, if it is found that the contribution of merchant popularity to conversion has decreased recently, while the contribution of bidding has increased, w2 can be automatically increased and w3 decreased.

[0115] For example, when the feedback event is a positive in-store redemption event, a positive reinforcement update strategy can be triggered. In the consumer preference vector update algorithm of the user profile, the feedback value of the target merchant's category can be set as a positive incentive, thereby increasing the preference score parameter of that dimension; at the same time, the weight parameters of the corresponding features in the intent recognition model can be fine-tuned to consolidate the successful matching logic.

[0116] When the feedback event is a negative ignored event or a push notification that is not redeemed, a negative penalty or boundary correction update strategy can be triggered. Negative feedback values ​​can be applied to the preference vector algorithm to lower the corresponding category score. Furthermore, macro-level spatial or threshold parameters can be adjusted based on the actual situation. For example, if the reason for non-redemption is determined to be that the distance is too far, causing the user to abandon walking, the maximum acceptable walking time parameter in the business district geographical range calculation formula can be dynamically lowered, thereby shrinking the effective radius of the next push notification. Alternatively, the preset matching threshold parameter when subsequently filtering target merchants can be appropriately increased to raise the push threshold.

[0117] In this implementation, the parameter update step transforms every actual user interaction into data nourishment for optimizing the underlying algorithm model. This allows the preceding steps, such as intent prediction, target merchant matching, and personalized discount calculation, to continuously self-correct and iterate over long-term operation, ultimately achieving a sustained spiral increase in the accuracy and conversion rate of business district marketing.

[0118] It should be noted that the various embodiments described in this application can be combined with each other or implemented individually without conflict, and this application does not limit this.

[0119] It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other.

[0120] For ease of understanding, a specific embodiment will be used as an example: (1) [Event Trigger] Vehicle Car_i enters parking lot P at time T.

[0121] (2) [Data Acquisition] The system collects the entry record for this entry: (plate_id, T, P, space_type). plate_id is the vehicle identifier, and space_type is the parking lot type.

[0122] (3) [Portrait Loading] Load the user profile User_Profile_i from the profile library using plate_id as the key (if not available, generate the initial default user profile based on vehicle information and parking information in the parking lot).

[0123] (4) [Intent Inference] Input (T, P, User_Profile_i) into the intent inference model (HMM or rule tree) and output the parking intent Intent.

[0124] (5) [POI Recall] Recall the set of candidate merchants {M} that match the Intent category within a radius R_walkable centered on the parking lot P.

[0125] (6) [Real-time bidding] Initiate a bidding request to the merchants in the set {M} and obtain their real-time bid Bid_Price and basic discount information Base_Coupon.

[0126] (7) [Personalized scoring] For each candidate merchant M_j, calculate its score, i.e. matching degree, according to the above formula (3).

[0127] (8) [Merchant Ranking] Sort the candidate merchants in descending order according to the score and select the top K.

[0128] (9) [Dynamic Discount Generation] For the top K merchants, personalized coupons can be generated using the dynamic discount generation algorithm of formula (4) above.

[0129] (10) [Multi-channel push] Push information containing {merchant, coupon} to the user's Car_i terminal via APP, SMS or partner mini-program.

[0130] (11) [User behavior collection] 11a. [Click / Ignore] Record whether the user clicked.

[0131] 11b. [Redemption Verification] Obtain the redemption status (Redemption_Status) of coupons through periodic polling or proactive reporting by merchants.

[0132] (12) [Effect Evaluation] Clicks, cancellations, and other behaviors are used as positive / negative feedback signals (Reward).

[0133] (13) [Online Model Update] 13a. Update the preference vector of User_Profile_i with Reward.

[0134] 13b. Update the rating weight w and intent inference model parameters with (Intent, push record, Reward) data.

[0135] (14) [End] The process ends, waiting for the next event.

[0136] like Figure 3 As shown, Figure 3 This application provides an embodiment of an interactive diagram illustrating the applicability of the information push method to a system, which may include a user terminal, a parking system, a profile service, a business engine, and a merchant system. Specifically, the information push method provided in this application embodiment can be applied to a business engine. The user terminal sends a vehicle entry identification signal to the parking system, triggering the parking system to start the information push process.

[0137] The parking system sends a request to the profiling service, carrying the plate_id (vehicle identifier) ​​of the target vehicle, in order to obtain the target user profile for that vehicle.

[0138] The profile service queries the profile database based on plate_id and returns the target user profile of the target vehicle to the parking system.

[0139] The parking system infers and generates a contextual profile based on the target user profile, which is a real-time profile combined with the current parking scenario. The contextual profile is then sent to the business engine to request business district calculation.

[0140] After receiving the contextual profile, the business engine performs the following sub-steps: a. Recall POIs (Points of Interest) and local data marts (i.e., filter candidate merchants from the geographical area of ​​the business district corresponding to the parking lot); b. Conduct batch bidding on candidate merchants (in conjunction with the merchants' real-time bidding strategies). c. Generate coupons (i.e., coupons corresponding to personalized discount values) based on the bidding results and merchant promotion information; d. Return the generated coupons and push notifications to the parking system.

[0141] The parking system receives push content returned by the business engine and pushes push information of the target merchants to the user terminal.

[0142] The user terminal responds to user actions and sends feedback events to the business engine.

[0143] Based on feedback events, the business engine adopts corresponding update strategies to update the parameters used to determine the content to be pushed (such as updating the consumption preference vector in the user profile, optimizing the intent recognition model or the business district calculation model).

[0144] In some embodiments, the information push method provided in this application can be applied to an information push system. The information push system may include a data acquisition layer, a data processing layer, an application service layer, and a user terminal service layer. Specifically: Data Acquisition Layer: Used to acquire heterogeneous data from multiple sources.

[0145] Data processing layer: Used for preprocessing multi-source heterogeneous data.

[0146] Application Service Layer: This layer executes core business logic based on preprocessed data output from the data processing layer. This includes user profile building, parking intent recognition, merchant matching and push strategy generation, and feedback event handling. Specifically: User profile construction sub-layer: Based on the preprocessed multi-source data, construct a multi-dimensional vectorized profile of the target user.

[0147] Parking Intent Recognition Sublayer: Based on the method of obtaining the user profile, adaptively select the intent recognition model (Markov model or decision tree model) and output the parking intent; Merchant matching and push strategy sub-layer: Based on parking intention, candidate merchants are filtered within the geographical area of ​​the business district. The matching degree is calculated by combining user preferences, merchant discount information, and popularity information. Merchants with a matching degree greater than the threshold are selected as target merchants. Based on the price sensitivity in the target user profile, personalized discount values ​​are generated and a differentiated push strategy is adopted. Feedback event handling sublayer: Receives feedback events transmitted from the client service layer, adjusts parameters using corresponding update strategies, and achieves iterative optimization of the model.

[0148] User-side service layer: Used to send push information from the target merchant to the terminal device corresponding to the target vehicle, and to receive user feedback events on the push information, and transmit the feedback events to the application service layer; wherein, the push information includes the target merchant's marketing information and personalized discount value, and the terminal device generates feedback events in response to user operation.

[0149] like Figure 4 As shown, Figure 4 The second flowchart illustrating the information push method provided in this application embodiment is as follows: New User Phase: When the target vehicle parks for the first time, the new user process is triggered, which includes: Collect basic data: Obtain basic attribute data such as vehicle brand, model, and license plate location from data sources such as parking management systems and license plate recognition systems; Establish initial profile: Based on basic data, an initial user profile is constructed through preset static attribute mapping rules, which serves as the initial basis for subsequent push notifications.

[0150] Multiple parking cycle phase: When the target vehicle stops multiple times, a cycle process begins, which specifically includes: Parking event triggered: When a vehicle entry / exit event is detected, the process for the current parking cycle is initiated; Intent Inference: Based on the current parking scenario and historical user profiles, infer the user's parking intent; Merchant matching: Within the geographical area of ​​the business district corresponding to the parking lot, filter candidate merchants that match the user's intent and calculate the matching degree; Discount Push: Generate personalized discount values ​​based on price sensitivity in user profiles and push push information containing target merchants and coupons to user terminals; User feedback: Receive user feedback on push notifications; User profile update: Based on feedback events, adopt corresponding update strategies to update user profiles and iteratively optimize the model.

[0151] Mature User Stage: After multiple parking cycles, users enter the mature user stage based on their profile maturity (based on a parking frequency threshold, such as parking frequency ≥ N times). Users are then categorized into different types based on their parking behavior characteristics, and a differentiated push strategy is adopted: where N is a positive integer.

[0152] For high-frequency active users: push notifications at a moderate frequency to avoid excessive disturbance; Commuters: Push notifications are sent during weekday morning and evening rush hours to match their travel patterns; For low-frequency users: increase incentives and boost conversion rates; For users at risk of churn: Push high-value coupons to win them back.

[0153] Based on the information push method provided in the above embodiments, this application also provides specific implementations of the information push device. Please refer to the following embodiments.

[0154] See Figure 5 The information push device 300 provided in this application embodiment may include: The determination module is used to determine the parking intention of a target vehicle in response to the acquisition of a target vehicle entering the parking lot, based on the target user profile of the target vehicle. The matching module is used to determine the target merchants within the geographical area of ​​the business district corresponding to the parking lot based on the parking intention; The push module is used to send push information from the target merchant to the terminal device corresponding to the target vehicle.

[0155] In one embodiment of this application, the target user profile is obtained through the following steps: Based on the target vehicle's identifier, query the preset profile database; the preset profile database stores the vehicle identifiers of each vehicle and the user profiles corresponding to the vehicle identifiers. In response to finding the target vehicle identifier in the preset profile database, the user profile corresponding to the target vehicle identifier in the preset profile database is determined as the target user profile; In response to the failure to find the target vehicle identifier in the preset profile database, a target user profile is generated based on the target vehicle's vehicle information and its parking information in the parking lot.

[0156] In one embodiment of this application, the determining module may include: The first determination submodule is used to determine the target intent recognition model based on the acquisition method of the target user profile; the target intent recognition model is either a Markov model or a decision tree model; wherein, when the acquisition method is query acquisition, the target intent recognition model is a Markov model; when the acquisition method is generation acquisition, the target intent recognition model is a decision tree model. The intent output submodule is used to input the target user profile into the target intent recognition model, and the target intent recognition model outputs the parking intent of the target vehicle.

[0157] In one embodiment of this application, the matching module may include: The second determination submodule is used to determine each candidate merchant within the geographical area of ​​the business district corresponding to the parking lot based on the parking intention; The third determination submodule is used to determine the matching degree between each candidate merchant and the target vehicle based on the target user profile and the marketing information of each candidate merchant; The fourth submodule is used to determine the target merchant based on the matching degree.

[0158] In one embodiment of this application, the marketing information includes business category, promotional information, and popularity information; The third determination submodule includes: The first determining unit is used to analyze the correlation between the consumption preference characteristics in the target user profile and the business categories of each candidate merchant, and to determine the preference matching score between each candidate merchant and the target user profile. The second determining unit is used to determine the matching degree between each candidate merchant and the target vehicle based on the promotional information, popularity information, and preference matching score of each candidate merchant.

[0159] In one embodiment of this application, the information push device 300 may further include: Based on the target user profile, determine the price sensitivity of the target vehicle's users; Based on the target merchant's basic discount range and price sensitivity, determine the personalized discount value for the target vehicle; Based on personalized discount values, push notifications are generated.

[0160] In one embodiment of this application, the push module may include: The fifth determination submodule is used to determine the user type corresponding to the target vehicle based on the target vehicle's historical consumption information within the geographical area of ​​the parking lot when the number of times the target vehicle parks in the parking lot exceeds a preset parking number threshold. The strategy push submodule is used to send push information from the target merchant to the terminal device corresponding to the target vehicle using a push strategy that corresponds to the user type.

[0161] In one embodiment of this application, the information push device 300 may further include: The update module is used to respond to user feedback events for push information received from the target vehicle, and to update the information used to determine the push information using the update strategy corresponding to the feedback event.

[0162] The information push device provided in this application embodiment can implement all the processes implemented in the aforementioned information push method embodiment and achieve the same technical effect. To avoid repetition, it will not be described again here.

[0163] Figure 6 A schematic diagram of the hardware structure of the information push device provided in an embodiment of this application is shown.

[0164] The information push device may include a processor 401 and a memory 402 storing computer program instructions.

[0165] Specifically, the processor 401 may include a central processing unit (CPU), an application-specific integrated circuit (ASIC), or one or more integrated circuits that can be configured to implement the embodiments of this application.

[0166] Memory 402 may include mass storage for data or instructions. For example, and not limitingly, memory 402 may include a hard disk drive (HDD), floppy disk drive, flash memory, optical disk, magneto-optical disk, magnetic tape, or Universal Serial Bus (USB) drive, or a combination of two or more of these. Where appropriate, memory 402 may include removable or non-removable (or fixed) media. Where appropriate, memory 402 may be internal or external to the integrated gateway disaster recovery device. In a particular embodiment, memory 402 is non-volatile solid-state memory.

[0167] Memory may include read-only memory (ROM), random access memory (RAM), disk storage media devices, optical storage media devices, flash memory devices, and electrical, optical, or other physical / tangible memory storage devices. Therefore, typically, memory includes one or more tangible (non-transitory) computer-readable storage media (e.g., memory devices) encoded with software including computer-executable instructions, and when the software is executed (e.g., by one or more processors), it is operable to perform the operations described with reference to the methods according to one aspect of this disclosure.

[0168] The processor 401 reads and executes computer program instructions stored in the memory 402 to implement any of the information push methods in the above embodiments.

[0169] In some embodiments, the information push device can be a vehicle. In these embodiments, the processor 401 implements the above by reading and executing computer program instructions stored in the memory 402. Figure 2 Any of the information push methods in the method embodiments.

[0170] In one example, the information push device may also include a communication interface 403 and a bus 410. For example, Figure 6 As shown, the processor 401, memory 402, and communication interface 403 are connected through bus 410 and complete communication with each other.

[0171] The communication interface 403 is mainly used to realize communication between various modules, devices, units and / or equipment in the embodiments of this application.

[0172] Bus 410 includes hardware, software, or both, that couples the components of the information pushing device together. This is an example, not a limitation. The bus may include an Accelerated Graphics Port (AGP) or other graphics bus, an Extended Industry Standard Architecture (EISA) bus, a Front Side Bus (FSB), a Hyper Transport (HT) interconnect, an Industry Standard Architecture (ISA) bus, an Infinite Bandwidth Interconnect, a Low Pin Count (LPC) bus, a memory bus, a Micro Channel Architecture (MCA) bus, a Peripheral Component Interconnect (PCI) bus, a Serial Advanced Technology Attachment (SATA) bus, a Video Electronics Standards Association Local Bus (VLB) bus, or other suitable buses, or a combination of two or more of these. Where appropriate, bus 410 may include one or more buses. Although specific buses are described and illustrated in embodiments of this application, this application contemplates any suitable bus or interconnect.

[0173] Furthermore, in conjunction with the information push methods in the above embodiments, this application embodiment can provide a computer storage medium for implementation. The computer storage medium stores computer program instructions; when these computer program instructions are executed by a processor, they implement any of the information push methods in the above embodiments.

[0174] This application embodiment may also provide a computer program product, wherein when the instructions in the computer program product are executed by the processor of the information push device, the information push device performs any of the information push methods in the above embodiments.

[0175] It should be clarified that this application is not limited to the specific configurations and processes described above and shown in the figures. For the sake of brevity, detailed descriptions of known methods are omitted here. In the above embodiments, several specific steps are described and shown as examples. However, the method process of this application is not limited to the specific steps described and shown. Those skilled in the art can make various changes, modifications, and additions, or change the order of steps, after understanding the spirit of this application.

[0176] The functional blocks shown in the above-described block diagram can be implemented as hardware, software, firmware, or a combination thereof. When implemented in hardware, they can be, for example, electronic circuits, application-specific integrated circuits (ASICs), appropriate firmware, plug-ins, function cards, etc. When implemented in software, the elements of this application are programs or code segments used to perform the required tasks. The programs or code segments can be stored on a machine-readable medium or transmitted over a transmission medium or communication link via data signals carried on a carrier wave. "Machine-readable medium" can include any medium capable of storing or transmitting information. Examples of machine-readable media include electronic circuits, semiconductor memory devices, ROM, flash memory, erasable ROM, floppy disks, CD-ROMs, optical disks, hard disks, fiber optic media, radio frequency links, etc. Code segments can be downloaded via computer networks such as the Internet, intranets, etc.

[0177] It should also be noted that the exemplary embodiments mentioned in this application describe methods or systems based on a series of steps or apparatus. However, this application is not limited to the order of the above steps; that is, the steps can be performed in the order mentioned in the embodiments, or in a different order, or several steps can be performed simultaneously.

[0178] The aspects of this disclosure have been described above with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this disclosure. It should be understood that each block in the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing apparatus to produce a machine such that these instructions, executable via the processor of the computer or other programmable data processing apparatus, enable the implementation of the functions / actions specified in one or more blocks of the flowchart illustrations and / or block diagrams. Such a processor can be, but is not limited to, a general-purpose processor, a special-purpose processor, a special application processor, or a field-programmable logic circuit. It is also understood that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can also be implemented by special-purpose hardware performing the specified functions or actions, or can be implemented by a combination of special-purpose hardware and computer instructions.

[0179] The above description is merely a specific implementation of this application. Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, modules, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here. It should be understood that the protection scope of this application is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in this application, and these modifications or substitutions should all be covered within the protection scope of this application.

Claims

1. An information push method, characterized in that, The method includes: In response to the acquisition of a target vehicle entering the parking lot, the parking intention of the target vehicle is determined based on the target user profile of the target vehicle; Based on the parking intention, the target merchants are determined within the geographical area of ​​the business district corresponding to the parking lot; Send the target merchant's push information to the terminal device corresponding to the target vehicle.

2. The method according to claim 1, characterized in that, The target user profile is obtained through the following steps: Based on the target vehicle identifier of the target vehicle, query the preset profile database; wherein, the preset profile database stores the vehicle identifier of each vehicle and the user profile corresponding to the vehicle identifier; In response to finding the target vehicle identifier in the preset profile database, the user profile corresponding to the target vehicle identifier in the preset profile database is determined as the target user profile; In response to the fact that the target vehicle identifier is not found in the preset profile database, the target user profile is generated based on the vehicle information of the target vehicle and the parking information in the parking lot.

3. The method according to claim 1, characterized in that, Determining the parking intention of the target vehicle based on the target user profile includes: Based on the acquisition method of the target user profile, a target intent recognition model is determined; wherein, when the acquisition method is query acquisition, the target intent recognition model is a Markov model; when the acquisition method is generation acquisition, the target intent recognition model is a decision tree model. The target user profile is input into the target intent recognition model, and the target intent recognition model outputs the parking intent of the target vehicle.

4. The method according to claim 1, characterized in that, The step of determining the target merchants matching the parking intention within the geographical area of ​​the business district corresponding to the parking lot includes: Based on the parking intention, each candidate merchant is determined within the geographical area of ​​the business district corresponding to the parking lot; Based on the target user profile and the marketing information of each candidate merchant, determine the matching degree between each candidate merchant and the target vehicle; The target merchant is determined based on the matching degree.

5. The method according to claim 4, characterized in that, The marketing information includes business categories, promotional information, and popularity information; The step of determining the matching degree between each candidate merchant and the target vehicle based on the target user profile and the marketing information of each candidate merchant includes: The correlation analysis is performed between the consumption preference features in the target user profile and the business categories of each candidate merchant to determine the preference matching score between each candidate merchant and the target user profile. Based on the promotional information, popularity information, and preference matching score of each candidate merchant, the matching degree between each candidate merchant and the target vehicle is determined.

6. The method according to claim 1, characterized in that, Before sending the push information of the target merchant to the terminal device corresponding to the target vehicle, the method further includes: Based on the target user profile, determine the price sensitivity of the target vehicle's users; Based on the target merchant's basic discount rate and the price sensitivity, determine the personalized discount value for the target vehicle; The push notification is generated based on the personalized discount value.

7. The method according to claim 1, characterized in that, Sending the push information of the target merchant to the terminal device corresponding to the target vehicle includes: In response to the target vehicle parking more times in the parking lot than a preset parking number threshold, the user type corresponding to the target vehicle is determined based on the historical consumption information of the target vehicle within the geographical area of ​​the parking lot's business district. Using a push strategy corresponding to the user type, push information of the target merchant is sent to the terminal device corresponding to the target vehicle.

8. The method according to claim 1, characterized in that, After sending the push information of the target merchant to the terminal device corresponding to the target vehicle, the method further includes: In response to a feedback event received from the user of the target vehicle regarding the push notification, the information used to determine the push notification is updated using the update strategy corresponding to the feedback event.

9. An information push device, characterized in that, The device includes: The determination module is used to determine the parking intention of the target vehicle based on the target user profile of the target vehicle in response to the acquisition of the target vehicle entering the parking lot. The matching module is used to determine the target merchants that match the parking intention within the geographical area of ​​the business district corresponding to the parking lot, based on the parking intention. The push module is used to send push information of the target merchant to the terminal device corresponding to the target vehicle.

10. An information push device, characterized in that, The information push device includes: a processor and a memory storing computer program instructions; When the processor executes the computer program instructions, it implements the information push method as described in any one of claims 1-8.

Images