A method for recommending home appliance repair services

Abstract

The present invention is applicable to the field of computer technology, and provides a method, device and terminal device for recommending home appliance repair services, the method comprising: obtaining classification labels of all users; setting classification labels according to the brand of electrical products used by users, product classification and product service life; calculating the initial user similarity of target users according to classification labels, and calculating the recommendation degree of N home appliance repair service items based on target users according to the browsing behavior of target users; obtaining K neighbors of target users and the cell similarity of the cells to which K neighbors belong, recalculating the initial user similarity, and obtaining the secondary user similarity; arranging the home appliance repair service items to be recommended according to the secondary user similarity and recommendation degree, and recommending the arranged home appliance repair service items to target users. The present invention can improve the effect of recommending home appliance repair services based on target users.

Description

Technical Field

[0001] The present invention relates to the field of computer technology, and in particular to a method for recommending home appliance repair services. Background Art

[0002] Preference algorithms have been widely used in the e-commerce industry to recommend products that users may like and increase product sales. However, the home appliance repair industry is different from the e-commerce industry. The user's needs are not closely related to the user's preferences. For users in the home appliance repair industry, they only need home appliance repair services when there is a problem with the home appliance.

[0003] Therefore, the currently commonly used preference algorithms are not applicable to the home appliance repair industry. Summary of the invention

[0004] The main purpose of the present invention is to propose a method, device and terminal device for recommending home appliance repair services, so as to solve the problem that the preference algorithm used in the e-commerce industry in the prior art is not suitable for the home appliance repair industry.

[0005] To achieve the above-mentioned purpose, a first aspect of an embodiment of the present invention provides a method for recommending home appliance repair services, comprising:

[0006] Obtaining classification labels for all users; the classification labels are set according to the brand, product classification and product age of the electrical products used by the users;

[0007] Calculate the initial user similarity of the target user according to the classification label, and calculate the recommendation degree of N home appliance repair service items based on the target user according to the browsing behavior of the target user;

[0008] Obtain K neighbors of the target user and the cell similarities of the cells to which the K neighbors belong, recalculate the initial user similarity, and obtain the secondary user similarity;

[0009] Wherein, N and K are positive integers;

[0010] The home appliance repair service items to be recommended are arranged according to the secondary user similarity and the recommendation degree, and the arranged home appliance repair service items are recommended to the target user.

[0011] In combination with the first aspect of the embodiment of the present invention, in a first implementation of the present invention, the step of calculating the initial user similarity of users according to the classification labels includes:

[0012] A second user group is obtained, and the similarity between the target user and each user in the second user group is calculated according to the classification label as the initial user similarity.

[0013] In combination with the first implementation manner of the first aspect of the embodiment of the present invention, in the second implementation manner of the present invention, obtaining K neighbors of the target user and the cell similarities of the cells to which the K neighbors belong, recalculating the initial user similarity, and obtaining the secondary user similarity includes:

[0014] If the user in the second user group is one of the K neighbors, recalculating the secondary user similarity between the target user and the users in the second user group using the first similarity calculation method;

[0015] If the user in the second user group is not one of the K neighbors, the secondary user similarity between the target user and the users in the second user group is recalculated in a second similarity calculation manner.

[0016] In combination with the second implementation manner of the first aspect of the embodiment of the present invention, in the third implementation manner of the present invention, the first similarity calculation method is set according to the cell similarity between the cell to which the target user belongs and the cell with the highest similarity in its vicinity;

[0017] The second similarity calculation method is set according to the cell similarity between the cell to which the target user belongs and the cell to which the users in the second user group belong.

[0018] In combination with the first aspect of the embodiment of the present invention, in a fourth implementation mode of the present invention, the home appliance repair service items to be recommended are arranged according to the secondary user similarity and the recommendation degree, including:

[0019] For any home appliance repair service item to be recommended, the ranking of any home appliance repair service item to be recommended is calculated based on the recommendation degree given by the user with the highest value in the secondary user similarity and the secondary similarity between the user and the target user.

[0020] A second aspect of an embodiment of the present invention provides a home appliance repair service recommendation device, comprising:

[0021] A classification label acquisition module is used to acquire classification labels of all users; the classification labels are set according to the brand, product category and product service life of the electrical product used by the user;

[0022] An initial user similarity and recommendation degree calculation module, used to calculate the initial user similarity of the target user according to the classification label, and calculate the recommendation degree of N home appliance repair service items based on the target user according to the browsing behavior of the target user;

[0023] A secondary user similarity calculation module is used to obtain K neighbors of the target user and the cell similarity of the cells to which the K neighbors belong, recalculate the initial user similarity, and obtain the secondary user similarity;

[0024] The home appliance repair service item recommendation module is used to arrange the home appliance repair service items to be recommended according to the secondary user similarity and the recommendation degree, and recommend the arranged home appliance repair service items to the target user.

[0025] The third aspect of an embodiment of the present invention provides a terminal device, including a memory, a processor, and a computer program stored in the above memory and executable on the above processor, and when the above processor executes the above computer program, the steps of the method provided in the first aspect are implemented.

[0026] A fourth aspect of an embodiment of the present invention provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the method provided in the first aspect are implemented.

[0027] The embodiment of the present invention proposes a method for recommending home appliance repair services. First, all users are classified using classification tags. When a target user wants to obtain a home appliance repair service item, the initial user similarity of the target user is calculated based on the classification tag, and then similar users are found more accurately through similar cells to obtain secondary user similarity. Combined with the recommendation degree of N home appliance repair service items based on the target user, the arranged home appliance repair service items are provided to the target user. The method for recommending home appliance repair services provided by the embodiment of the present invention fully considers the user needs in the home appliance repair industry through user classification and similar cells, improves the search accuracy of similar users, and thus improves the effect of recommending home appliance repair services based on the target user. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] Figure 1 A schematic diagram of the implementation flow of the method for recommending home appliance repair services provided by an embodiment of the present invention;

[0029] Figure 2 A schematic diagram of the structure of a device for recommending home appliance repair services provided by an embodiment of the present invention.

[0030] The realization of the purpose, functional features and advantages of the present invention will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0031] It should be understood that the specific embodiments described herein are only used to explain the present invention, and are not used to limit the present invention.

[0032] It should be noted that, in this article, the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, an element defined by the sentence "comprises a ..." does not exclude the existence of other identical elements in the process, method, article or device including the element.

[0033] In this document, suffixes such as "module", "component" or "unit" used to represent elements are only used to facilitate the description of the present invention, and have no specific meanings. Therefore, "module" and "component" can be used interchangeably.

[0034] like Figure 1 As shown, an embodiment of the present invention provides a method for recommending home appliance repair services, which is applied to an application software that provides home appliance repair services, and the current user using this application software is the target user. The method for recommending home appliance repair services includes but is not limited to the following steps:

[0035] S101. Obtain classification labels for all users; the classification labels are set according to the brand, product category and product usage period of the electrical products used by the users.

[0036] Before the above step S101, the embodiment of the present invention classifies all users, so each user has a classification label.

[0037] Regarding the classification label of product brand, in specific applications, electrical appliances of different product brands have different service qualities, which affect the service life of the electrical appliances, the frequency of cleaning required, etc.

[0038] Regarding the classification label of product classification, in specific applications, different types of electrical appliances have different material properties, which affect the service life of the electrical appliances, the frequency of cleaning, etc.

[0039] As for the classification label of product age, in specific applications, different years of use of electrical appliances will lead to different fault problems. For example, a computer that has been used for more than five years may face problems mostly caused by hardware failures, while a computer that has been used for about one year may face problems mostly caused by software problems.

[0040] S102: Calculate the initial user similarity of the target user according to the classification label, and calculate the recommendation degree of N home appliance repair service items according to the browsing behavior of the target user.

[0041] In the above step S102, when the initial user similarity of the target user is calculated according to the classification label, the target user has been classified. It can be imagined that the initial user similarity value of two users with the same classification label is higher.

[0042] In one embodiment, the implementation of calculating the initial user similarity of the target user according to the classification label in the above step S102 includes:

[0043] A second user group is obtained, and a similarity between the user and each user in the second user group is calculated according to the classification label as the initial user similarity.

[0044] The second user group may be a plurality of users similar to the target user obtained according to a similarity algorithm, and the acquisition range is relatively large. The embodiment of the present invention also analyzes the similarity between each user in the second user group and the target user through secondary user similarity.

[0045] Combined with the above step S101, the classification label is set according to the brand of the electrical product used by the user, the product classification and the product usage age. In actual application, when calculating the similarity between the target user and each user in the second user group, the calculation is performed according to the product brand, product classification and product usage age. For any user in the second user group, when using the classification label of product brand, let the product brand of the target user u be Cu, and the product brand of any user v in the second user group be Cv, then the product brand similarity C(u, v) between user u and user v is as follows:

[0046]

[0047] When using the product classification label, let the classification of the product used by the target user u be Du, and the classification of the product used by any user v in the second user group be Dv, then the product classification similarity D(u, v) between user u and user v is as follows:

[0048]

[0049] When using the classification label of product age, let the product age used by the target user u be Yu, and the product age used by any user v in the second user group be Yv, then the product age similarity Y(u, v) between user u and user v is defined as follows:

[0050]

[0051] Finally, the initial user similarity between the target user and any user in the second user group is calculated based on the sum of the similarities of the above three features. Different features should have different weights when calculating the initial user similarity. Therefore, for example, the initial user similarity between the target user u and any user v in the second user group is calculated as follows:

[0052] Simp(u,v)=αC(u,v)+βD(u,v)+γY(u,v)

[0053] Among them, α, β, and γ are weight coefficients for calculating the similarity of primary users by product brand, product category, and product age, respectively. Among them, α, β, and γ∈(0,1), and α+β+γ=1.

[0054] It is conceivable that when calculating the similarity of users in different areas and cells, the value can be dynamically adjusted to assign different weights. In addition, after obtaining the initial user similarity through the above steps, decentralized processing can be performed to reduce errors.

[0055] In the above step S102, the recommendation degree of the N home appliance repair service items based on the target user is calculated according to the browsing behavior of the target user. The preference algorithm can be used for calculation, such as calculating the user's score for the product according to the browsing behavior of the target user. Among them, the user's browsing behavior is divided into: page visit, search for goods, add to shopping cart, generate order, complete payment, add rating, add complaint, etc. At the same time, the page visit time needs to be recorded. Adding ratings and adding complaints are the target user's displayed ratings of the product. And page visits, searching for goods, viewing service details, adding to shopping cart, generating orders, and completing payment are the target user's implicit ratings of the product. Based on this, exemplarily, the following scoring formula is used to calculate the recommendation degree of N home appliance repair service items:

[0056] Score ui =R ui +M ui

[0057] R ui is the positive rating of target user u for home appliance repair service item i, and M ui is the negative rating of target user u on home appliance repair service item i. ui The definition is as follows:

[0058] R ui =β 1 f ui +β 2 s ui +β 3 g ui +β 4 o ui+β 5 b ui +β 6 p ui

[0059] f ui Represents the page access behavior of target user u to home appliance repair service item i.

[0060] s ui Represents the search behavior of target user u for home appliance repair service item i.

[0061] g ui Represents the target user u’s behavior of adding home appliance repair service item i to the shopping cart.

[0062] o ui Represents the ordering behavior of target user u for home appliance repair service item i.

[0063] b ui Represents the target user u’s payment behavior for home appliance repair service item i.

[0064] p ui Represents the target user u's behavior of adding evaluation to home appliance repair service item i.

[0065] in,

[0066] β i The table is the score weight of the behavior.

[0067] M ui Represents the negative rating of user u on home appliance repair service item i.

[0068] M ui =β 6 t ui

[0069] t ui Represents the adding behavior of user u to home appliance repair service item i.

[0070] in,

[0071] β i Indicates the score weight of the behavior. There are many ways to determine this value. It can be determined by experts through experience, or it can be randomly distributed first and then determined by the back propagation algorithm based on the quality of the recommendation results. Among them, the user's display score β 5 The weight of β is very important. 1 The weight of is relatively small; β can also be determined by the Pearson correlation coefficient i The specific value of .

[0072] S103: Acquire K neighbors of the target user and the cell similarities of the cells to which the K neighbors belong, recalculate the initial user similarity, and obtain the secondary user similarity.

[0073] In specific applications, the products used by users in different communities in the same area are in the same weather environment and similar usage conditions. In addition, the brands and years of use of electrical appliances used by users in similar communities are also roughly the same. Therefore, the accuracy of initial user similarity can be improved based on similar communities to find similar users more accurately.

[0074] In the embodiment of the present invention, the user similarity is recalculated according to the cell similarity through the above step S103, and the implementation method includes:

[0075] If the user in the second user group is one of the K neighbors, recalculating the secondary user similarity between the target user and the users in the second user group using the first similarity calculation method;

[0076] If the user in the second user group is not one of the K neighbors, the secondary user similarity between the target user and the users in the second user group is recalculated in a second similarity calculation manner.

[0077] Wherein, the first similarity calculation method is set according to the cell similarity between the cell to which the target user belongs and the most similar cell in its vicinity;

[0078] The second similarity calculation method is set according to the cell similarity between the cell to which the target user belongs and the cell to which the users in the second user group belong.

[0079] In the embodiment of the present invention, the cell similarity is the product of the service classification name similarity, the service classification quantity ratio and the product damage feature ratio.

[0080] The embodiment of the present invention also exemplarily uses the formula to illustrate the calculation of the above-mentioned secondary user similarity:

[0081] If the user in the second user group is one of the K neighbors, it means that the user in the second user group and the target user are in the same cell, then

[0082] Secondary user similarity = initial user similarity × (1 + similarity of the most similar community near the area).

[0083] If the user in the second user group is not one of the K neighbors, it means that the user in the second user group and the target user are not in the same cell, then

[0084] Secondary user similarity = user similarity × (1 + cell similarity)

[0085] Among them, the cell similarity = service category name similarity × service category quantity ratio × product damage feature ratio.

[0086] Among them, the similarity of the secondary service name = (A cell service category name ∩ B cell service category name) / (A cell service category name ∪ B cell service category name)

[0087] Service category quantity ratio: When the quantity is greater than 10, quantity ratio = 1-(|A cell Xi category quantity - B cell Xi category quantity) / ([A cell Xi category quantity, B cell Xi category quantity])

[0088] Xi is the name of the TOP I service category;

[0089] [x, y] means taking the smaller value between X and Y;

[0090] When the number of a service is less than 10, the service classification number ratio = 1;

[0091] Product damage characteristic ratio (A cell Pi product damage characteristic ∩B cell Pi product damage characteristic) /

[0092] (Damage characteristics of Pi product in community A ∪ damage characteristics of Pi product in community B).

[0093] S104: Arrange the home appliance repair service items to be recommended according to the secondary user similarity and the recommendation degree, and recommend the arranged home appliance repair service items to the user.

[0094] In an embodiment of the present invention, the implementation method of arranging the home appliance repair service items to be recommended in the above step S104 includes:

[0095] For any home appliance repair service item to be recommended, the ranking of any home appliance repair service item to be recommended is calculated based on the recommendation degree given by the user with the highest value in the secondary user similarity and the secondary similarity between the user and the target user.

[0096] Take any home appliance repair service item to be recommended as an example, and sort them by the values ​​obtained by the following formula:

[0097] Pscoreui=∑ v∈S w uv r vi

[0098] Pscoreui represents the score of target user u on product i.

[0099] S is a set of similar users to user u, ie, the second user group.

[0100] w uvis the secondary user similarity between the target user u and any user v in the second user group.

[0101] r vi is the rating of user v on product i.

[0102] like Figure 2 As shown, the embodiment of the present invention further provides a home appliance repair service recommendation device 20, comprising:

[0103] The classification label acquisition module 21 is used to acquire the classification labels of all users; the classification labels are set according to the brand of the electrical product used by the user, the product category and the product usage period;

[0104] The initial user similarity and recommendation degree calculation module 22 is used to calculate the initial user similarity of the target user according to the classification label, and calculate the recommendation degree of N home appliance repair service items based on the target user according to the browsing behavior of the target user;

[0105] The secondary user similarity calculation module 23 is used to obtain K neighbors of the target user and the cell similarities of the cells to which the K neighbors belong, recalculate the initial user similarity, and obtain the secondary user similarity;

[0106] The home appliance repair service item recommendation module 24 is used to arrange the home appliance repair service items to be recommended according to the secondary user similarity and the recommendation degree, and recommend the arranged home appliance repair service items to the target user.

[0107] An embodiment of the present invention also provides a terminal device including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, each step of the home appliance repair service recommendation method described in the above embodiment is implemented.

[0108] An embodiment of the present invention further provides a storage medium, which is a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, each step in the home appliance repair service recommendation method described in the above embodiment is implemented.

[0109] The embodiments described above are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the foregoing embodiments describe the present invention in detail, those skilled in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or replace some of the technical features therein by equivalents. Such modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included in the protection scope of the present invention.

Claims

1. A method for recommending home appliance repair services, characterized in that: include: Get the category labels of all users; The classification label is set according to the brand, product classification and product age of the electrical product used by the user; Calculate the initial user similarity of the target user according to the classification label, and calculate the recommendation degree of N home appliance repair service items based on the target user according to the browsing behavior of the target user; Obtain K neighbors of the target user and the cell similarities of the cells to which the K neighbors belong, recalculate the initial user similarity, and obtain the secondary user similarity; Wherein, N and K are positive integers; Arranging the home appliance repair service items to be recommended according to the secondary user similarity and the recommendation degree, and recommending the arranged home appliance repair service items to the target user; The calculating the initial user similarity of the user according to the classification label includes: Obtain a second user group, and calculate the similarity between the target user and each user in the second user group according to the classification label as the initial user similarity; If the user in the second user group is one of the K neighbors, it means that the user in the second user group and the target user are in the same cell, then Secondary user similarity = initial user similarity × (1 + similarity of the highest similar community near the region); If the user in the second user group is not one of the K neighbors, it means that the user in the second user group and the target user are not in the same cell, then Secondary user similarity = initial user similarity × (1 + cell similarity) Among them, cell similarity = service category name similarity × service category quantity ratio × product damage feature ratio; Among them, service category name similarity = (A cell service category name ∩ B cell service category name) / (A cell service category name ∪ B cell service category name) Service category quantity ratio: When the quantity is greater than 10, quantity ratio = 1-(|A cell Xi category quantity - B cell Xi category quantity) / ([A cell Xi category quantity, B cell Xi category quantity]) [x, y] means taking the smaller value between x and y; Xi is the name of the TOP I service category; When the number of a service is less than 10, the service classification number ratio = 1; Product damage characteristic ratio = (product damage characteristic of cell A Pi ∩ product damage characteristic of cell B Pi) / (product damage characteristic of cell A Pi ∪ product damage characteristic of cell B Pi); For any home appliance repair service item to be recommended, the ranking of any home appliance repair service item to be recommended is calculated based on the recommendation degree given by the user with the highest value in the secondary user similarity and the secondary similarity between the user and the target user; Take any home appliance repair service item to be recommended as an example, and sort them by the values ​​obtained by the following formula: Pscoreui=∑ v∈S In uv r vi Pscoreui represents the score of target user u on product i; S is a set of similar users to user u, i.e., the second user group; w uv is the secondary user similarity between the target user u and any user v in the second user group; r vi is the rating of user v on product i.

2. A device for recommending home appliance repair services, characterized in that: include: Category label acquisition module, used to obtain category labels of all users; The classification label is set according to the brand, product classification and product age of the electrical product used by the user; An initial user similarity and recommendation degree calculation module, used to calculate the initial user similarity of the target user according to the classification label, and calculate the recommendation degree of N home appliance repair service items based on the target user according to the browsing behavior of the target user; A secondary user similarity calculation module is used to obtain K neighbors of the target user and the cell similarities of the cells to which the K neighbors belong, recalculate the initial user similarity, and obtain the secondary user similarity; A home appliance repair service item recommendation module, used for arranging the home appliance repair service items to be recommended according to the secondary user similarity and the recommendation degree, and recommending the arranged home appliance repair service items to the target user; The calculating the initial user similarity of the user according to the classification label includes: Obtain a second user group, and calculate the similarity between the target user and each user in the second user group according to the classification label as the initial user similarity; If the user in the second user group is one of the K neighbors, it means that the user in the second user group and the target user are in the same cell, then Secondary user similarity = initial user similarity × (1 + similarity of the highest similar community near the region); If the user in the second user group is not one of the K neighbors, it means that the user in the second user group and the target user are not in the same cell, then Secondary user similarity = initial user similarity × (1 + cell similarity) Among them, cell similarity = service category name similarity × service category quantity ratio × product damage feature ratio; Among them, service category name similarity = (A cell service category name ∩ B cell service category name) / (A cell service category name ∪ B cell service category name) Service category quantity ratio: When the quantity is greater than 10, quantity ratio = 1-(|A cell Xi category quantity - B cell Xi category quantity) / ([A cell Xi category quantity, B cell Xi category quantity]) [x, y] means taking the smaller value between x and y; Xi is the name of the TOPI service category; When the number of a service is less than 10, the service classification number ratio = 1; Product damage characteristic ratio = (Product damage characteristic of A cell Pi ∩Product damage characteristic of B cell Pi) / (A cell Pi product damage characteristics ∪ B cell Pi product damage characteristics); For any home appliance repair service item to be recommended, the ranking of any home appliance repair service item to be recommended is calculated based on the recommendation degree given by the user with the highest value in the secondary user similarity and the secondary similarity between the user and the target user; Take any home appliance repair service item to be recommended as an example, and sort them by the values ​​obtained by the following formula: Pscoreui=∑ v∈S In uv r vi Pscoreui represents the score of target user u on product i; S is a set of similar users to user u, i.e., the second user group; w uv is the secondary user similarity between the target user u and any user v in the second user group; r vi is the rating of user v on product i.

3. A terminal device, characterized in that: The method comprises a memory, a processor and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, each step of the method for recommending home appliance repair services as claimed in claim 1 is implemented.

4. A storage medium, the storage medium being a computer-readable storage medium, on which a computer program is stored, characterized in that: When the computer program is executed by a processor, each step of the home appliance repair service recommendation method as claimed in claim 1 is implemented.

Images