Housing recommendation method and device

Abstract

Embodiments of the present invention provide a property recommendation method and apparatus. By determining a target user's first set of properties of interest and a user cluster's second set of properties of interest, the similarity between the first and second sets of properties of interest is calculated. Combined with the user cluster's interest in the second set of properties within a target area, the target user's estimated interest in the difference set of properties within the target area is determined. Finally, properties are recommended to the target user based on the estimated interest. By incorporating the interest levels of users who have viewed a large number of properties within the target area, this method not only avoids data sparsity but also ensures the accuracy of recommendation results, improving the user experience.

Description

Technical Field

[0001] The present invention relates to the field of computer software application technology, and in particular to a housing recommendation method and device. Background Art

[0002] With the proliferation of housing rental and sales platforms, users often look up relevant listings online before visiting properties in person, in order to shorten their search time. Therefore, it is crucial for housing rental and sales platforms to develop a feature that recommends properties to users.

[0003] Typically, when recommending properties, housing rental and sales platforms primarily utilize massive amounts of user preference data to calculate the similarity between individual users' preferences and those of target users. Based on the preferences of users with similar preferences to the target user, they recommend properties that the target user may be interested in, thereby increasing user stickiness, the number of viewings of properties, and transaction volume.

[0004] However, existing property recommendation methods use the interests of individual users as consideration criteria to recommend properties to users with similar preferences. This not only results in inaccurate recommendations but also suffers from data sparsity. Summary of the Invention

[0005] To overcome the above problems or at least partially solve the above problems, an embodiment of the present invention provides a housing recommendation method and apparatus.

[0006] In a first aspect, an embodiment of the present invention provides a housing recommendation method, comprising:

[0007] Determining a first set of properties of interest to a target user and a second set of properties of interest to a user cluster; elements in the first set of properties of interest are used to represent the target user's interest in a first portion of properties within a target area, and elements in the second set of properties of interest are used to represent the user cluster's interest in a second portion of properties within the target area;

[0008] Calculating the similarity between the first set of interested properties and the second set of interested properties, and determining an estimated interest level of the target user in the difference set of properties within the target area based on the similarity and the interest level of the user cluster in the second portion of properties within the target area;

[0009] Based on the estimated level of interest, a property recommendation is made to the target user.

[0010] Preferably, the user cluster groups are determined based on the following method:

[0011] Calculate the preference similarity matrix and degree matrix of sample users;

[0012] Determining a Laplacian matrix and an eigenvector matrix of the Laplacian matrix based on the preference similarity matrix and the degree matrix;

[0013] Based on the eigenvector matrix, the preferences of the sample users are clustered to obtain the user cluster groups.

[0014] Preferably, the step of calculating the preference similarity matrix of the sample users specifically includes:

[0015] determining preference variables of the sample users;

[0016] Calculate the difference between the preference variables of every two sample users;

[0017] Based on the difference, a preference similarity matrix of the sample users is determined.

[0018] Preferably, clustering the sample users based on the eigenvector matrix to obtain the user cluster groups specifically includes:

[0019] Based on the k-means algorithm, all rows in the feature vector matrix are clustered into multiple clusters, each cluster being the user cluster group.

[0020] Preferably, determining the estimated interest level of the target user in the difference set of properties within the target area based on the similarity and the interest level of the user cluster group in the second part of properties within the target area specifically includes:

[0021] Calculating the product of the similarity and the interest level of the user cluster group in the difference set of housing listings in the target area to obtain a first value;

[0022] The sum of the first values ​​corresponding to all user clusters is used as the estimated interest level of the target user in the difference set of properties in the target area.

[0023] Preferably, the interest level of the user cluster group in the second part of the housing listings in the target area is determined in the following manner:

[0024] Determine the number of views and attentions of the user cluster group on the second part of the housing listings in the target area;

[0025] Based on the number of views and the number of attentions, the interest level of the user cluster group in the second part of the housing resources in the target area is determined.

[0026] Preferably, recommending housing to the target user based on the estimated interest level specifically includes:

[0027] sorting the estimated interest of the target user in the second part of the properties in the target area from greatest to least;

[0028] Based on the ranking results, properties are recommended to the target user.

[0029] In a second aspect, an embodiment of the present invention further provides a housing recommendation device, comprising: a housing set determination module, an estimated interest level determination module, and a housing recommendation module.

[0030] The housing set determination module is configured to determine a first housing set of interest to a target user and a second housing set of interest to a user cluster; elements in the first housing set of interest are used to represent the target user's level of interest in a first portion of housing within a target area, and elements in the second housing set of interest are used to represent the user cluster's level of interest in a second portion of housing within the target area;

[0031] The estimated interest level determination module is configured to calculate a similarity between the first set of interested listings and the second set of interested listings, and determine an estimated interest level of the target user in the difference set of listings within the target area based on the similarity and the interest level of the user cluster in the second portion of listings within the target area;

[0032] The housing recommendation module is used to recommend housing to the target user based on the estimated interest level.

[0033] In a third aspect, an embodiment of the present invention further provides an electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the steps of any of the above-described housing recommendation methods are implemented.

[0034] In a fourth aspect, an embodiment of the present invention further provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of any of the above-described housing recommendation methods.

[0035] The housing recommendation method and apparatus provided by embodiments of the present invention determine a target user's first set of interested listings and a user cluster's second set of interested listings. The similarity between the first and second sets of interested listings is then calculated. Combined with the user cluster's interest in the second set of listings within the target area, the estimated interest level of the target user in the difference set of listings within the target area is determined. Finally, based on the estimated interest level, housing recommendations are made to the target user. By incorporating the interest levels of users who have viewed a large number of listings within the target area, this method not only avoids data sparsity but also ensures the accuracy of recommendation results, enhancing the user experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following is a brief introduction 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 invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0037] Figure 1 This is a flowchart of a housing recommendation method provided by an embodiment of the present invention;

[0038] Figure 2 1 is a schematic structural diagram of a housing recommendation device provided by an embodiment of the present invention;

[0039] Figure 3 It is a structural diagram of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0040] To make the objectives, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts shall fall within the scope of protection of the present invention.

[0041] Existing housing rental and sales platforms primarily utilize massive amounts of user preference data to recommend properties. This calculation calculates the similarity between individual users' preferences and those of a target user, and based on the preferences of users with similar preferences, recommends properties that may be of interest to the target user. This improves user engagement, increases viewings, and increases transaction volume. This approach of recommending properties based on individual user interests to users with similar preferences not only results in inaccurate recommendations but also suffers from data sparsity. Therefore, embodiments of the present invention provide a method for recommending properties.

[0042] Figure 1 Schematic diagram of a flow chart of a housing recommendation method provided in an embodiment of the present invention. Figure 1 As shown, the property recommendation method includes:

[0043] S1: Determine a first set of properties of interest to a target user and a second set of properties of interest to a user cluster; the elements in the first set of properties of interest are used to represent the target user's interest in a first portion of properties within a target area, and the elements in the second set of properties of interest are used to represent the user cluster's interest in a second portion of properties within the target area;

[0044] S2: Calculate the similarity between the first set of interested properties and the second set of interested properties, and determine the estimated interest level of the target user in the difference set of properties within the target area based on the similarity and the interest level of the user cluster in the second portion of properties within the target area.

[0045] S3: Recommending properties to the target user based on the estimated interest level.

[0046] Specifically, the housing recommendation method provided in the embodiment of the present invention is used to recommend housing to target users. The target users are users who have housing recommendation needs, that is, users to whom housing needs to be recommended.

[0047] First, step S1 is executed to determine a first set of properties of interest to the target user and a second set of properties of interest to the user cluster. The elements in the first set of properties of interest are used to represent the target user's level of interest in a first portion of properties within a target area. The target area refers to the area to be studied, and the target area includes properties that the target user has viewed as well as properties that the target user has not viewed. The first portion of properties refers to properties within the target area that the target user has viewed, and the specific number of properties in the first portion may include one or more. The target user's level of interest in the first portion of properties within the target area can be determined specifically by the number of times the target user has followed or viewed the first portion of properties, or can be determined by both the number of times the target user has followed and viewed the first portion of properties.

[0048] The elements in the second set of interested listings are used to represent the user cluster's level of interest in the second portion of listings within the target area. The second portion of listings refers to listings already viewed by the user cluster within the target area. User clusters are clusters formed by users who have viewed a large number of listings within the target area based on their preferences for listings. User preferences for listings may include size, price, orientation, location, age, transportation, and amenities. The clustering method can be spectral clustering or general clustering, which is not specifically limited in the embodiments of the present invention. The number of user clusters is at least two, and the specific number of the second portion of listings may include one or more. The target user's level of interest in the second portion of listings within the target area can be determined by the number of times the target user has viewed or viewed the second portion of listings, or can be determined by both the number of times the target user has viewed and viewed the second portion of listings.

[0049] Then, step S2 is executed to calculate the similarity between the first set of interested properties and the second set of interested properties. This similarity represents the degree of similarity between the preferences of the target user and each user cluster. A higher similarity indicates a higher degree of similarity in preferences, while a lower similarity indicates a lower degree of similarity in preferences. When calculating the similarity, the first set of interested properties and the second set of interested properties can be expanded into new sets with the same number of elements, thereby obtaining a third set of interested properties and a fourth set of interested properties, respectively. The third set of interested properties is obtained by expanding the first set of interested properties, and the fourth set of interested properties is obtained by expanding the second set of interested properties. The third set of interested properties and the fourth set of interested properties contain the same number of elements. Except for the elements in the first set of interested properties, all other elements in the third set of interested properties can be 0 or empty. Except for the elements in the second set of interested properties, all other elements in the fourth set of interested properties can be 0 or empty. This ensures that the similarity between the first set of interested properties and the second set of interested properties can be successfully calculated. The similarity between the first set of interested properties and the second set of interested properties can be calculated using the cosine similarity formula, or the similarity can be expressed by calculating the Minkowski distance, Manhattan distance, Euclidean distance, etc., which is not specifically limited in the embodiments of the present invention.

[0050] Since there may be more than one user cluster and more than one second set of properties of interest, there may also be more than one calculated similarity between the first set of properties of interest and the second set of properties of interest. After calculating the similarity between the first set of properties of interest and the second set of properties of interest, the target user's estimated interest in the difference set of properties within the target area can be determined based on the calculated similarity and the user cluster's interest in the second portion of properties within the target area. The difference set of properties within the target area refers to properties within the target area that the target user has not viewed and that have been viewed by the user cluster corresponding to the higher similarity. The estimated interest level can be determined based on the product of the calculated similarity and each user cluster's interest in the second portion of properties within the target area, which is not specifically limited in the embodiments of the present invention.

[0051] Finally, step S3 is executed to recommend properties to the target user based on the estimated interest level determined in step S2. Specifically, the properties in the difference set with the highest estimated interest level may be recommended to the target user. Alternatively, the properties in the difference set may be sorted by estimated interest level and the sorted result recommended to the target user.

[0052] The housing recommendation method provided in embodiments of the present invention determines a target user's first set of interested listings and a user cluster's second set of interested listings. It then calculates the similarity between the first and second sets of interested listings. Combined with the user cluster's interest in the second set of listings within the target area, it determines the target user's estimated interest in the difference set of listings within the target area. Finally, based on the estimated interest, a housing recommendation is made to the target user. By incorporating the interest levels of users who have viewed a large number of listings within the target area, this method not only mitigates data sparsity but also ensures the accuracy of recommendation results, enhancing the user experience.

[0053] Based on the above embodiment, in the housing recommendation method provided in the embodiment of the present invention, the similarity between the first set of interested housing resources and the second set of interested housing resources can be calculated using the following formula:

[0054]

[0055] Among them, w uv is the similarity between the first property set of interest to target user u and the second property set of interest to user cluster v, N(u) is the third property set of interest to target user u, and N(v) is the fourth property set of interest to user cluster v. n is the total number of properties in the target area, that is, both the third property set and the fourth property set contain n elements, Nui is the interest level of target user u in the i-th house in the target area, N vi The interest level of user cluster v in the i-th house in the target area can be represented by a value between 0 and 1, which is not specifically limited in the embodiment of the present invention.

[0056] Based on the above embodiment, in the housing recommendation method provided in the embodiment of the present invention, the user clusters are determined based on the following method:

[0057] Calculate the preference similarity matrix and degree matrix of sample users;

[0058] Determining a Laplacian matrix and an eigenvector matrix of the Laplacian matrix based on the preference similarity matrix and the degree matrix;

[0059] Based on the eigenvector matrix, the preferences of the sample users are clustered to obtain the user cluster groups.

[0060] Specifically, in the embodiment of the present invention, when determining user clusters by clustering, spectral clustering can be used. That is, multiple sample users who have viewed a large number of properties in the target area are selected, and the properties viewed by all sample users can be all properties in the target area. The number of sample users can be set as needed. In the embodiment of the present invention, n can be used to represent the number of sample users. The sample users' preference variables for properties can include preferences such as house area, price, orientation, location, year, transportation, and facilities. All sample users can be represented as X = {x1, x2, ..., x i ,…,x n}, where x i Represents the set of preference variables of the i-th sample user.

[0061] The preference similarity matrix of sample users refers to the preference similarity between any two sample users. When calculating the preference similarity matrix W of sample users, the element w in the i-th row and j-th column of W is ij It represents the similarity of preferences between sample user i and sample user j, which can be expressed by the following formula:

[0062] w ij =w(x i ,x j ).

[0063] The degree matrix of the sample user refers to the diagonal matrix corresponding to each sample user. When calculating the degree matrix D of the sample user, the i-th element d in D is i It represents the sum of the similarities between the preferences of sample user i and all sample users. The specific calculation formula is as follows:

[0064]

[0065] Among them, the degree matrix D is composed of d i An n*n diagonal matrix.

[0066] After calculating the preference similarity matrix and degree matrix of the sample users, the Laplace matrix L and the eigenvector matrix U of the Laplace matrix L are determined based on the preference similarity matrix and the degree matrix. The Laplace matrix L can be determined by the difference between the degree matrix D and the preference similarity matrix W, which can be specifically expressed as:

[0067] L=DW.

[0068] Calculate the eigenvalues ​​of the Laplace matrix L and sort them from small to large. Assuming the number of clusters in spectral clustering is K, we can select the first K eigenvalues ​​in the sorting result and calculate the eigenvectors of each eigenvalue in the first K eigenvalues, which can be expressed as:

[0069] u1, u2, ..., u K ;

[0070] Among them, the eigenvector of each eigenvalue is a column vector; the eigenvector matrix U of the Laplace matrix L can be expressed as:

[0071] U={u1,u2,…,u K}, U∈R n*K .

[0072] Finally, according to the eigenvector matrix U, the preferences of all sample users can be clustered to obtain multiple user cluster groups.

[0073] The clustering method can be: first, the vector of the i-th row of the eigenvector matrix U is replaced by y i ∈R K Represented by, where i=1,2,…,n. Then the eigenvector matrix U can be expressed as:

[0074] U={y1;y2;…;y n}.

[0075] Then, according to the commonly used clustering algorithm in the prior art, the eigenvector matrix U can be re-clustered into multiple clusters, each cluster representing a user cluster group.

[0076] In the embodiment of the present invention, user clusters are determined by spectral clustering, which can make the determined user clusters more accurate and can be used to characterize all sample users with the same or similar preferences.

[0077] Based on the above embodiment, the housing recommendation method provided in the embodiment of the present invention, wherein the step of calculating the preference similarity matrix of sample users specifically includes:

[0078] determining preference variables of the sample users;

[0079] Calculate the difference between the preference variables of every two sample users;

[0080] Based on the difference, a preference similarity matrix of the sample users is determined.

[0081] Specifically, in the embodiments of the present invention, when calculating the preference similarity matrix of sample users, the preference variables of the sample users can be determined first, which can include preferences such as house area, price, orientation, location, age, transportation, and facilities. Then, the difference between the preference variables of each two sample users is calculated; and the preference similarity matrix of the sample users is determined based on all the calculated differences. The specific calculation formula is as follows:

[0082]

[0083] Among them, σ is the variance of the preference variables of all sample users.

[0084] In the embodiment of the present invention, a specific calculation method of the preference similarity matrix of the sample users is provided, which can determine the preference similarity matrix of the sample users more quickly and conveniently.

[0085] Based on the above embodiment, the housing recommendation method provided in the embodiment of the present invention, clustering the sample users based on the eigenvector matrix to obtain the user cluster groups, specifically includes:

[0086] Based on the k-means algorithm, all rows in the feature vector matrix are clustered into multiple clusters, each cluster being the user cluster group.

[0087] Specifically, in the embodiment of the present invention, when clustering sample users according to the eigenvector matrix U to obtain user cluster groups, U can be specifically represented as follows: n Clustering is performed by rows, that is, clustering n rows in U into K rows, that is, obtaining a cluster set C consisting of multiple clusters:

[0088] C=C1、C2、…、C K ;

[0089] Among them, all sample users belonging to each user cluster have the same or similar preferences.

[0090] Based on the above embodiment, the housing recommendation method provided in the embodiment of the present invention, wherein the method determines the estimated interest level of the target user in the difference set of housing listings within the target area based on the similarity and the interest level of the user cluster group in the second portion of housing listings within the target area, specifically includes:

[0091] Calculating the product of the similarity and the interest level of the user cluster group in the difference set of housing listings in the target area to obtain a first value;

[0092] The sum of the first values ​​corresponding to all user clusters is used as the estimated interest level of the target user in the difference set of properties in the target area.

[0093] Specifically, in an embodiment of the present invention, when determining the target user's estimated interest in the difference set of listings within the target area, the product of the similarity and the user cluster group's interest in the difference set of listings within the target area can be first calculated to obtain a first value; then the sum of the first values ​​corresponding to all user cluster groups is used as the target user's estimated interest in the difference set of listings within the target area; and the top K user cluster groups that have viewed the i-th listing and have the highest similarity with the target user u can be selected from all user cluster groups, and the sum of the first values ​​corresponding to the selected top K user cluster groups is used as the target user's estimated interest in the difference set of listings within the target area. This can be specifically shown in the following formula:

[0094]

[0095] Here, v∈S(u,K)∩N(i) represents the top K user clusters v that have viewed the i-th listing and have the highest similarity to the target user u. It should be noted that the above formula represents the target user's estimated interest in all listings within the target area. This includes both listings viewed by the top K user clusters with the highest similarity to the target user and listings viewed by the target user u. Subtracting these two factors yields the target user's estimated interest in the difference set of listings within the target area.

[0096] Based on the above embodiment, in the housing recommendation method provided in the embodiment of the present invention, the user cluster's interest level in the second portion of housing within the target area is determined specifically in the following manner:

[0097] Determine the number of views and attentions of the user cluster group on the second part of the housing listings in the target area;

[0098] Based on the number of views and the number of attentions, the interest level of the user cluster group in the second part of the housing resources in the target area is determined.

[0099] Specifically, in the embodiment of the present invention, when expressing the interest level of the user cluster group in the second part of the housing resources in the target area, the number of views and the number of attentions can be combined. That is, the number of views and the number of attentions of the user cluster group in the second part of the housing resources in the target area are first determined, for example, they are set as z vi and y vi , respectively represent the number of views of user cluster v on the second part of the house i and the number of attentions of user cluster v on the second part of the house i. According to the number of views z vi and the number of followers y vi , the interest level of user cluster v in the second part of listings i in the target area can be determined by the following formula:

[0100] N vi =kz vi +hy vi

[0101] Where k and h are constants.

[0102] In the embodiment of the present invention, the user cluster group's interest in the second part of the housing resources in the target area is determined by considering both the number of views and the number of attentions of the user cluster group on the second part of the housing resources i, which can make the determination result more accurate.

[0103] On the basis of the above embodiment, in an embodiment of the present invention, when determining the target user's interest level in the first part of housing resources in the target area, the target user's interest level in the first part of housing resources in the target area can also be determined in combination with the number of times the target user browses and follows the first part of housing resources, which can make the determination result more accurate.

[0104] It should be noted that, for the third set of properties of interest and the fourth set of properties of interest, all properties therein can be represented by i, that is, no distinction is made between the first part of properties and the second part of properties. This is because the number of elements in the third set of properties of interest and the fourth set of properties of interest are the same, which include both the first part of properties and the second part of properties, but the viewing objects of the two are different.

[0105] Based on the above embodiment, the housing recommendation method provided in the embodiment of the present invention recommends housing to the target user based on the estimated interest level, specifically including:

[0106] sorting the estimated interest of the target user in the second part of the properties in the target area from greatest to least;

[0107] Based on the ranking results, properties are recommended to the target user.

[0108] Specifically, in embodiments of the present invention, when recommending properties to a target user, the target user's estimated interest in the second set of properties within the target area can be sorted from highest to lowest. The top M properties from the sorted results are then selected and recommended to the target user in that order. M properties can be recommended to the target user simultaneously, allowing the target user to see the properties most likely to be of interest first.

[0109] Figure 2 FIG. 1 is a structural diagram of a housing recommendation device provided in an embodiment of the present invention. Figure 2 As shown, the housing recommendation device includes: a housing set determination module 21, an estimated interest level determination module 22 and a housing recommendation module 23.

[0110] The housing set determination module 21 is configured to determine a first housing set of interest to a target user and a second housing set of interest to a user cluster; the elements in the first housing set of interest are used to represent the target user's interest in a first portion of housing within a target area, and the elements in the second housing set of interest are used to represent the user cluster's interest in a second portion of housing within the target area.

[0111] The estimated interest level determination module 22 is configured to calculate the similarity between the first set of interested listings and the second set of interested listings, and determine the estimated interest level of the target user in the difference set of listings within the target area based on the similarity and the interest level of the user cluster in the second portion of listings within the target area.

[0112] The housing recommendation module 23 is configured to recommend housing to the target user based on the estimated interest level.

[0113] Specifically, the functions of each module in the housing recommendation device provided in the embodiment of the present invention correspond one-to-one to the operation process of each step in the above-mentioned method embodiment, and the effects achieved are also consistent. Please refer to the above-mentioned embodiment for details, and no further details will be given in the embodiment of the present invention.

[0114] Based on the above embodiment, the housing recommendation device provided in the embodiment of the present invention further includes: a user cluster group determination module; the user cluster group includes: a first matrix calculation module, a second matrix calculation module and a user cluster group determination submodule;

[0115] The first matrix calculation module is used to calculate the preference similarity matrix and degree matrix of the sample users;

[0116] The second matrix calculation module is used to: determine a Laplacian matrix and an eigenvector matrix of the Laplacian matrix based on the preference similarity matrix and the degree matrix;

[0117] The user cluster group determination submodule is used to cluster the preferences of sample users based on the eigenvector matrix to obtain the user cluster groups.

[0118] Based on the above embodiment, in the housing recommendation device provided in the embodiment of the present invention, the first matrix calculation module is specifically configured to:

[0119] determining preference variables of the sample users;

[0120] Calculate the difference between the preference variables of every two sample users;

[0121] Based on the difference, a preference similarity matrix of the sample users is determined.

[0122] Based on the above embodiment, in the housing recommendation device provided in the embodiment of the present invention, the user cluster determination submodule is specifically configured to:

[0123] Based on the k-means algorithm, all rows in the feature vector matrix are clustered into multiple clusters, each cluster being the user cluster group.

[0124] Based on the above embodiment, in the housing recommendation device provided in the embodiment of the present invention, the estimated interest level determination module is specifically configured to:

[0125] Calculating the product of the similarity and the interest level of the user cluster group in the difference set of housing listings in the target area to obtain a first value;

[0126] The sum of the first values ​​corresponding to all user clusters is used as the estimated interest level of the target user in the difference set of properties in the target area.

[0127] Based on the above embodiment, the housing recommendation device provided in the embodiment of the present invention further includes: an interest level determination module; the interest level determination module is configured to:

[0128] Determine the number of views and attentions of the user cluster group on the second part of the housing listings in the target area;

[0129] Based on the number of views and the number of attentions, the interest level of the user cluster group in the second part of the housing resources in the target area is determined.

[0130] Based on the above embodiment, in the housing recommendation device provided in the embodiment of the present invention, the housing recommendation module is specifically configured to:

[0131] sorting the estimated interest of the target user in the second part of the properties in the target area from greatest to least;

[0132] Based on the ranking results, properties are recommended to the target user.

[0133] Figure 3 An example of a physical structure diagram of an electronic device is shown below. Figure 3 As shown, the electronic device may include: a processor 310, a communications interface 320, a memory 330, and a communications bus 340, wherein the processor 310, the communications interface 320, and the memory 330 communicate with each other via the communications bus 340. The processor 310 may invoke logic instructions in the memory 330 to execute a housing recommendation method, including: determining a first set of housing listings of interest to a target user and a second set of housing listings of interest to a user cluster; elements in the first set of housing listings of interest are used to represent the target user's interest in a first portion of housing listings within a target area, and elements in the second set of housing listings of interest are used to represent the user cluster's interest in a second portion of housing listings within the target area; calculating the similarity between the first set of housing listings of interest and the second set of housing listings of interest, and determining an estimated interest level of the target user in a difference set of housing listings within the target area based on the similarity and the user cluster's interest level in the second portion of housing listings within the target area; and recommending housing listings to the target user based on the estimated interest level.

[0134] It should be noted that the electronic device in this embodiment can be a server, a PC, or other devices in specific implementation, as long as its structure includes the following: Figure 3 The processor 310, communication interface 320, memory 330, and communication bus 340 shown are shown. The processor 310, communication interface 320, and memory 330 communicate with each other via the communication bus 340, and the processor 310 can call the logic instructions in the memory 330 to execute the above method. This embodiment does not limit the specific implementation form of the electronic device.

[0135] In addition, the logic instructions in the above-mentioned memory 330 can be implemented in the form of a software functional unit and can be stored in a computer-readable storage medium when sold or used as an independent product. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

[0136] On the other hand, an embodiment of the present invention further provides a computer program product, which includes a computer program stored on a non-transitory computer-readable storage medium, and the computer program includes program instructions. When the program instructions are executed by a computer, the computer can execute the housing recommendation method provided by the above-mentioned method embodiments, including: determining a first set of housing sources of interest for a target user and a second set of housing sources of interest for a user cluster group; the elements in the first set of housing sources of interest are used to characterize the target user's interest in a first part of housing sources in a target area, and the elements in the second set of housing sources of interest are used to characterize the user cluster group's interest in a second part of housing sources in the target area; calculating the similarity between the first set of housing sources of interest and the second set of housing sources of interest, and determining the target user's estimated interest in the difference set of housing sources in the target area based on the similarity and the user cluster group's interest in the second part of housing sources in the target area; and recommending housing to the target user based on the estimated interest.

[0137] On the other hand, an embodiment of the present invention further provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, is implemented to perform the housing recommendation method provided by the above embodiments, including: determining a first set of housing sources of interest for a target user and a second set of housing sources of interest for a user cluster group; the elements in the first set of housing sources of interest are used to characterize the degree of interest of the target user in a first part of housing sources in a target area, and the elements in the second set of housing sources of interest are used to characterize the degree of interest of the user cluster group in a second part of housing sources in the target area; calculating the similarity between the first set of housing sources of interest and the second set of housing sources of interest, and determining the estimated degree of interest of the target user in the difference set of housing sources in the target area based on the similarity and the degree of interest of the user cluster group in the second part of housing sources in the target area; and recommending housing to the target user based on the estimated degree of interest.

[0138] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, i.e., they may be located in one location or distributed across multiple network units. Some or all of the modules may be selected based on actual needs to achieve the objectives of the present embodiment. Persons of ordinary skill in the art will be able to understand and implement the present invention without inventive effort.

[0139] Through the description of the above embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus a necessary general hardware platform, or of course, by hardware. Based on this understanding, the essence of the above technical solution or the part that contributes to the existing technology can be embodied in the form of a software product. The computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a magnetic disk, an optical disk, etc., and includes a number of instructions for enabling 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 certain parts of the embodiments.

[0140] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention 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 various embodiments of the present invention.

Claims

1. A housing recommendation method, characterized in that: include: Determine a first set of properties of interest to the target user and a second set of properties of interest to the user cluster; The elements in the first set of interested properties represent the target user's interest in the first set of properties within the target area. The elements in the second set of interested properties represent the user cluster's interest in the second set of properties within the target area. The first set of properties refers to properties within the target area that the target user has viewed, while the second set of properties refers to properties within the target area that the user cluster has viewed. User clusters are clusters formed by clustering users who have viewed a large number of properties within the target area based on their preferences for properties. Calculating the similarity between the first set of interested properties and the second set of interested properties, and determining the target user's estimated interest in the difference set of properties in the target area based on the similarity and the user cluster group's interest in the second part of the properties in the target area, including: calculating the product of the similarity and the user cluster group's interest in the difference set of properties in the target area to obtain a first value; taking the sum of the first values ​​corresponding to all user cluster groups as the target user's estimated interest in the difference set of properties in the target area; or selecting the top K user cluster groups that have viewed the i-th property and have the highest similarity with the target user from all user cluster groups, and taking the sum of the first values ​​corresponding to the selected top K user cluster groups as the target user's estimated interest in the difference set of properties in the target area; the difference set of properties in the target area refers to properties in the target area that have not been viewed by the target user and are viewed by the user cluster groups corresponding to the similarity that falls within a preset range; Recommend properties to target users based on estimated interest levels.

2. The housing recommendation method according to claim 1, characterized in that: The user cluster groups are determined based on the following method: Calculate the preference similarity matrix and degree matrix of sample users; Determining a Laplacian matrix and an eigenvector matrix of the Laplacian matrix based on the preference similarity matrix and the degree matrix; Based on the eigenvector matrix, the preferences of the sample users are clustered to obtain the user cluster groups.

3. The housing recommendation method according to claim 2, characterized in that: The calculating of the preference similarity matrix of the sample users specifically includes: determining preference variables of the sample users; Calculate the difference between the preference variables of every two sample users; Based on the difference, a preference similarity matrix of the sample users is determined.

4. The housing recommendation method according to claim 2, characterized in that: Clustering the sample users based on the eigenvector matrix to obtain the user cluster groups specifically includes: Based on the k-means algorithm, all rows in the feature vector matrix are clustered into multiple clusters, each cluster being the user cluster group.

5. The housing recommendation method according to claim 1, wherein: The user cluster's interest in the second portion of properties within the target area is determined specifically by: Determine the number of views and attentions of the user cluster group on the second part of the housing listings in the target area; Based on the number of views and the number of attentions, the interest level of the user cluster group in the second part of the housing resources in the target area is determined.

6. The housing recommendation method according to any one of claims 1 to 5, characterized in that: The recommending of properties to the target user based on the estimated interest level specifically includes: sorting the estimated interest of the target user in the second part of the properties in the target area from greatest to least; Based on the ranking results, properties are recommended to the target user.

7. A housing recommendation device, characterized in that: include: A housing set determination module is used to determine a first housing set of interest to a target user and a second housing set of interest to a user cluster; The elements in the first set of interested properties represent the target user's interest in the first set of properties in the target area. The elements in the second set of interested properties represent the user cluster's interest in the second set of properties in the target area. The first set of properties are properties that the target user has viewed in the target area, and the second set of properties are properties that the user cluster has viewed in the target area. User clusters are clusters formed by clustering users who have viewed a large number of properties in the target area based on their preferences for properties. an estimated interest level determination module, configured to calculate the similarity between the first set of interested listings and the second set of interested listings, and determine the target user's estimated interest level in the difference set of listings within the target area based on the similarity and the user cluster's interest level in the second set of listings within the target area; The difference set of listings within the target area refers to listings within the target area that have not been viewed by the target user and have been viewed by the user cluster group whose similarity falls within a preset range. The method is further configured to: calculate the product of the similarity and the user cluster group's interest in the difference set of listings within the target area to obtain a first value; take the sum of the first values ​​corresponding to all user cluster groups as the target user's estimated interest in the difference set of listings within the target area; or, select the top K user cluster groups that have viewed the i-th listing and have the highest similarity with the target user from all user cluster groups, and take the sum of the first values ​​corresponding to the selected top K user cluster groups as the target user's estimated interest in the difference set of listings within the target area. The property recommendation module is used to recommend properties to target users based on the estimated level of interest.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the program, the steps of the housing recommendation method according to any one of claims 1 to 6 are implemented.

9. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the housing recommendation method according to any one of claims 1 to 6 are implemented.

Images