Mobile phone recommendation method for deposit-free online leasing

Abstract

The invention provides a mobile phone recommendation method for deposit-free online leasing, and relates to the technical field of mobile phone online leasing, and the mobile phone recommendation method for deposit-free online leasing comprises the steps: comparing the personal basic information of a new leasing user with the personal basic information of historical leasing users; obtaining similar customer groups of the new leasing user in historical leasing users of each mobile phone in the mobile phone model list; comparing the personal preference information of the new leasing user with the personal preference information of historical leasing users, and screening out a nearest neighbor set of the new leasing user in a similar customer group; and calculating the score of each historical leasing user for the corresponding mobile phone type in the nearest neighbor set, selecting a top-N commodity set, and recommending the top-N commodity set. According to the method, a relatively accurate prediction result can be obtained without providing online behavior data such as website browsing and online shopping by a new lease user, a relatively suitable mobile phone model is recommended to the new lease user, and the mobile phone online lease requirement of the new lease user is well met.

Description

Technical Field

[0001] The present application relates to the technical field of online mobile phone rental, and in particular to a mobile phone recommendation method for online rental without deposit. Background Art

[0002] Due to the accelerated replacement of smart devices and the increasing demand of consumers for flexible use and low-cost solutions, as well as the continuous improvement of the credit system and the application of big data technology, the market for deposit-free mobile phone rental services has continued to expand. Traditional deposit-free mobile phone rental services require users to browse different mobile phone models on the rental platform and choose the right mobile phone according to their own needs and budget. At the same time, the mobile phone recommendation algorithm of the rental platform recommends suitable mobile phone models to users based on factors such as the user's personal preferences, historical behavior, and the characteristics of the mobile phone.

[0003] For example, the invention patent with publication number CN117726416A discloses a method, system and rental platform for online mobile phone rental. The method for online mobile phone rental in the invention patent discloses a mobile phone recommendation algorithm, including: collecting user online behavior data such as user browsing history, shopping habits and application preferences; analyzing and processing the collected user behavior data to predict the user's mobile phone usage needs; and recommending mobile phone models that suit the user's needs based on the predicted mobile phone usage needs.

[0004] However, when the above mobile phone recommendation algorithm recommends mobile phone models that suit users' needs, it needs to collect online behavior data such as users' browsing history, shopping habits and application preferences. For most new users, due to the lack of information such as website browsing and online shopping, there is often a lack of sufficient online behavior data, which leads to inaccurate recommendation results. There is a certain gap between the recommended mobile phone models and the expectations of new users, and they cannot better meet the online mobile phone rental needs of new users. Summary of the invention

[0005] The purpose of this application is to provide a method for recommending mobile phones for online rental without deposit, which is used to solve the problem that the mobile phone recommendation algorithm in the related art is prone to lack of sufficient online behavior data for new users, resulting in inaccurate recommendation results, and there is a certain gap between the recommended mobile phone models and the expectations of new users, and the online mobile phone rental needs of new users cannot be met well.

[0006] The present application provides a method for recommending mobile phones for online rental without deposit, which adopts the following technical solution:

[0007] A method for recommending a mobile phone for online rental without deposit, comprising:

[0008] Collect basic personal information and personal preference information of historical rental users and new rental users respectively;

[0009] Obtain the mobile phone model rental demand information of new rental users;

[0010] Generate a list of mobile phone models that meet the mobile phone model rental needs of new rental users;

[0011] Compare the basic personal information of the new rental user with that of the historical rental users, and obtain similar customer groups of the new rental user among the historical rental users of each mobile phone in the mobile phone model list through cluster analysis;

[0012] Compare the personal preference information of the new rental user with the personal preference information of the historical rental users, and filter out the nearest neighbor set of the new rental user in the similar customer group;

[0013] Calculate the ratings of the corresponding mobile phone models by each historical rental user in the nearest neighbor set, arrange the obtained scores in descending order, select the top-N product set and recommend them.

[0014] Optionally, the step of screening out a nearest neighbor set of the new rental user in the similar customer group includes:

[0015] The similarity between each historical rental user and the new rental user in the similar customer group is calculated, and the similarities are arranged from high to low. The top-ranked historical rental users are determined based on a predetermined similarity threshold to form a nearest neighbor set of the new rental user in the similar customer group.

[0016] Optionally, the calculating the similarity between each historical rental user and the new rental user in the similar customer group includes: calculating the similarity between each historical rental user and the new rental user in the similar customer group according to the following formula:

[0017]

[0018] Among them, sim(u i ,u j ) is the historical rental user u in the similar customer group j With new rental user u i The similarity of I ij For new rental users i and historical rental user u j The set of identical items in the personal preference information of ik and r jk They are new rental users u i and historical rental user u j The score of each identical item in the personal preference information, and They are new rental users u iand historical rental user u j The score of each category of the same item in the personal preference information.

[0019] Optionally, the calculating of the score of the corresponding mobile phone model by each historical rental user in the nearest neighbor set includes: calculating the score of the corresponding mobile phone model by each historical rental user in the nearest neighbor set based on the user experience of the historical rental user in the mobile phone rental process, combined with the item attributes of the mobile phone model rented by the historical rental user, and the preference degree of the historical rental user for the rented mobile phone model.

[0020] Optionally, calculating the score of each historical rental user in the nearest neighbor set for the corresponding mobile phone model includes: calculating the score of each historical rental user in the nearest neighbor set for the corresponding mobile phone model according to the following formula:

[0021]

[0022] Among them, r ui is the rating of mobile phone model i by historical rental user u, pop i is the popularity of mobile phone models i, e ui is the user experience coefficient of historical rental user u for mobile phone model i, is the transpose of the corresponding vector of the historical rental user u in the feature matrix P, representing the user feature, Q i is the vector corresponding to mobile phone model i in the feature matrix Q, representing the item features, The preference of historical rental user u for mobile phone model i is obtained by dot product operation of user characteristics and item characteristics.

[0023] Optionally, the user experience coefficient e ui The calculation formula is as follows:

[0024]

[0025] Among them, e ui is the user experience coefficient of historical rental user u for mobile phone model i, X uik is the number of historical rental user u’s feedback behaviors k on mobile phone model i, W uk is the weight value of the historical rental user u on the feedback behavior k.

[0026] Optionally, the popularity of the mobile phone model i is i The calculation formula is as follows:

[0027] pop i =σ(c i , c avg ),

[0028]

[0029] Among them, pop i is the popularity of mobile phone model i, c i is the number of times mobile phone model i is rented in the online rental system, c avg is the average number of times each mobile phone model is rented in the online rental system, c all is the number of times all mobile phone models are rented in the online rental system, c ai It is the number of types of mobile phone models rented in the online rental system.

[0030] Optionally, the basic personal information includes the renter's age, gender, education background, housing situation, work unit, occupation, job level, means of transportation, income level and family status.

[0031] Optionally, the comparing the basic personal information of the new rental user with the basic personal information of the historical rental users, and obtaining similar customer groups of the new rental user among the historical rental users of each mobile phone in the mobile phone model list through cluster analysis, includes:

[0032] The basic personal information of new rental users and historical rental users is normalized and converted into numerical information of unified standards. The numerical information of new rental users and historical rental users is compared and arranged from high to low according to similarity. The top-ranked historical rental users are determined based on a predetermined similarity threshold, forming a similar customer group of new rental users among the historical rental users of each mobile phone in the mobile phone model list.

[0033] Optionally, the basic personal information of new and historical rental users is normalized to convert the basic personal information into numerical information of a unified standard, including:

[0034] For non-numeric data in personal basic information, different numeric values ​​in the range of [0, 1] are assigned to each non-numeric data in personal basic information according to the set weight; for numeric data in personal basic information, the normal distribution S function is used to convert the numeric data in the personal basic information into numeric values ​​distributed in (0, 1).

[0035] In summary, this application at least includes the following beneficial technical effects:

[0036] 1. The mobile phone recommendation method for deposit-free online rental of the present application collects personal basic information and personal preference information of historical rental users and new rental users, compares the personal basic information of new rental users with the personal basic information of historical rental users, obtains similar customer groups among historical rental users of each mobile phone in the mobile phone model list of the new rental user, and then compares the personal preference information of the new rental user with the personal preference information of the historical rental user, screens out the nearest neighbor set of the new rental user in the similar customer group, and finally calculates the score of each historical rental user in the nearest neighbor set for the corresponding mobile phone model, arranges the obtained scores in order from large to small, selects the top-N product set and recommends it. Therefore, the mobile phone recommendation method for deposit-free online rental of the present application only needs to obtain the personal basic information and personal preference information of historical rental users and new rental users, and does not require new rental users to provide online behavior data such as website browsing and online shopping, and can obtain more accurate prediction results, recommend more suitable mobile phone models for new rental users, better meet the mobile phone online rental needs of new rental users, and improve the mobile phone online rental experience.

[0037] 2. When searching for the user set with the highest similarity to the new rental user from historical rental users, the mobile phone recommendation method for deposit-free online rental of the present application first compares the personal basic information of the new rental user with the personal basic information of the historical rental users, obtains the similar customer group among the historical rental users of each mobile phone in the mobile phone model list of the new rental user, and then compares the personal preference information of the new rental user with the personal preference information of the historical rental users, and filters out the nearest neighbor set of the new rental user in the similar customer group, thereby effectively improving the similarity between the filtered historical rental user set and the new rental user, and then improving the matching degree of the mobile phone model recommendation.

[0038] 3. Based on the user experience of historical rental users in the process of mobile phone rental, combined with the item attributes of the mobile phone models rented by historical rental users, and the preference of historical rental users for the rented mobile phone models, the score of each historical rental user for the corresponding mobile phone model in the nearest neighbor set is calculated, thereby improving the accuracy of the score, making the recommended mobile phone models closer to the expectations of new users, and further improving the online mobile phone rental experience of new rental users. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1 This is a flow chart of a method for recommending a mobile phone for online rental without deposit in an embodiment of the present application. DETAILED DESCRIPTION

[0040] The following is combined with Figure 1 , further details of this application are given.

[0041] The embodiment of the present application discloses a method for recommending mobile phones for online rental without deposit.

[0042] Reference Figure 1 , a mobile phone recommendation method for online rental without deposit, comprising the following steps:

[0043] Step S1: Collect basic personal information and personal preference information of historical rental users and new rental users respectively.

[0044] The basic personal information includes but is not limited to the following: the renter's age, gender, education, housing conditions, work unit, occupation, job level, means of transportation, income level and family status.

[0045] The personal preference information includes but is not limited to the following:

[0046] Sports: basketball, badminton, table tennis, football, tennis, swimming, fitness, skateboarding, skating, running and skipping rope, etc.;

[0047] Culture: reading, writing, painting, music, film and drama, etc.;

[0048] Handicrafts: weaving, embroidery, woodworking, origami, paper cutting, pottery and baking, etc.;

[0049] Technology: computers, mobile phones, photography and games;

[0050] Tourism: domestic tourism, foreign tourism and outdoor adventure, etc.;

[0051] Social: parties, clubs and societies, etc.

[0052] Intelligence: puzzles, Rubik's cubes, chess, etc.;

[0053] Collections: cultural relics, jewelry, coins, stamps, ceramics and jade, etc.;

[0054] Musical instruments: guitar, piano, saxophone, Hulusi, tuba and trumpet, etc.

[0055] Step S2. Obtaining the mobile phone model rental demand information of the new rental user, wherein the mobile phone model rental demand information includes at least one of the mobile phone brand, mobile phone price, mobile phone model, rental price, processor information, memory parameters, storage space parameters, screen parameters, camera parameters and battery parameters;

[0056] Step S3. Generate a list of mobile phone models that meet the new rental user's mobile phone model rental demand information. For example, in step S2, the new rental user's mobile phone model rental demand information is a 6.7-inch screen and a 50-megapixel camera, then generate a list of all mobile phone models that meet the 6.7-inch screen and 50-megapixel camera requirements.

[0057] Step S4: compare the basic personal information of the new rental user with the basic personal information of the historical rental users, and obtain similar customer groups of the new rental user among the historical rental users of each mobile phone in the mobile phone model list through cluster analysis.

[0058] In an optional embodiment, the step of comparing the basic personal information of the new rental user with the basic personal information of the historical rental users, and obtaining a similar customer group of the new rental user among the historical rental users of each mobile phone in the mobile phone model list through cluster analysis, includes the following steps:

[0059] The basic personal information of new rental users and historical rental users is normalized and converted into numerical information of unified standards. The numerical information of new rental users and historical rental users is compared and arranged from high to low according to similarity. The top-ranked historical rental users are determined based on a predetermined similarity threshold, forming a similar customer group of new rental users among the historical rental users of each mobile phone in the mobile phone model list.

[0060] In an optional embodiment, normalizing the basic personal information of new rental users and historical rental users to convert the basic personal information into numerical information of a unified standard includes the following steps:

[0061] For non-numeric data in personal basic information, different numeric values ​​in the range of [0, 1] are assigned to each non-numeric data in personal basic information according to the set weight; for numeric data in personal basic information, the normal distribution S function is used to convert the numeric data in the personal basic information into numeric values ​​distributed in (0, 1).

[0062] Step S5. Compare the personal preference information of the new rental user with the personal preference information of the historical rental users, and filter out the nearest neighbor set of the new rental user in the similar customer group;

[0063] In an optional embodiment, the screening out of the nearest neighbor set of the new rental user in the similar customer group includes the following steps: calculating the similarity between each historical rental user in the similar customer group and the new rental user, and arranging the similarities from high to low, determining the top-ranked historical rental users based on a predetermined similarity threshold, and forming the nearest neighbor set of the new rental user in the similar customer group.

[0064] In an optional embodiment, the calculating of the similarity between each historical rental user and the new rental user in the similar customer group includes the following steps: calculating the similarity between each historical rental user and the new rental user in the similar customer group according to the following formula:

[0065]

[0066] Among them, sim(u i ,u j ) is the historical rental user u in the similar customer group j With new rental user u i The similarity of I ij For new rental users i and historical rental user u j The set of identical items in the personal preference information of ik and r jk They are new rental users u i and historical rental user u j The score of each identical item in the personal preference information, and They are new rental users u i and historical rental user u j The score of each category of the same item in the personal preference information.

[0067] Step S6: Calculate the scores of each historical rental user on the corresponding mobile phone model in the nearest neighbor set, arrange the scores in descending order, select the top-N product set and recommend it.

[0068] In an optional embodiment, the calculation of the score of each historical rental user in the nearest neighbor set for the corresponding mobile phone model includes the following steps: according to the user experience of the historical rental user in the mobile phone rental process, combined with the item attributes of the mobile phone model rented by the historical rental user, and the degree of preference of the historical rental user for the rented mobile phone model, the score of each historical rental user in the nearest neighbor set for the corresponding mobile phone model is calculated. This can improve the accuracy of the score, make the recommended mobile phone model closer to the expectations of the new user, and further enhance the online mobile phone rental experience of the new rental user.

[0069] In an optional embodiment, the step of calculating the score of each historical rental user in the nearest neighbor set for the corresponding mobile phone model comprises the following steps: calculating the score of each historical rental user in the nearest neighbor set for the corresponding mobile phone model according to the following formula:

[0070]

[0071] Among them, r ui is the rating of mobile phone model i by historical rental user u, pop i is the popularity of mobile phone models i, e ui is the user experience coefficient of historical rental user u for mobile phone model i, is the transpose of the corresponding vector of the historical rental user u in the feature matrix P, representing the user feature, Q i is the vector corresponding to mobile phone model i in the feature matrix Q, representing the item features, The preference of historical rental user u for mobile phone model i is obtained by dot product operation of user characteristics and item characteristics.

[0072] The user experience coefficient e ui The calculation formula is as follows:

[0073]

[0074] Among them, e ui is the user experience coefficient of historical rental user u for mobile phone model i, X uik is the number of historical rental user u’s feedback behaviors k on mobile phone model i, W uk is the weight value of the historical rental user u on the feedback behavior k.

[0075] The popularity of the mobile phone model ipop i The calculation formula is as follows:

[0076] pop i =σ(c i , c avg ),

[0077]

[0078] Among them, pop i is the popularity of mobile phone model i, c i is the number of times mobile phone model i is rented in the online rental system, c avg is the average number of times each mobile phone model is rented in the online rental system, c all is the number of times all mobile phone models are rented in the online rental system, c ai It is the number of types of mobile phone models rented in the online rental system.

[0079] The mobile phone recommendation method for deposit-free online rental of the present application collects personal basic information and personal preference information of historical rental users and new rental users, compares the personal basic information of new rental users with the personal basic information of historical rental users, obtains similar customer groups among historical rental users of each mobile phone in the mobile phone model list of the new rental user, and then compares the personal preference information of the new rental user with the personal preference information of the historical rental users, screens out the nearest neighbor set of the new rental user in the similar customer group, and finally calculates the score of each historical rental user in the nearest neighbor set for the corresponding mobile phone model, arranges the obtained scores in order from large to small, selects the top-N product set and recommends it. Therefore, the mobile phone recommendation method for deposit-free online rental of the present application only needs to obtain the personal basic information and personal preference information of historical rental users and new rental users, and does not require the new rental user to provide online behavior data such as website browsing and online shopping, and can obtain more accurate prediction results, recommend more suitable mobile phone models for new rental users, better meet the mobile phone online rental needs of new rental users, and improve the mobile phone online rental experience.

[0080] The embodiments of this specific implementation are all preferred embodiments of the present application, and are not intended to limit the protection scope of the present application. The same components are represented by the same figure marks. Therefore, any equivalent changes made based on the structure, shape, and principle of the present application should be included in the protection scope of the present application.

Claims

1. A method for recommending mobile phones for online rental without deposit, characterized in that: include: Collect basic personal information and personal preference information of historical rental users and new rental users respectively; Obtain the mobile phone model rental demand information of new rental users; Generate a list of mobile phone models that meet the mobile phone model rental needs of new rental users; Compare the basic personal information of the new rental user with that of the historical rental users, and obtain similar customer groups of the new rental user among the historical rental users of each mobile phone in the mobile phone model list through cluster analysis; Compare the personal preference information of the new rental user with the personal preference information of the historical rental users, and filter out the nearest neighbor set of the new rental user in the similar customer group; Calculate the ratings of the corresponding mobile phone models by each historical rental user in the nearest neighbor set, arrange the obtained scores in descending order, select the top-N product set and recommend them.

2. A method for recommending mobile phones for online rental without deposit according to claim 1, characterized in that: The step of screening out the nearest neighbor set of the new rental user in the similar customer group includes: The similarity between each historical rental user and the new rental user in the similar customer group is calculated, and the similarities are arranged from high to low. The top-ranked historical rental users are determined based on a predetermined similarity threshold to form a nearest neighbor set of the new rental user in the similar customer group.

3. A method for recommending mobile phones for online rental without deposit according to claim 2, characterized in that: The calculating the similarity between each historical rental user and the new rental user in the similar customer group includes: calculating the similarity between each historical rental user and the new rental user in the similar customer group according to the following formula: Among them, sim(u i ,u j ) is the historical rental user u in the similar customer group j With new rental user u i The similarity of I ij For new rental users i and historical rental user u j The set of identical items in the personal preference information of ik and r jk They are new rental users u i and historical rental user u j The score of each identical item in the personal preference information, and They are new rental users u i and historical rental user u j The score of each category of the same item in the personal preference information.

4. A method for recommending mobile phones for online rental without deposit according to claim 1, characterized in that: The calculating of the score of the corresponding mobile phone model by each historical rental user in the nearest neighbor set includes: calculating the score of the corresponding mobile phone model by each historical rental user in the nearest neighbor set according to the user experience of the historical rental user in the mobile phone rental process, and combining the item attributes of the mobile phone model rented by the historical rental user, and the preference degree of the historical rental user for the rented mobile phone model.

5. A method for recommending mobile phones for online rental without deposit according to claim 4, characterized in that: The step of calculating the score of the corresponding mobile phone model by each historical rental user in the nearest neighbor set includes: calculating the score of the corresponding mobile phone model by each historical rental user in the nearest neighbor set according to the following formula: Among them, r ui is the rating of mobile phone model i by historical rental user u, pop i is the popularity of mobile phone models i, e ui is the user experience coefficient of historical rental user u for mobile phone model i, is the transpose of the corresponding vector of the historical rental user u in the feature matrix P, representing the user feature, Q i is the vector corresponding to mobile phone model i in the feature matrix Q, representing the item features, The preference of historical rental user u for mobile phone model i is obtained by dot product operation of user characteristics and item characteristics.

6. A method for recommending mobile phones for online rental without deposit according to claim 5, characterized in that: The user experience coefficient e ui The calculation formula is as follows: Among them, e ui is the user experience coefficient of historical rental user u for mobile phone model i, X uik is the number of historical rental user u’s feedback behaviors k on mobile phone model i, W uk is the weight value of the historical rental user u on the feedback behavior k.

7. A method for recommending mobile phones for online rental without deposit according to claim 5, characterized in that: The popularity of the mobile phone model ipop i The calculation formula is as follows: pop i =σ(c i ,c avg ), Among them, pop i is the popularity of mobile phone model i, c i is the number of times mobile phone model i is rented in the online rental system, c avg is the average number of times each mobile phone model is rented in the online rental system, c all is the number of times all mobile phone models are rented in the online rental system, c ai It is the number of types of mobile phone models rented in the online rental system.

8. A method for recommending mobile phones for online rental without deposit according to claim 1, characterized in that: The basic personal information includes the renter's age, gender, education level, housing situation, work unit, occupation, job level, means of transportation, income level and family status.

9. A method for recommending mobile phones for online rental without deposit according to claim 8, characterized in that: The personal basic information of the new rental user is compared with the personal basic information of the historical rental users, and a similar customer group of the new rental user among the historical rental users of each mobile phone in the mobile phone model list is obtained through cluster analysis, including: The basic personal information of new rental users and historical rental users is normalized and converted into numerical information of unified standards. The numerical information of new rental users and historical rental users is compared and arranged from high to low according to similarity. The top-ranked historical rental users are determined based on a predetermined similarity threshold, forming a similar customer group of new rental users among the historical rental users of each mobile phone in the mobile phone model list.

10. A method for recommending mobile phones for online rental without deposit according to claim 9, characterized in that: Normalize the basic personal information of new and historical users and convert it into numerical information of unified standards, including: For non-numeric data in personal basic information, different numeric values ​​in the range of [0, 1] are assigned to each non-numeric data in personal basic information according to the set weight; for numeric data in personal basic information, the normal distribution S function is used to convert the numeric data in the personal basic information into numeric values ​​distributed in (0, 1).

Images