Vehicle rental information recommendation and reward settlement method and system based on user portrait
By building user portraits and dynamically adjusting reward mechanisms, the problem that the existing vehicle rental recommendation system cannot accurately capture user needs is solved, the degree of personalization and user experience of recommendations is improved, and a more accurate recommendation and incentive mechanism is achieved.
Patent Information
- Application Number
- CN202411406347.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-10-10
- Publication Date
- 2024-11-08
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
The existing vehicle rental recommendation system cannot accurately capture the user's personalized needs, resulting in insufficient recommendation results, poor user experience, and the reward mechanism cannot be personalized according to different user behaviors.
By obtaining the user's personal information and historical leasing behavior data, building user portraits, analyzing user preference tags, filtering out vehicle information that meets user preference tags from the vehicle database, and using the recommendation algorithm to generate a recommendation list, and at the same time, based on the user's leasing behavior records, Use preset reward rules to calculate reward points and dynamically adjust the user's preference tags and reward mechanisms.
It improves the personalization and user experience of vehicle rental recommendations, ensures that the recommendation results are more in line with the actual needs of users, and encourages user activity and loyalty through a dynamic reward mechanism.
Smart Images

Figure CN118916552A_ABST
Abstract
Description
Technical Field
[0001] The embodiments of the present application relate to the field of data processing technology, and in particular to a method and system for recommending vehicle rental information and settling rewards based on user portraits. Background Art
[0002] With the development of Internet technology and the rise of the sharing economy, vehicle rental services have become an important part of people's daily travel. Users expect to get a more personalized service experience when choosing a rental vehicle, and hope to get more rewards through their positive behavior. In vehicle rental services, how to recommend suitable vehicles to users based on their historical behavior and preference tags, and encourage user activity and loyalty through an effective reward mechanism has become an important technical requirement. This requires the rental service platform to not only have strong data analysis capabilities, but also be able to update recommendation strategies and reward rules in real time.
[0003] Most existing vehicle rental recommendation systems make recommendations based on simple user behavior data, such as geographic location and rental history. The reward mechanism is also mostly a fixed point system, which cannot be personalized according to different user behaviors. Although this recommendation and service method can basically meet the rental needs of users, it is insufficient in terms of refined management and personalized services.
[0004] However, due to the lack of in-depth exploration of user portraits, existing recommendation systems often fail to accurately capture the real needs of users, resulting in insufficient personalization of recommendation results and poor user experience. Summary of the invention
[0005] The embodiments of the present application provide a method and system for recommending vehicle rental information and settling rewards based on user portraits, so as to solve the problem of inaccurate recommendation mechanism in the prior art.
[0006] In a first aspect, an embodiment of the present application provides a vehicle rental information recommendation and reward settlement method based on user portrait, comprising the following steps: Obtain the user's personal information and build a user profile based on historical rental behavior data; Analyze the user's preference tags according to the user portrait, and filter out vehicle information that matches the user's preference tags from a vehicle database; Generate a recommendation list of the vehicle information that meets the user's preference tags through a recommendation algorithm and push it to the user; Record the user's feedback on the recommendation list, update the user portrait, and adjust the recommendation strategy; According to the rental behavior record of the user, the reward points of the user are calculated using the preset reward rules, and the user is given a corresponding reward according to the reward points.
[0007] Optionally, analyzing the user's preference tags according to the user portrait includes: Classify the historical rental records of the user and identify the user's usage habits and preference tag types; Use machine learning algorithms to train models that can predict users’ future rental behavior; According to the result predicted by the model, the preference label of the user is adjusted.
[0008] Optionally, the calculating the reward points of the user according to the rental behavior record of the user using a preset reward rule includes: Different points values are set corresponding to different rental behaviors, and the rental behaviors include rental duration and rental frequency; Establishing a user points pool and dynamically adjusting the size of the points pool based on the activity level of the user; According to the rental behavior record of the user, the corresponding reward points are calculated, and the reward points are added to the point pool of the user to determine the final reward points of the user.
[0009] Optionally, the reward points of the user are calculated using a preset reward rule, and the specific calculation formula is as follows:
[0010] in, Indicates The increment of reward points earned by each user; Indicates the total number of rental behavior records of the user; Indicates the total number of vehicle types; Indicates The duration of the lease; Represents the average of all rental durations; Indicates The rental frequency of each user; represents the average rental frequency of all users; Indicates driving distance of each user; represents the average driving distance of all users; Indicates Number of rentals of this type of vehicle; Indicates the average number of rentals for each type of vehicle; They represent the weight coefficients of rental duration, rental frequency, driving distance, and number of rentals respectively.
[0011] Optionally, the reward includes a total amount of free rental time obtained by the user; Giving the user a corresponding reward according to the reward points includes: The user is given a corresponding reward according to the reward points through the following formula:
[0012] in, Indicates the total amount of free rental time obtained by the user, Indicates a round-down operation; Indicates the total number of reward points in the user's point pool; is the time decay factor of the bonus points; is the user’s credit score; It is a quantitative indicator of the user’s social network influence; is the preference label coefficient of the user's rental behavior in different time periods; is the frequency of user interaction on social media; It is the time series characteristic of leasing behavior; Indicates the length of rental time that a unit of points can be redeemed for; is the user's historical behavior score.
[0013] Optionally, giving the user a corresponding reward according to the reward points further includes: Calculating the effective value of the reward points, and giving the user a corresponding reward based on the effective value; The effective value of the reward points is calculated by the following formula:
[0014] in, Indicates the valid value of reward points; Indicates the final reward points; It is a coefficient adjusted according to user activity, ranging from 0 to 1; It is a coefficient adjusted according to the user's social influence, ranging from 0 to 1; is the user's preference label coefficient for different types of vehicles; It is the user’s loyalty to any one brand; is the trend of user activity over time; Also includes: Dynamic adjustment using Bayesian update rule and to update the effective value of the reward points; Among them, the following formula is used to dynamically adjust and Values: in, and They are The user's preference label coefficients for different types of vehicles and the user's loyalty to any brand after the update; and It is the actual obtained user preference label coefficients for different types of vehicles and the user's loyalty to any brand; and is the smoothing parameter; Indicates The user's preference label coefficients for different types of vehicles after the update; Indicates The user's loyalty to any brand after the update.
[0015] Optionally, the calculating the reward points of the user according to the rental behavior record of the user using a preset reward rule includes: Updating a user preference tag, and adjusting the user's reward points according to the user preference tag; The updating of the user preference tag and adjusting the reward points of the user according to the user preference tag includes: Based on the Bayesian formula, using the user's rental behavior record Dynamically update the user's preferences for multiple vehicle types Preferences tab , The calculation formula of the Bayesian formula includes:
[0016] in, Indicates the updated preference label, which is the observed rental behavior record In the case of user preference label vehicle type probability; Indicates the user's preference label vehicle type Under the assumption that probability; is the prior probability, indicating that without considering any rental behavior records Previously, the user preferred the label vehicle type probability; Indicates all assumptions Evidence The weighted sum of the probabilities of , namely the marginal probability, is used as a normalization factor to ensure The value of is between 0 and 1; Take advantage of the updated preference tab adjusting the user's reward points; Among them, the adjustment formula is defined as:
[0017] in, represents the increment of the adjusted reward points; Indicates the increment of reward points before adjustment; It is the adjustment coefficient for each vehicle type, which is used to control the impact of changes in user preference labels on reward points; The total number of different types of vehicles representing user preference labels; Indicates the updated preference label; represents the prior probability; is the importance weight of each vehicle type.
[0018] In a second aspect, the embodiment of the present application provides a vehicle rental information recommendation and reward settlement system based on user portraits, including: Acquisition building module, used to obtain the user's personal information and build a user portrait based on historical rental behavior data; An analysis and screening module, used to analyze the user's preference tags according to the user portrait, and screen out vehicle information that meets the user's preference tags from the vehicle database; A push module is used to generate a recommendation list of the vehicle information that meets the user's preference tags through a recommendation algorithm and push it to the user; An updating and adjusting module, used to record the user's feedback behavior on the recommendation list, update the user portrait, and adjust the recommendation strategy; A calculation and granting module, used to calculate the user's reward points according to the user's rental behavior record and using a preset reward rule, and grant the user a corresponding reward according to the reward points; The calculation and granting module is specifically used to set different point values corresponding to different rental behaviors, and the rental behaviors include rental duration and rental frequency; establish a user point pool, and dynamically adjust the size of the point pool according to the user's activity level; calculate the corresponding reward points based on the user's rental behavior record, and add the reward points to the user's point pool to determine the user's final reward points.
[0019] In a third aspect, an embodiment of the present application provides a computing device, comprising a processing component and a storage component; the storage component stores one or more computer instructions; the one or more computer instructions are used to be called and executed by the processing component to implement a vehicle rental information recommendation and reward settlement method based on user portrait as described in any one of the first aspects.
[0020] In a fourth aspect, an embodiment of the present application provides a computer storage medium storing a computer program. When the computer program is executed by a computer, it implements a vehicle rental information recommendation and reward settlement method based on user portrait as described in any one of the first aspects.
[0021] In the embodiment of the present application, the user's personal information is obtained, and a user portrait is constructed in combination with historical rental behavior data; the user's preference tags are analyzed according to the user portrait, and vehicle information that meets the user's preference tags is screened out from the vehicle database; the vehicle information that meets the user's preference tags is generated into a recommendation list through a recommendation algorithm, and pushed to the user; the user's feedback behavior on the recommendation list is recorded, the user portrait is updated, and the recommendation strategy is adjusted; according to the user's rental behavior record, the user's reward points are calculated using preset reward rules, and the user is given corresponding rewards according to the reward points. The technical solution provided by the present application improves user experience and service quality.
[0022] These and other aspects of the present application will become more clearly understood in the description of the following embodiments. BRIEF DESCRIPTION OF THE DRAWINGS
[0023] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, a brief introduction will be given below to the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.
[0024] Figure 1 A flowchart of a method for recommending vehicle rental information and reward settlement based on user portraits provided in an embodiment of the present application; Figure 2 A schematic diagram of the structure of a vehicle rental information recommendation and reward settlement system based on user portraits provided in an embodiment of the present application; Figure 3 A schematic diagram of the structure of a computing device provided in an embodiment of the present application. DETAILED DESCRIPTION
[0025] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application.
[0026] In some of the processes described in the specification and claims of this application and the above-mentioned figures, multiple operations that appear in a specific order are included, but it should be clearly understood that these operations may not be executed in the order in which they appear in this article or executed in parallel. The serial numbers of the operations, such as 101, 102, etc., are only used to distinguish between different operations, and the serial numbers themselves do not represent any execution order. In addition, these processes may include more or fewer operations, and these operations may be executed in sequence or in parallel. It should be noted that the descriptions of "first", "second", etc. in this article are used to distinguish different messages, devices, modules, etc., do not represent the order of precedence, and do not limit the "first" and "second" to be different types.
[0027] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of this application.
[0028] Figure 1 A flowchart of a method for recommending vehicle rental information and reward settlement based on user portraits is provided for an embodiment of the present application, such as Figure 1 As shown, the method includes: 101. Obtain the user's personal information and build a user profile based on historical rental behavior data; In this step, user personal information refers to data that can identify the user or describe the user's characteristics, including but not limited to the user's name, age, gender, occupation, contact information, address and other basic information.
[0029] Historical rental behavior data: refers to the data generated when users rent vehicles in the past period of time, including but not limited to the type, brand, rental duration, rental frequency, return location, mileage, etc. of the rented vehicle.
[0030] User portrait: It is a comprehensive and in-depth understanding of users formed by collecting, organizing and analyzing user information, including multiple dimensions such as users' basic attributes, interests, behaviors, consumption patterns, etc. In the field of vehicle rental, user portrait can help the platform understand user needs more accurately and provide personalized services.
[0031] In the present embodiment, it is assumed that a certain vehicle rental service platform (hereinafter referred to as the "platform") needs to build a user portrait for a newly registered user A. The following is a complete embodiment of this step: Collection of user personal information: User A filled in the following personal information when registering: name (Zhang San), age (30 years old), gender (male), occupation (IT engineer), contact information (mobile phone number), and permanent address (XX community in a high-tech zone in a certain city).
[0032] Get historical rental behavior data (for new users, this part of the data may be empty, but for users with existing rental records, it is necessary to obtain): Assuming that user A has no previous rental record on the platform, this step can be skipped. If user A is an old user, the platform will retrieve his historical rental data, such as the latest three rental records: The first time: The rental model is SUV, the brand is X, the rental period is 3 days, and the mileage is 200 kilometers.
[0033] The second time: The rental vehicle type is a sedan, the brand is Y, the rental period is 1 day, and the mileage is 50 kilometers.
[0034] The third time: rented an SUV again, the brand was Z, the rental period was 2 days, and the mileage was 150 kilometers.
[0035] Build user portraits: Based on user A’s personal information and (hypothetical) historical rental behavior data, the platform can initially construct a user profile of user A: Basic attributes: name (Zhang San), age (30 years old), gender (male), occupation (IT engineer), permanent residence (XX community in a high-tech zone of a certain city).
[0036] Preference tags (based on historical rental data, assuming user A is an old user): preference tags for SUV models (rented SUV twice), with certain preference tags for brands but not fixed (X, Z brands).
[0037] Behavioral characteristics: possible demand for short trips (shorter mileage), preference for short-term rentals (up to 3 days).
[0038] Note: Since user A is a new user and has no historical rental data, the above historical rental behavior data is only an example. In actual applications, for new users, the platform may need to use other methods (such as questionnaires, social media data, etc.) to preliminarily estimate their preference tags and behavioral characteristics, or directly build a relatively basic user portrait based on the user's personal information, and then continuously revise and improve it through the user's actual behavior data.
[0039] Through the above steps, the platform can initially build a user profile of user A, providing a basis for subsequent vehicle rental information recommendations and reward settlement.
[0040] 102. Analyze the user's preference tags according to the user portrait, and filter out vehicle information that matches the user's preference tags from a vehicle database; Optionally, the analyzing of the user's preference tags according to the user portrait in step 102 includes: classifying the user's historical rental records to identify the user's usage habits and preference tag types; using a machine learning algorithm to train a model that can predict the user's future rental behavior; and adjusting the user's preference tags according to the model's prediction results.
[0041] In this step, user portrait: a user model formed by collecting and analyzing the user's basic information, behavioral data, preference tag information, etc.
[0042] Preference tags: Identifiers set based on user behavior patterns and preference tags to distinguish different types of users.
[0043] Historical rental records: All detailed records of the user's past vehicle rentals, including rental time, location, model, etc.
[0044] Usage habits: The behavior patterns that users exhibit when renting vehicles, such as liking to rent on weekends, preferring to label a certain type of vehicle, etc.
[0045] Preference label type: user's preference labels for different types of vehicles, such as luxury cars, economy cars, electric cars, etc.
[0046] Machine learning algorithm: A type of algorithm that can automatically "learn" patterns from data and make predictions or decisions based on them.
[0047] Model training: The process of training a machine learning model using an existing dataset.
[0048] Model prediction: Use the trained model to predict future unknown data.
[0049] In the embodiment of the present application, the user's preference tags are analyzed and the user's historical rental records are classified: Suppose we have a user whose historical rental records we classify to identify his usage habits and preference tag types. For example, user B has rented a total of 10 vehicles in the past year, 7 of which were electric sedans, 2 luxury SUVs, and 1 minivan. Through these records, we can preliminarily determine that user B prefers environmentally friendly and energy-saving vehicles, and occasionally chooses luxurious and comfortable options.
[0050] Use machine learning algorithms to train a model that can predict users' future rental behavior: In order to more accurately predict the user's future rental behavior, we need to build a machine learning model. Here we can use classification algorithms in supervised learning, such as Random Forest, Support Vector Machine (SVM), etc. The training data set contains the user's rental records (time, location, car model, etc.), and the label is the car model category selected by the user.
[0051] Suppose we use a random forest model, which, after training, can predict the type of vehicle a user is most likely to rent next time based on his or her historical rental records.
[0052] Adjust the user's preference label based on the model's prediction results Once the model is trained, it can be used to predict the future rental behavior of user B. For example, the model predicts that user B is likely to choose an electric car for his next rental behavior. Based on this prediction, we can further adjust the user's preference label, such as increasing the weight of the user's preference label for electric cars.
[0053] The specific steps are as follows: Collect historical rental records of user B.
[0054] Preprocess the records, such as cleaning data, removing outliers, and standardizing.
[0055] The random forest model is used for training. The input features are rental time, location, model, etc., and the output is the model category selected by the user.
[0056] Use the trained model to predict user B’s future behavior.
[0057] Adjust user B’s preference label based on the prediction results.
[0058] For example, if the model predicts that user B has a high probability of renting an electric car in the next week, the weight of "preferred label electric car" can be increased in the user's preference labels.
[0059] Specific numerical values in the following embodiments: Sample of historical rental records: User B has rented 10 vehicles in the past, of which 7 were electric sedans (Tesla Model 3), 2 were luxury SUVs (Mercedes-Benz GLE), and 1 was a minivan (Ford Transit).
[0060] Random forest model training: implemented using Python’s scikit-learn library, which contains feature vectors (such as rental time, location, vehicle model code, etc.) and target variables (vehicle model category).
[0061] Model prediction results: According to the trained model prediction, the probability that user B will choose an electric car in the next week is 80%.
[0062] In summary, by analyzing the user's historical rental records and combining machine learning algorithms to predict the user's future behavior patterns, and then adjusting the user's preference tags, we can better provide users with personalized services.
[0063] 103. Generate a recommendation list using a recommendation algorithm for the vehicle information that matches the user's preference tags, and push the list to the user; In this step, recommendation algorithm: an information filtering technology is used to recommend relevant content to users based on their preference tags and historical behaviors.
[0064] Recommendation list: A collection of recommended items generated based on user preference tags and behavior data.
[0065] Push: The process of proactively sending information to users, usually via email, text message, app notifications, etc.
[0066] Personalized recommendation system: a system that customizes recommended content based on users' personal preferences and behavior patterns.
[0067] The following are examples: The vehicle information that matches the user's preference tags is generated into a recommendation list through the recommendation algorithm: Assume that we have completed step 102, that is, analyzing the user's preference tags based on the user portrait, and we already have a database containing a large amount of vehicle information. Now all we have to do is use the recommendation algorithm to generate a recommendation list and push this information to the user.
[0068] Analyze user preference tags: User B has been tagged with a preference tag of electric sedans, and occasionally chooses luxury SUVs. In addition, the user's rental behavior data shows that he prefers to rent vehicles on weekends and tends to pick up vehicles in the city center area.
[0069] There are many choices for recommendation algorithms. Here we choose collaborative filtering as an example, which is a recommendation method based on the similarity of user behavior and preference tags. There are two main types of collaborative filtering: user-based collaborative filtering and item-based collaborative filtering. Here, we choose item-based collaborative filtering because it is more suitable for recommending specific types of vehicles.
[0070] The core idea of the item-based collaborative filtering algorithm is to find items that are similar to the user's known preference tags and recommend these items to the user. The specific steps are as follows: Calculate similarity: Calculate the similarity between vehicles in the historical rental records of user B. The similarity can be calculated using cosine similarity or other similarity measurement methods.
[0071] Suppose we calculate the similarity between TSL Model 3 (Electric Car A) and BC GLE (LuxurySUV B) to be 0.85 and 0.40 respectively, while the similarity between TSL Model 3 and other electric sedans is higher, reaching 0.95.
[0072] Generate candidate set: Based on the similarity score, select the vehicles closest to the user's preferred label as the candidate set. In this example, the candidate set may include several different electric sedans and other luxury SUVs.
[0073] Sorting: Sort the vehicles in the candidate set according to the strength of the user's preference tag and historical rental records. Vehicles with higher preference tag strength are ranked first.
[0074] Generate recommendation list: Select the top N items from the sorted candidate set as the final recommendation list.
[0075] Once the recommendation list is generated, we can push this information to users in a variety of ways. For example, we can inform users through the notification function of the mobile application or send emails.
[0076] The following are specific embodiments: Suppose we have the following vehicles in our vehicle database: TSL Model3(Electric Car A) WLE S6 (Electric Car B) BM iX3 (Electric Car C) BC GLE (Luxury SUV B) AD Q7 (Luxury SUV C) According to user B's preference tags, we find that he prefers electric sedans and sometimes considers luxury SUVs. Therefore, we can generate the following recommendation list: TSL Model 3(Electric Car A) WLES6 (Electric Car B) BM iX3 (Electric Car C) BC GLE (Luxury SUV B) We can also further optimize the recommendation list based on the user's rental time and location preference tags, such as recommending vehicles available on weekends and recommending pick-up points near the city center.
[0077] Finally, we push these recommendations to user B through the notification function of the application. The notification contains the information of the recommended vehicle and the booking link, so that the user can proceed directly to the next step.
[0078] This completes the entire process of generating a recommendation list based on user preference tags and pushing it to users.
[0079] 104. Record the user's feedback behavior on the recommendation list, update the user portrait, and adjust the recommendation strategy; In this step, feedback behavior: the user's response to the recommendation list, such as click, view details, rent, etc.
[0080] User portrait: A user model formed by collecting and analyzing the user's basic information, behavioral data, preference tag information, etc.
[0081] Recommendation strategy: a method or algorithm for adjusting recommended content based on user profiles and feedback behavior.
[0082] Update user profile: Modify the user model based on new feedback data to make it closer to the user's actual preference tags.
[0083] Adjust recommendation strategy: Based on the updated user profile, change the parameters or logic of the recommendation algorithm to improve the accuracy and satisfaction of the recommendation.
[0084] The following is an example of recording user feedback on a recommendation list: Assume that we have generated a recommendation list based on user B's preference tags and pushed information to him. Next, we need to record the user's feedback behavior on these recommendations.
[0085] Record feedback behavior: User B viewed the vehicles in the recommended list and finally selected TSL Model 3 (ElectricCar A) for rental. This behavior shows that user B not only viewed the information in the recommended list, but also responded positively to the recommended content.
[0086] The specific records are as follows: Viewing behavior: User B viewed all the vehicles in the recommended list.
[0087] Leasing behavior: User B finally chooses TSL Model 3 (Electric Car A) for leasing.
[0088] Feedback time: User B rented the property at 10 a.m. the next day after receiving the recommendation list.
[0089] Update user profile: Based on user feedback, we need to update the user portrait. The process of updating the user portrait includes but is not limited to the following points: Strengthening preference labels: User B selected TSL Model 3, which further strengthened his preference label for electric sedans.
[0090] Add rental record: Add user B's rental behavior to his historical rental record.
[0091] Adjust behavior patterns: Record the time when user B responds after receiving the recommendation to understand his behavior patterns.
[0092] Adjust the recommended strategy: Based on the updated user profile, we need to adjust the recommendation strategy. Adjustment strategies may include: Increase the weight of electric cars: Since user B selected the electric car again, the weight of the electric car in his preference label can be appropriately increased.
[0093] Optimize push time: Considering that user B rented at around 10 am, future push times can be adjusted to the morning hours to increase the response rate.
[0094] Personalized recommendations: Based on user B’s feedback, other electric sedans similar to the TSL Model 3 can be recommended in a more personalized way. At the same time, some luxury SUVs can also be recommended as supplementary options.
[0095] The following are specific embodiments: Assume that after receiving the recommendation list, user B chooses TSL Model 3 for rental and gives a good review during the rental process. We can update the user profile and adjust the recommendation strategy through the following steps: Record feedback behavior: After receiving the recommendation list, User B chose TSL Model 3 for rental and gave a positive review during the rental process.
[0096] User B viewed all the vehicles in the recommended list and rented them at 10 am the next day.
[0097] Update user profile: Add a new rental record to the user portrait: TSL Model 3 (Electric Car A).
[0098] Enhance user B's preference label for electric cars, for example, increase the preference label weight from 0.8 to 0.9.
[0099] Update the user's behavior pattern and record his rental behavior around 10 am.
[0100] Adjust the recommended strategy: Next time when pushing a recommendation list, electric cars will be given priority, especially TSL Model 3 and its similar models.
[0101] The push time is adjusted to the morning period to better match users' rental habits.
[0102] Based on user B’s positive feedback, continue to recommend other models of the TSL brand, such as TSL Model S or Model Y.
[0103] The following are specific numerical examples: Assume that user B chooses TSL Model 3 for rental after receiving the recommendation list and gives a good review during the rental process. After updating the user profile, the user's preference label may become: Electric car preference label weight: 0.9 Luxury SUV preference label weight: 0.4 Rental time preference tag: around 10 am After adjusting the recommendation strategy, the next recommendation list may include: TSL Model 3(Electric Car A) TSL Model S (Electric Car D) TSL Model Y(Electric Car E) BC GLE (Luxury SUV B) In this way, we can continuously optimize the recommendation strategy so that the recommendations are more in line with the users' actual needs and preference tags.
[0104] 105. Calculate the user's reward points based on the user's rental behavior record using preset reward rules, and give the user a corresponding reward based on the reward points.
[0105] Optionally, the step 105 calculates the user's reward points based on the user's rental behavior record using preset reward rules, including: setting different point values corresponding to different rental behaviors, the rental behaviors including rental duration and rental frequency; establishing a user point pool, and dynamically adjusting the size of the point pool according to the user's activity level; calculating the corresponding reward points based on the user's rental behavior record, and adding the reward points to the user's point pool to determine the user's final reward points.
[0106] Optionally, the reward points of the user are calculated using a preset reward rule in step 105, and the specific calculation formula is as follows:
[0107] in, Indicates The increment of reward points earned by each user; Indicates the total number of rental behavior records of the user; Indicates the total number of vehicle types; Indicates The duration of the lease; Represents the average of all rental durations; Indicates The rental frequency of each user; represents the average rental frequency of all users; Indicates driving distance of each user; represents the average driving distance of all users; Indicates Number of rentals of this type of vehicle; Indicates the average number of rentals for each type of vehicle; They represent the weight coefficients of rental duration, rental frequency, driving distance, and number of rentals respectively.
[0108] Optionally, the reward in step 105 includes the total amount of free rental time obtained by the user; Giving the user a corresponding reward according to the reward points includes: The user is given a corresponding reward according to the reward points through the following formula:
[0109] in, Indicates the total amount of free rental time obtained by the user, Indicates a round-down operation; Indicates the total number of reward points in the user's point pool; is the time decay factor of the bonus points; is the user’s credit score; It is a quantitative indicator of the user’s social network influence; is the preference label coefficient of the user's rental behavior in different time periods; is the frequency of user interaction on social media; It is the time series characteristic of leasing behavior; Indicates the length of rental time that a unit of points can be redeemed for; is the user's historical behavior score.
[0110] Optionally, the step of giving the user a corresponding reward according to the reward points in step 105 further includes: calculating an effective value of the reward points, and giving the user a corresponding reward based on the effective value; The effective value of the reward points is calculated by the following formula:
[0111] in, Indicates the valid value of reward points; Indicates the final reward points; It is a coefficient adjusted according to user activity, ranging from 0 to 1; It is a coefficient adjusted according to the user's social influence, ranging from 0 to 1; is the user's preference label coefficient for different types of vehicles; It is the user’s loyalty to any one brand; is the trend of user activity over time; Also includes: Dynamic adjustment using Bayesian update rule and to update the effective value of the reward points; Among them, the following formula is used to dynamically adjust and Values: in, and They are The user's preference label coefficients for different types of vehicles and the user's loyalty to any brand after the update; and It is the actual obtained user preference label coefficients for different types of vehicles and the user's loyalty to any brand; and is the smoothing parameter; Indicates The user's preference label coefficients for different types of vehicles after the update; Indicates The user's loyalty to any brand after the update.
[0112] In this step, rental behavior records: user activity records in the rental service, including rental duration, rental frequency, driving distance, etc.
[0113] Points Pool: A virtual account used to store a user's reward points.
[0114] Reward points: A form of reward given based on the user's rental behavior, which can be exchanged for different rewards.
[0115] Reward rules: A set of preset algorithms or formulas used to calculate a user's reward points.
[0116] Time decay factor ( ): The value of points may change over time, and this factor reflects this change; Credit score ( ): An indicator to measure user reputation; Social network influence ( ): The influence of users on social networks; Preference label coefficient ( ): The preference label coefficient of the user's rental behavior in different time periods; Interaction frequency ( ): Frequency of user interaction on social media; Time series features ( ): Time series characteristics of leasing behavior; Unit points redeem rental time length ( ): The length of rental time that a unit point can be redeemed for; Historical Behavior Score ( ): Rating of user's historical behavior; Activity adjustment factor ( ): Coefficient adjusted according to user activity; Social influence adjustment coefficient ( ): Coefficient adjusted according to the user's social influence; Preference label coefficient ( ): User preference label coefficients for different types of vehicles; Brand Loyalty ( ): User loyalty to a brand; Activity Trends ): The trend of user activity over time; Smoothing parameter ( ): Smoothing parameter used in the Bayesian update rule.
[0117] The following are examples: Calculate reward points based on the user's rental behavior record: Suppose we have a user and we have recorded his rental behavior. Now we need to calculate the reward points based on these records and give corresponding rewards.
[0118] Set different rental behaviors to correspond to different points values: Assume that we set the following integral values: Rental time: 1 point per hour; Rental frequency: 5 points for each rental; Driving distance: 0.1 point per kilometer; Create a user points pool: User A is an active user. We set up a points pool for him and dynamically adjust the size of the points pool based on his activity level.
[0119] Calculate reward points: User A's rental behavior record is as follows: Rental duration ( ): The last few rentals were 2 hours, 3 hours, 2 hours, and 4 hours respectively; Lease frequency ( ): Lease 3 times per month; Driving distance ( ): The driving distances of the most recent rentals were 20 km, 30 km, 25 km, and 40 km respectively; Number of rentals of various types of vehicles ( ): Two types of cars were rented, one was an electric sedan (rented 3 times in total) and the other was a luxury SUV (rented 1 time in total).
[0120] We set the following averages: Average rental duration ( ): 2.5 hours; Average rental frequency ( ): 2 times / month; Average driving distance ( ): 25 km; Average number of leases ( ): 2 times; Weight coefficient ( ) are 0.4, 0.3, 0.1 and 0.2 respectively.
[0121] According to the formula:
[0122] Substitute specific values:
[0123] Simplified:
[0124]
[0125] Therefore, the reward point increment for user A is 30.4 points.
[0126] Give corresponding rewards: Give users corresponding rewards based on their reward points. Assume that user A has 2000 points in his points pool, and the time decay factor is The credit score is 0.9. 0.9, social network influence is 0.8, the preference label coefficient is 0.7, the interaction frequency is 0.6, time series characteristics 0.5, historical behavior score 0.4, unit points can be exchanged for rental time For 1 minute.
[0127]
[0128] Substitute specific values:
[0129] Simplified:
[0130] Therefore, user A can obtain about 3 minutes of free rental time.
[0131] Calculate the effective value of reward points: Assume that user A's activity adjustment coefficient The social influence adjustment coefficient is 0.8. is 0.7, the preference label coefficient 0.9, brand loyalty is 0.8, activity trend is 0.6.
[0132] According to the formula:
[0133] Substitute specific values:
[0134] Simplified:
[0135] Therefore, the effective value of user A's reward points is 403.2 points.
[0136] Dynamically adjust preference label coefficients and brand loyalty: Assuming smoothing parameters is 0.5, is 0.6, the actual obtained preference label coefficient 0.85, brand loyalty is 0.75, the preference label coefficient after the last update 0.9, brand loyalty is 0.8.
[0137] According to the formula: Substitute specific values:
[0138] Therefore, the updated preference label coefficient of user A is 0.875 and the brand loyalty is 0.78.
[0139] Through the above steps, we can calculate the reward points based on the user's rental behavior records and give corresponding rewards. At the same time, we can continuously adjust the user's preference tags and brand loyalty based on feedback to make the recommendation more accurate.
[0140] Optionally, the step 105 of calculating the user's reward points according to the user's rental behavior record and using a preset reward rule includes: updating a user preference tag, and adjusting the user's reward points according to the user preference tag; The updating of the user preference tag and adjusting the user's reward points according to the user preference tag includes: using the user's rental behavior record based on the Bayesian formula Dynamically update the user's preferences for multiple vehicle types Preferences tab , The calculation formula of the Bayesian formula includes:
[0141] in, Indicates the updated preference label, which is the observed rental behavior record In the case of user preference label vehicle type probability; Indicates the user's preference label vehicle type Under the assumption that probability; is the prior probability, indicating that without considering any rental behavior records Previously, the user preferred the label vehicle type probability; Indicates all assumptions Evidence The weighted sum of the probabilities of , namely the marginal probability, is used as a normalization factor to ensure The value of is between 0 and 1; Take advantage of the updated preference tab adjusting the user's reward points; Among them, the adjustment formula is defined as:
[0142] in, represents the increment of the adjusted reward points; Indicates the increment of reward points before adjustment; It is the adjustment coefficient for each vehicle type, which is used to control the impact of changes in user preference labels on reward points; The total number of different types of vehicles representing user preference labels; Indicates the updated preference label; represents the prior probability; is the importance weight of each vehicle type.
[0143] In this step, we use the Bayesian probability model to dynamically update the user's preference tags and adjust the reward points accordingly. The following is an explanation of the concepts involved: Posterior probability : This is new evidence After that, the updated estimate of the user's preference label is obtained. It indicates that after observing the user's specific rental behavior, they prefer a certain car model. The degree of preference for labels.
[0144] Likelihood : This is in the known user preference tag In the case of In short, it is a measure of the probability of new evidence. In the assumption The possibility of occurrence under establishment.
[0145] Prior probability : This refers to the initial estimate of the user's preference labels before any new information is obtained. It represents the system's understanding of the user's preference labels before considering the user's latest behavior.
[0146] Normalization constant : The denominator is the sum of the posterior probabilities of all possible hypotheses, which ensures that the posterior probability Between 0 and 1.
[0147] Adjustment factor and importance weight :These two parameters are used to adjust the points reward mechanism. Adjustment coefficient Controls the impact of changes in user preference tags on point rewards, and the importance weight It reflects the relative importance of different types of vehicles in the points adjustment.
[0148] In the present embodiment, we assume that a car rental platform wants to dynamically adjust the points reward based on the user's rental behavior. For example, if user A often rents an SUV model (assuming ), then the platform will update the preference tag To increase the points the user earns when leasing an SUV.
[0149] Suppose there are two vehicle types: SUV and sedan , and we know that user A prefers SUV in the past. Assume that the prior estimate of the platform is and , and new evidence It shows that user A recently rented an SUV again. Through Bayesian updating, we get the updated preference label and .
[0150] Hypothesis Likelihood and , then the posterior probability is calculated as follows:
[0151] Next, we adjust the points reward, assuming that the points increment before adjustment is , adjustment coefficient , importance weight , then the adjusted integral increment is:
[0152] Therefore, based on user A's latest rental behavior, the adjusted points increment is approximately 9.752 points. This process can be dynamically adjusted based on different user behaviors to encourage users to use their favorite models more often. Figure 2 A structural diagram of a vehicle rental information recommendation and reward settlement system based on user portrait is provided for the embodiment of the present application, such as Figure 2 As shown, the system includes: The acquisition construction module 21 is used to obtain the user's personal information and construct a user portrait in combination with historical rental behavior data; An analysis and screening module 22, configured to analyze the user's preference tags according to the user portrait, and screen out vehicle information that matches the user's preference tags from a vehicle database; A push module 23 is used to generate a recommendation list of the vehicle information that meets the user's preference tags through a recommendation algorithm and push it to the user; An updating and adjusting module 24 is used to record the user's feedback behavior on the recommendation list, update the user portrait, and adjust the recommendation strategy; The calculation and granting module 25 is used to calculate the reward points of the user according to the rental behavior record of the user and the preset reward rules, and grant the user a corresponding reward according to the reward points; The calculation and granting module is specifically used to set different point values corresponding to different rental behaviors, and the rental behaviors include rental duration and rental frequency; establish a user point pool, and dynamically adjust the size of the point pool according to the user's activity level; calculate the corresponding reward points based on the user's rental behavior record, and add the reward points to the user's point pool to determine the user's final reward points.
[0153] Figure 2 The vehicle rental information recommendation and reward settlement system based on user portrait can be executed Figure 1 The implementation principle and technical effect of the vehicle rental information recommendation and reward settlement method based on user portrait described in the illustrated embodiment will not be described in detail. The specific manner in which each module and unit performs operations in the vehicle rental information recommendation and reward settlement system based on user portrait in the above embodiment has been described in detail in the embodiment of the method, and will not be elaborated here.
[0154] In one possible design, Figure 2 The vehicle rental information recommendation and reward settlement system based on user portrait of the illustrated embodiment can be implemented as a computing device, such as Figure 3 As shown, the computing device may include a storage component 31 and a processing component 32; The storage component 31 stores one or more computer instructions, wherein the one or more computer instructions are called and executed by the processing component 32 .
[0155] The processing component 32 is used to: obtain the user's personal information and build a user portrait in combination with historical rental behavior data; analyze the user's preference tags based on the user portrait, and filter out vehicle information that meets the user's preference tags from the vehicle database; generate a recommendation list for the vehicle information that meets the user's preference tags through a recommendation algorithm, and push it to the user; record the user's feedback on the recommendation list, update the user portrait, and adjust the recommendation strategy; calculate the user's reward points based on the user's rental behavior record using preset reward rules, and give the user corresponding rewards based on the reward points.
[0156] The processing component 32 may include one or more processors to execute computer instructions to complete all or part of the steps in the above method. Of course, the processing component may also be implemented by one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors or other electronic components to perform the above method.
[0157] The storage component 31 is configured to store various types of data to support operations at the terminal. The storage component can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk or optical disk.
[0158] Of course, the computing device may also include other components, such as input / output interfaces, display components, communication components, etc.
[0159] The input / output interface provides an interface between the processing component and the peripheral interface module, which may be an output device, an input device, etc.
[0160] The communication component is configured to facilitate, among other things, wired or wireless communications between the computing device and other devices.
[0161] Among them, the computing device can be a physical device or an elastic computing host provided by a cloud computing platform, etc. In this case, the computing device can refer to a cloud server, and the above-mentioned processing components, storage components, etc. can be basic server resources rented or purchased from the cloud computing platform.
[0162] The present application also provides a computer storage medium storing a computer program, wherein the computer program can achieve the above-mentioned Figure 1 The illustrated embodiment shows a method for recommending vehicle rental information and reward settlement based on user portraits.
[0163] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.
[0164] The system embodiment described above is merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment. Those of ordinary skill in the art may understand and implement it without creative work.
[0165] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.
[0166] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit it. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application.
Claims
1. A vehicle rental information recommendation and reward settlement method based on user portrait, characterized in that: The following steps are involved: Obtain the user's personal information and build a user profile based on historical rental behavior data; Analyze the user's preference tags according to the user portrait, and filter out vehicle information that matches the user's preference tags from a vehicle database; Generate a recommendation list of the vehicle information that meets the user's preference tags through a recommendation algorithm and push it to the user; Record the user's feedback on the recommendation list, update the user portrait, and adjust the recommendation strategy; According to the rental behavior record of the user, the reward points of the user are calculated using the preset reward rules, and the user is given a corresponding reward according to the reward points; The step of calculating the user's reward points according to the user's rental behavior record and using a preset reward rule includes: Different points values are set corresponding to different rental behaviors, and the rental behaviors include rental duration and rental frequency; Establish a user points pool and dynamically adjust the size of the points pool based on the activity level of the user; According to the rental behavior record of the user, the corresponding reward points are calculated, and the reward points are added to the point pool of the user to determine the final reward points of the user.
2. The method according to claim 1, characterized in that Analyzing the user's preference tags according to the user portrait includes: Classify the historical rental records of the user and identify the user's usage habits and preference tag types; Use machine learning algorithms to train models that can predict users’ future rental behavior; According to the result predicted by the model, the preference label of the user is adjusted.
3. The method according to claim 1, characterized in that The reward points of the user are calculated using the preset reward rules, and the specific calculation formula is as follows: ; in, Indicates The increment of reward points earned by each user; Indicates the total number of rental behavior records of the user; Indicates the total number of vehicle types; Indicates The duration of the lease; Represents the average of all rental durations; Indicates The rental frequency of each user; represents the average rental frequency of all users; Indicates driving distance of each user; represents the average driving distance of all users; Indicates Number of rentals of this type of vehicle; Indicates the average number of rentals for each type of vehicle; They represent the weight coefficients of rental duration, rental frequency, driving distance, and number of rentals respectively.
4. The method according to claim 3, characterized in that The reward includes the total amount of free rental time obtained by the user; Giving the user a corresponding reward according to the reward points includes: The user is given a corresponding reward according to the reward points through the following formula: ; in, Indicates the total amount of free rental time obtained by the user, Indicates a round-down operation; Indicates the total number of reward points in the user's point pool; is the time decay factor of the bonus points; is the user’s credit score; It is a quantitative indicator of the user’s social network influence; is the preference label coefficient of the user's rental behavior in different time periods; is the frequency of user interaction on social media; It is the time series characteristic of leasing behavior; Indicates the length of rental time that a unit of points can be redeemed for; is the user's historical behavior score.
5. The method according to claim 4, characterized in that Giving the user a corresponding reward according to the reward points also includes: Calculating the effective value of the reward points, and giving the user a corresponding reward based on the effective value; The effective value of the reward points is calculated by the following formula: ; in, Indicates the valid value of reward points; Indicates the final reward points; It is a coefficient adjusted according to user activity, ranging from 0 to 1; It is a coefficient adjusted according to the user's social influence, ranging from 0 to 1; is the user's preference label coefficient for different types of vehicles; It is the user’s loyalty to any one brand; is the trend of user activity over time; Also includes: Dynamic adjustment using Bayesian update rule and to update the effective value of the reward points; Among them, the following formula is used to dynamically adjust and Values: ;in, and They are The user's preference label coefficients for different types of vehicles and the user's loyalty to any brand after the update; and It is the actual obtained user preference label coefficients for different types of vehicles and the user's loyalty to any brand; and is the smoothing parameter; Indicates The user's preference label coefficients for different types of vehicles after the update; Indicates The user's loyalty to any brand after the update.
6. The method according to claim 1, characterized in that The step of calculating the user's reward points according to the user's rental behavior record and using a preset reward rule includes: Updating a user preference tag, and adjusting the user's reward points according to the user preference tag; The updating of the user preference tag and adjusting the reward points of the user according to the user preference tag includes: Based on the Bayesian formula, using the user's rental behavior record Dynamically update the user's preferences for multiple vehicle types Preferences tab , The calculation formula of the Bayesian formula includes: ; in, Indicates the updated preference label, which is the observed rental behavior record In the case of user preference label vehicle type The probability of Indicates the user's preference label vehicle type Under the assumption that The probability of is the prior probability, indicating that without considering any rental behavior records Previously, the user preferred the label vehicle type The probability of Indicates all assumptions Evidence The weighted sum of the probabilities of , namely the marginal probability, is used as a normalization factor to ensure The value of is between 0 and 1; Take advantage of the updated preference tab adjusting the user's reward points; Among them, the adjustment formula is defined as: ; in, represents the increment of the adjusted reward points; Indicates the increment of reward points before adjustment; It is the adjustment coefficient for each vehicle type, which is used to control the impact of changes in user preference labels on reward points; The total number of different types of vehicles representing user preference labels; Indicates the updated preference label; represents the prior probability; is the importance weight of each vehicle type.
7. A vehicle rental information recommendation and reward settlement system based on user portrait, characterized in that: include: Acquisition building module, used to obtain the user's personal information and build a user portrait based on historical rental behavior data; An analysis and screening module, used to analyze the user's preference tags according to the user portrait, and screen out vehicle information that meets the user's preference tags from the vehicle database; A push module is used to generate a recommendation list of the vehicle information that meets the user's preference tags through a recommendation algorithm and push it to the user; An updating and adjusting module, used to record the user's feedback behavior on the recommendation list, update the user portrait, and adjust the recommendation strategy; A calculation and granting module, used to calculate the user's reward points according to the user's rental behavior record and using a preset reward rule, and grant the user a corresponding reward according to the reward points; The calculation and granting module is specifically used to set different points values corresponding to different rental behaviors, and the rental behaviors include rental duration and rental frequency; A user points pool is established, and the size of the points pool is dynamically adjusted according to the user's activity level; corresponding reward points are calculated according to the user's rental behavior record, and the reward points are added to the user's points pool to determine the user's final reward points.
8. A computing device, characterized in that It comprises a processing component and a storage component; the storage component stores one or more computer instructions; the one or more computer instructions are used to be called and executed by the processing component to implement a vehicle rental information recommendation and reward settlement method based on user portrait as described in any one of claims 1 to 6.
9. A computer storage medium, characterized in that A computer program is stored, and when the computer program is executed by a computer, a vehicle rental information recommendation and reward settlement method based on user portrait as described in any one of claims 1 to 6 is implemented.
Citation Information
Patent Citations
Recommendation method, recommendation device and terminal equipment
CN112102031A
Machine learning-based integral dynamic adjustment method and system
CN118586962A
Vehicle rental management device, vehicle rental management program, and vehicle rental management method
WO2023218737A1
Cited By
Product renting method based on AI big data dialogue generation
CN120429494A
Intelligent management method and system for rented vehicles
CN120509950A
Unmanned aerial vehicle leasing interaction method and device based on multi-modal data and medium
CN120563215A
