Employee Dormitory Recommendation Method, Device, Electronic Device, and Storage Medium

Abstract

The present invention provides a method, device, electronic device and storage medium for recommending employee dormitories. By determining the personal information and basic intended dormitory information of the candidates for the dormitory, as well as the recommended sampling step length and the number of recommended dormitories, obtaining the occupancy information of the occupants of the occupied dormitories that match the basic intended dormitory information, and based on the occupancy information and the personal information of the candidates for the dormitory, determining a set of matching occupants that match the candidates for the dormitory. Subsequently, the occupants in the set of matching occupants and the candidates for the dormitory are sorted, and based on the sorting serial number of the candidates for the dormitory and the recommended sampling step length, the sampled occupants are determined. Thus, based on the dormitory attribute information of the sampled occupants, the proportion of various room types is determined, and based on the proportion of various room types and the number of recommended dormitories, the recommended dormitories to be recommended to the candidates for the dormitory are selected from the current available vacant housing sources, improving the room selection efficiency and the fairness of dormitory allocation.

Description

Technical Field

[0001] The present invention relates to the technical field of data recommendation, and particularly to a method, device, electronic device and storage medium for recommending staff dormitories. Background Art

[0002] In the process of allocating staff dormitories, how to balance the personal choices of staff and the principle of fair distribution has always been an important problem in human resource management. Especially for newly recruited employees, due to limited dormitory resources, especially when dormitory conditions are relatively tight, they often cannot choose a suitable dormitory that meets their needs. This situation not only brings certain dissatisfaction to the staff, but also makes the dormitory allocation process appear more subjective and unfair, and is likely to cause unnecessary conflicts and contradictions.

[0003] In addition, the efficiency of dormitory allocation is also an issue that cannot be ignored. In some enterprises with loose management, there is often no effective management system and operation process for dormitory allocation, resulting in untimely update of dormitory information and slow execution of allocation decisions. Staff often need to wait for a long time to complete the move-in. This inefficient allocation method not only wastes a lot of time and resources, but may also affect the daily work arrangements of the staff, thereby reducing their work enthusiasm and satisfaction.

[0004] Therefore, how to improve the efficiency of dormitory allocation on the basis of providing fair distribution as much as possible has become an important challenge faced by many enterprise human resource managers. Summary of the Invention

[0005] The present invention provides a method, device, electronic device and storage medium for recommending staff dormitories to solve the defect that the efficiency of existing dormitory allocation is poor and it is difficult to ensure fair distribution.

[0006] The present invention provides a method for recommending staff dormitories, including:

[0007] Determine the personal information and basic intended dormitory information of the dormitory candidates, as well as the recommended sampling step size and the number of dormitory recommendations;

[0008] Obtain the occupancy information of the occupants of the occupied dormitories that match the basic intended dormitory information, and determine a set of matching occupants that match the dormitory candidates based on the occupancy information and the personal information of the dormitory candidates;

[0009] Sort the occupants in the set of matching occupants and the dormitory candidates, and determine the sampled occupants based on the sorting sequence number of the dormitory candidates and the recommended sampling step size;

[0010] Based on the dormitory attribute information of the sampled occupants, determine the proportion of each type of room, and based on the proportion of each type of room, the room type information of the current available rooms, and the number of dormitory recommendations, select recommended dormitories from the current available rooms to be recommended to the candidates for the dormitories.

[0011] According to an employee dormitory recommendation method provided by the present invention, the determining a set of matching occupants matching the candidate for the dormitory based on the occupant information and the personal information of the candidate for the dormitory includes:

[0012] Based on the occupant information of each occupant and the personal information of the candidate for the dormitory, construct an occupant vector for each occupant and a personal vector for the candidate for the dormitory respectively;

[0013] Based on the occupant vectors of each occupant, create a set of auxiliary person vectors; the set of auxiliary person vectors contains multiple auxiliary person vectors;

[0014] Fuse the occupant vectors of each occupant and the set of auxiliary person vectors to obtain a set of fused vectors, and perform density estimation on the set of fused vectors to obtain the density value of each vector in the set of fused vectors;

[0015] Obtain vectors with a preset number of density values higher than a preset threshold as central vectors, and perform clustering based on the central vectors to obtain multiple clusters;

[0016] Based on the occupants corresponding to the occupant vectors included in the cluster where the personal vector of the candidate for the dormitory is located, construct the set of matching occupants.

[0017] According to an employee dormitory recommendation method provided by the present invention, the creating a set of auxiliary person vectors based on the occupant vectors of each occupant includes:

[0018] Generate multiple virtual person vectors based on the generator of the generative adversarial network, classify the multiple virtual person vectors and the occupant vectors of each occupant based on the discriminator of the generative adversarial network to obtain the classification probabilities of each virtual person vector and occupant vector, and update the generator and discriminator of the generative adversarial network based on the classification probabilities of each virtual person vector and occupant vector;

[0019] Use the updated generator of the generative adversarial network to generate multiple auxiliary person vectors, and construct the set of auxiliary person vectors based on the multiple auxiliary person vectors.

[0020] A method for recommending employee dormitories provided by the present invention, updating the generator and discriminator of the generative adversarial network based on the classification probabilities of each virtual personnel vector and the occupancy personnel vector, includes:

[0021] Updating the model parameters of the discriminator based on the classification probabilities of each virtual personnel vector and the difference between the classification probability of each occupancy personnel vector and 1;

[0022] Updating the model parameters of the generator based on the difference between the classification probability of each virtual personnel vector and 1.

[0023] A method for recommending employee dormitories provided by the present invention, obtaining vectors with density values higher than a preset threshold as central vectors, includes:

[0024] Determining virtual personnel vectors with density values higher than the preset threshold as reference personnel vectors;

[0025] Determining occupancy personnel vectors with density values higher than the preset threshold as candidate personnel vectors;

[0026] Calculating the distances between each candidate personnel vector and each reference personnel vector, and selecting a preset number of candidate personnel vectors with the closest distances as the central vectors.

[0027] A method for recommending employee dormitories provided by the present invention, sorting the occupancy personnel in the matching occupancy personnel set and the dormitory candidate personnel, and determining the sampled occupancy personnel based on the sorting serial number of the dormitory candidate personnel and the recommended sampling step size, includes:

[0028] Sorting the occupancy personnel in the matching occupancy personnel set and the dormitory candidate personnel based on the rank, service time, total working time, children's education level, and border support priority information of the occupancy personnel in the matching occupancy personnel set and the dormitory candidate personnel;

[0029] Based on the sorting serial number of the dormitory candidate personnel and the recommended sampling step size, determining the occupancy personnel whose difference between the sorting serial numbers before and after the dormitory candidate personnel and the sorting serial number of the dormitory candidate personnel is less than the recommended sampling step size as the sampled occupancy personnel.

[0030] The present invention also provides an employee dormitory recommendation device, including:

[0031] An information determination unit, configured to determine the personal information and basic intended dormitory information of the dormitory candidate personnel, as well as the recommended sampling step size and the number of dormitory recommendations;

[0032] A similar personnel acquisition unit, configured to acquire the occupancy information of the occupants in the occupied dormitories that match the basic intended dormitory information, and determine a set of matching occupants that match the dormitory candidate based on the occupancy information and the personal information of the dormitory candidate;

[0033] A sampling unit, configured to sort the occupants in the set of matching occupants and the dormitory candidate, and determine sampled occupants based on the sorting serial number of the dormitory candidate and the recommended sampling step;

[0034] A recommended dormitory determination unit, configured to determine the proportion of each room type based on the dormitory attribute information of the sampled occupants, and select recommended dormitories to be recommended to the dormitory candidate from the current available vacant housing based on the proportion of each room type, the room type information of the current available vacant housing, and the number of recommended dormitories.

[0035] According to an employee dormitory recommendation device provided by the present invention, the determining a set of matching occupants that match the dormitory candidate based on the occupancy information and the personal information of the dormitory candidate includes:

[0036] Based on the occupancy information of each occupant and the personal information of the dormitory candidate, respectively construct an occupancy vector for each occupant and a personal vector for the dormitory candidate;

[0037] Based on the occupancy vectors of each occupant, create a set of auxiliary person vectors; the set of auxiliary person vectors contains multiple auxiliary person vectors;

[0038] Fuse the occupancy vectors of each occupant and the set of auxiliary person vectors to obtain a set of fused vectors, and perform density estimation on the set of fused vectors to obtain the density value of each vector in the set of fused vectors;

[0039] Obtain vectors with density values higher than a preset threshold for a preset number as central vectors, and perform clustering based on the central vectors to obtain multiple clusters;

[0040] Based on the occupants corresponding to the occupancy vectors included in the cluster where the personal vector of the dormitory candidate is located, construct the set of matching occupants.

[0041] The present invention also provides an electronic device, including a memory, a processor, and a computer program stored on the memory and executable on the processor, where the processor implements the employee dormitory recommendation method as described in any one of the above when executing the program.

[0042] The present invention also provides a non-transitory computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the method for recommending staff dormitories as described in any one of the above is implemented.

[0043] The present invention also provides a computer program product, including a computer program. When the computer program is executed by a processor, the method for recommending staff dormitories as described in any one of the above is implemented.

[0044] The method, device, electronic device and storage medium for recommending staff dormitories provided by the present invention determine the personal information and basic intended dormitory information of the candidates for the dormitory, as well as the recommended sampling step size and the number of recommended dormitories. The information of the occupants of the occupied dormitories that match the basic intended dormitory information is obtained, and based on the information of the occupants and the personal information of the candidates for the dormitory, a set of matching occupants that match the candidates for the dormitory is determined. Subsequently, the occupants in the set of matching occupants and the candidates for the dormitory are sorted, and based on the sorting serial number of the candidates for the dormitory and the recommended sampling step size, the sampled occupants are determined. Thus, based on the dormitory attribute information of the sampled occupants, the proportion of various room types is determined, and based on the proportion of various room types, the room type information of the current available vacant rooms and the number of recommended dormitories, the recommended dormitories to be recommended to the candidates for the dormitory are selected from the current available vacant rooms, improving the recommendation accuracy, thereby improving the room selection efficiency and at the same time enhancing the fairness of dormitory allocation. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings in the following description are some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0046] Figure 1 is a schematic flowchart of the method for recommending staff dormitories provided by the present invention;

[0047] Figure 2 is a schematic flowchart of the method for constructing a set of matching occupants provided by the present invention;

[0048] Figure 3 is a schematic structural diagram of the device for recommending staff dormitories provided by the present invention;

[0049] Figure 4 is a schematic structural diagram of the electronic device provided by the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0050] To make the objectives, technical solutions and advantages of the present invention clearer, the technical solutions in the present invention will be clearly and completely described below with reference to the accompanying drawings in the present invention. Apparently, the described embodiments are some but not all of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments in the present invention without creative efforts shall fall within the protection scope of the present invention.

[0051] Figure 1 is a schematic flowchart of the employee dormitory recommendation method provided by the present invention. As Figure 1 shown, the method includes:

[0052] Step 110: Determine the personal information and basic intended dormitory information of the dormitory candidates, as well as the recommended sampling step size and the number of recommended dormitories.

[0053] Step 120: Obtain the occupancy information of the occupants of the occupied dormitories that match the basic intended dormitory information, and determine a set of matching occupants that match the dormitory candidates based on the occupancy information and the personal information of the dormitory candidates.

[0054] Step 130: Sort the occupants in the set of matching occupants and the dormitory candidates, and determine the sampled occupants based on the sorting serial number of the dormitory candidates and the recommended sampling step size.

[0055] Step 140: Determine the proportion of each type of room based on the dormitory attribute information of the sampled occupants, and select the recommended dormitories to be recommended to the dormitory candidates from the current available vacant dormitories based on the proportion of each type of room, the room type information of the current available vacant dormitories, and the number of recommended dormitories.

[0056] Here, the personal information including work rank, tenure, total working hours, children's education level, priority for border support, work performance, physical condition, and family situation, etc. can be entered through the basic information collection system of the dormitory candidates, and at the same time, record their basic intention for the dormitory (such as the community, the number of rooms, etc.) as the basic intended dormitory information to provide complete personalized demand data and lay the foundation for accurate recommendation. In addition, the recommended sampling step size and the number of recommended dormitories can also be set (such as recommending 3 or 5 dormitories).

[0057] Subsequently, it is possible to obtain the occupancy information of all the occupants in the occupied dormitories that currently match the basic intended dormitory information (such as the same community and room information). Among them, the types of information included in the occupancy information of the occupants are the same as the personal information of any of the dormitory candidates, which will not be elaborated here. Based on the personal information of the dormitory candidates and the occupancy information of the above-mentioned occupants, it is possible to screen out the occupants with personal situations similar to those of the dormitory candidates and construct a set of matching occupants as the sampling object for subsequent recommendations. Here, by obtaining the occupants with personal situations similar to those of the dormitory candidates as the sampling object for recommendations, it is possible to recommend to the dormitory candidates housing units similar to those where the occupants with personal situations similar to theirs live, improving the recommendation accuracy, thereby enhancing the housing selection efficiency and at the same time enhancing the fairness of dormitory allocation.

[0058] In some embodiments, as Figure 2 shown, the following method can be used to determine the set of matching occupants matching the dormitory candidate:

[0059] Step 210, based on the occupancy information of each occupant and the personal information of the dormitory candidate, respectively construct the occupancy vector of each occupant and the personal vector of the dormitory candidate;

[0060] Step 220, based on the occupancy vectors of each occupant, create a set of auxiliary person vectors; the set of auxiliary person vectors contains multiple auxiliary person vectors;

[0061] Step 230, fuse the occupancy vectors of each occupant and the set of auxiliary person vectors to obtain a set of fused vectors, and perform density estimation on the set of fused vectors to obtain the density value of each vector in the set of fused vectors;

[0062] Step 240, obtain a preset number of vectors with density values higher than a preset threshold as the central vectors, and perform clustering based on the central vectors to obtain multiple clusters;

[0063] Step 250, based on the occupants corresponding to the occupancy vectors included in the cluster where the personal vector of the dormitory candidate is located, construct the set of matching occupants.

[0064] Here, considering that the dimensions of the occupancy information / personal information of the occupants / candidates for the dormitory are relatively high and the data is sparse, directly using the clustering method to process the occupancy information / personal information of the occupants / candidates for the dormitory to obtain occupants similar to the candidates for the dormitory may result in insufficient clustering accuracy due to complex shapes or uneven inter-cluster densities. Secondly, conventional clustering methods, such as K-Means and K-Medoids, rely on the initial center points obtained by random initialization, and their performance is affected by the initial center points, so the clustering accuracy cannot be guaranteed.

[0065] In response, first, based on the occupancy information of each occupant and the personal information of the candidates for the dormitory, the occupancy vectors of each occupant and the personal vectors of the candidates for the dormitory can be constructed respectively. Among them, the one-hot vector of the occupancy information of each occupant can be obtained as the above-mentioned occupancy vector, and the one-hot vector of the personal information of the candidates for the dormitory can be obtained as the above-mentioned personal vector. Subsequently, in order to improve the clustering accuracy, an auxiliary person vector set can be created based on the occupancy vectors of each occupant. The auxiliary person vector set contains multiple auxiliary person vectors, and the distribution of the auxiliary person vectors in the auxiliary person vector set in the vector space is similar to the distribution of the occupancy vectors of each occupant in the vector space. By constructing auxiliary person vectors with a data distribution similar to the occupancy vectors, additional data can be generated to make up for the above deficiencies of the real data, provide more potential information, and at the same time, points close to the distribution boundary may also be generated, which is of great help in improving the clustering accuracy for the boundary fuzzy problem existing in the current scenario.

[0066] In some other embodiments, multiple virtual person vectors can be generated based on the generator of the generative adversarial network, and the discriminator of the generative adversarial network can classify the above-mentioned multiple virtual person vectors and the occupancy vectors of each occupant to obtain the classification probabilities of each virtual person vector and occupancy vector. Among them, random noise can be input into the generative adversarial network to achieve data generation. The classification probabilities of the virtual person vectors and occupancy vectors output by the discriminator represent the probabilities that the discriminator believes that any virtual person vector / any occupancy vector conforms to the data distribution of the occupancy vectors. Based on the classification probabilities of each virtual person vector and occupancy vector, the generator and discriminator of the generative adversarial network are updated, so as to guide the generative adversarial network to learn the data distribution of the occupancy vectors, and then obtain the ability to generate data that is similar enough to the data distribution of the occupancy vectors.

[0067] Among them, the model parameters of the discriminator can be updated based on the classification probabilities of each virtual person vector and the differences between the classification probabilities of each resident vector and 1, and the model parameters of the generator can be updated based on the differences between the classification probabilities of each virtual person vector and 1. Specifically, the model parameters of the discriminator and the generator can be updated alternately. When the discriminator needs to be updated, multiple virtual person vectors can be generated by the generator, and after the discriminator classifies the above-mentioned multiple virtual person vectors and the resident vectors of each resident to obtain the classification probabilities of each virtual person vector and the resident vector, the model parameters of the discriminator can be updated based on the classification probabilities of each virtual person vector and the differences between the classification probabilities of each resident vector and 1. When the generator needs to be updated, multiple virtual person vectors are generated again by the generator, the discriminator classifies the above-mentioned multiple virtual person vectors and the resident vectors of each resident to obtain the classification probabilities of each virtual person vector and the resident vector, and the model parameters of the generator are updated based on the differences between the classification probabilities of each virtual person vector and 1.

[0068] In some embodiments, considering that the current training samples are few and may be uneven, which is likely to cause the problem of mode collapse, that is, the uneven training samples result in uneven training and the discriminator being too strong, so that the generator only learns to generate samples of a single mode or a few modes and ignores other modes in the data distribution. Therefore, to alleviate this problem, when updating the discriminator, the first model loss can be determined based on the difference between the classification probability of each resident vector and 1, the second model loss can be determined based on the larger value between the difference between the preset control boundary and the classification probability of the virtual person vector and 0, and the total model loss can be determined based on the first model loss and the second model loss to update the parameters of the discriminator based on the total model loss.

[0069] Subsequently, multiple auxiliary person vectors can be generated by the generator of the updated generative adversarial network, and an auxiliary person vector set can be constructed based on the multiple auxiliary person vectors. Similar to the model training stage, random noise can be input into the generative adversarial network to obtain the auxiliary person vectors output by the network.

[0070] To select the center point that can better represent the cluster center as the clustering basis to improve the clustering accuracy, the resident vectors of each resident and the auxiliary person vector set can be fused to obtain a fused vector set, and density estimation can be performed on each vector (resident vector or auxiliary person vector) in the fused vector set to obtain the density value of each vector in the fused vector set. Subsequently, a preset number of vectors with density values higher than the preset threshold are obtained as the center vectors, and clustering is performed based on the above center vectors using the K-Means or K-Medoids algorithm to obtain multiple clusters.

[0071] In some embodiments, considering that the number of occupants is limited and the number of vectors during clustering is also limited, in order to make more comprehensive use of the data distribution characteristics, virtual person vectors can be used as an effective supplement to obtain more representative central vectors as clustering centers. Among them, virtual person vectors with density values higher than a preset threshold can be determined as reference person vectors, and occupant vectors with density values higher than the preset threshold can be determined as candidate person vectors. Then, the distance dist ij (dist ij represents the distance between the i-th candidate person vector and the j-th reference person vector) is calculated pairwise for each candidate person vector and each reference person vector, and a preset number of the candidate person vectors with the closest distances are selected as the central vectors.

[0072] After obtaining the clustering result, a matching occupant set can be constructed based on the occupants corresponding to the occupant vectors included in the cluster where the personal vector of the dormitory candidate is located.

[0073] The occupants in the above-mentioned matching occupant set and the dormitory candidates are sorted, and based on the sorting serial number of the dormitory candidates and the above-mentioned recommended sampling step size, sampling occupants are determined from the matching occupant set. Then, based on the dormitory attribute information of each sampling occupant (including room type information, such as floor and orientation, etc.), the proportion of each room type in the dormitories where these sampling occupants live is determined, and based on the proportion of each room type, the room type information of the current available empty rooms, and the above-mentioned dormitory recommendation quantity, recommended dormitories are selected from the current available empty rooms to be recommended to the dormitory candidates, realizing precise recommendation of staff dormitories. Here, several dormitories with the highest proportion can be selected as the above-mentioned recommended dormitories based on the dormitory recommendation quantity. In some embodiments, based on the rank, tenure, total working hours, children's education level, and border support priority information of the occupants in the matching occupant set and the dormitory candidates, the occupants in the matching occupant set and the dormitory candidates are sorted, and based on the sorting serial number of the dormitory candidates and the above-mentioned recommended sampling step size, the occupants whose difference between the sorting serial numbers before and after the dormitory candidates and the sorting serial number of the dormitory candidates is less than the recommended sampling step size are determined as sampling occupants.

[0074] In summary, the method provided by the embodiments of the present invention determines the personal information and basic intended dormitory information of the dormitory candidates, as well as the recommended sampling step and the number of dormitory recommendations, obtains the occupancy information of the occupants in the occupied dormitories that match the basic intended dormitory information, and determines a set of matching occupants that match the dormitory candidates based on the occupancy information and the personal information of the dormitory candidates. Subsequently, the occupants in the set of matching occupants and the dormitory candidates are sorted, and the sampled occupants are determined based on the sorting sequence number of the dormitory candidates and the recommended sampling step. Therefore, based on the dormitory attribute information of the sampled occupants, the proportion of various room types is determined, and based on the proportion of various room types, the room type information of the current available rooms, and the number of dormitory recommendations, the recommended dormitories to be recommended to the dormitory candidates are selected from the current available rooms, improving the recommendation accuracy, thereby improving the room selection efficiency and at the same time enhancing the fairness of dormitory allocation.

[0075] Next, the employee dormitory recommendation device provided by the present invention will be described. The employee dormitory recommendation device described below can be correspondingly referred to the employee dormitory recommendation method described above.

[0076] Based on any of the above embodiments, Figure 3 is a schematic structural diagram of the employee dormitory recommendation device provided by the present invention, as Figure 3 shown. The device includes:

[0077] An information determination unit 310, configured to determine the personal information and basic intended dormitory information of the dormitory candidates, as well as the recommended sampling step and the number of dormitory recommendations;

[0078] A similar person obtaining unit 320, configured to obtain the occupancy information of the occupants in the occupied dormitories that match the basic intended dormitory information, and determine a set of matching occupants that match the dormitory candidates based on the occupancy information and the personal information of the dormitory candidates;

[0079] A sampling unit 330, configured to sort the occupants in the set of matching occupants and the dormitory candidates, and determine the sampled occupants based on the sorting sequence number of the dormitory candidates and the recommended sampling step;

[0080] A recommended dormitory determination unit 340, configured to determine the proportion of various room types based on the dormitory attribute information of the sampled occupants, and select the recommended dormitories to be recommended to the dormitory candidates from the current available rooms based on the proportion of various room types, the room type information of the current available rooms, and the number of dormitory recommendations.

[0081] The device provided by the embodiment of the present invention determines the personal information of the candidates for dormitories and the basic intended dormitory information, as well as the recommended sampling step and the number of recommended dormitories. It obtains the information of the occupants of the occupied dormitories that match the basic intended dormitory information, and determines a set of matching occupants that match the candidates for dormitories based on the occupant information and the personal information of the candidates for dormitories. Subsequently, it sorts the occupants in the set of matching occupants and the candidates for dormitories, and determines the sampled occupants based on the sorting serial number of the candidates for dormitories and the recommended sampling step. Thus, based on the dormitory attribute information of the sampled occupants, it determines the proportion of various room types, and based on the proportion of various room types, the room type information of the current available vacant rooms, and the number of recommended dormitories, it selects the recommended dormitories to be recommended to the candidates for dormitories from the current available vacant rooms, improving the recommendation accuracy, thereby improving the room selection efficiency and at the same time enhancing the fairness of dormitory allocation.

[0082] Based on any of the above embodiments, the determining a set of matching occupants that match the candidates for dormitories based on the occupant information and the personal information of the candidates for dormitories includes:

[0083] Based on the occupant information of each occupant and the personal information of the candidates for dormitories, respectively construct the occupant vector of each occupant and the personal vector of the candidates for dormitories;

[0084] Based on the occupant vectors of each occupant, create a set of auxiliary person vectors; the set of auxiliary person vectors contains multiple auxiliary person vectors;

[0085] Fuse the occupant vectors of each occupant and the set of auxiliary person vectors to obtain a set of fused vectors, and perform density estimation on the set of fused vectors to obtain the density value of each vector in the set of fused vectors;

[0086] Obtain a preset number of vectors with density values higher than a preset threshold as the central vectors, and perform clustering based on the central vectors to obtain multiple clusters;

[0087] Based on the occupants corresponding to the occupant vectors included in the cluster where the personal vector of the candidates for dormitories is located, construct the set of matching occupants.

[0088] Based on any of the above embodiments, the creating a set of auxiliary person vectors based on the occupant vectors of each occupant includes:

[0089] The generator based on the generative adversarial network generates multiple virtual personnel vectors. The discriminator based on the generative adversarial network classifies the multiple virtual personnel vectors and the occupancy personnel vectors of each occupancy personnel to obtain the classification probabilities of each virtual personnel vector and occupancy personnel vector, and updates the generator and discriminator of the generative adversarial network based on the classification probabilities of each virtual personnel vector and occupancy personnel vector;

[0090] The generator of the updated generative adversarial network is used to generate multiple auxiliary personnel vectors, and the auxiliary personnel vector set is constructed based on the multiple auxiliary personnel vectors.

[0091] Based on any of the above embodiments, updating the generator and discriminator of the generative adversarial network based on the classification probabilities of each virtual personnel vector and occupancy personnel vector includes:

[0092] Based on the classification probabilities of each virtual personnel vector and the difference between the classification probabilities of each occupancy personnel vector and 1, update the model parameters of the discriminator;

[0093] Based on the difference between the classification probability of each virtual personnel vector and 1, update the model parameters of the generator.

[0094] Based on any of the above embodiments, obtaining vectors with density values higher than a preset threshold as the central vectors includes:

[0095] Determine the virtual personnel vectors with density values higher than the preset threshold as the reference personnel vectors;

[0096] Determine the occupancy personnel vectors with density values higher than the preset threshold as the candidate personnel vectors;

[0097] Calculate the distances between each candidate personnel vector and each reference personnel vector, and select a preset number of candidate personnel vectors with the closest distances as the central vectors.

[0098] Based on any of the above embodiments, sorting the occupancy personnel in the matching occupancy personnel set and the dormitory candidate personnel, and determining the sampled occupancy personnel based on the sorting serial numbers of the dormitory candidate personnel and the recommended sampling step length, includes:

[0099] Based on the rank, service time, total working time, children's education level, and border support priority information of the occupancy personnel in the matching occupancy personnel set and the dormitory candidate personnel, sort the occupancy personnel in the matching occupancy personnel set and the dormitory candidate personnel;

[0100] Based on the ranking sequence numbers of the dormitory candidates and the recommended sampling step, the occupants whose ranking sequence numbers before and after the dormitory candidates and whose difference with the ranking sequence number of the dormitory candidates is less than the recommended sampling step are determined as sampling occupants.

[0101] Figure 4 is a schematic diagram of the structure of the electronic device provided by the present invention, such as Figure 4 As shown, the electronic device may include: a processor (processor) 410, a memory (memory) 420, a communication interface (Communications Interface) 430 and a communication bus 440, wherein the processor 410, the memory 420, and the communication interface 430 communicate with each other through the communication bus 440. The processor 410 can call the logic instructions in the memory 420 to execute the staff dormitory recommendation method, which includes: determining the personal information and basic intended dormitory information of dormitory candidates, as well as the recommended sampling step and the number of dormitory recommendations; obtaining the occupant information of the occupants who have moved into the dormitory and match the basic intended dormitory information, and determining a set of matching occupants that match the dormitory candidates based on the occupant information and the personal information of the dormitory candidates; sorting the occupants in the set of matching occupants and the dormitory candidates, and determining the sampled occupants based on the sorting sequence number of the dormitory candidates and the recommended sampling step; determining the proportion of each type of room based on the dormitory attribute information of the sampled occupants, and selecting a recommended dormitory from the current vacant rooms to recommend to the dormitory candidates based on the proportion of each type of room, the room type information of the current vacant rooms, and the recommended number of dormitories.

[0102] In addition, the logic instructions in the above-mentioned memory 420 can be implemented in the form of a software functional unit and can be stored in a computer-readable storage medium when it is sold or used as an independent product. Based on such an understanding, the technical solution of the present invention, in essence, 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, and the computer software product is stored in a storage medium, including several instructions for 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: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc. Various media that can store program codes.

[0103] On the other hand, the present invention also provides a computer program product, which includes a computer program stored on a non-transitory computer-readable storage medium. The computer program includes program instructions. When the program instructions are executed by a computer, the computer can execute the employee dormitory recommendation method provided by each of the above methods. The method includes: determining the personal information and basic intended dormitory information of the dormitory candidates, as well as the recommendation sampling step and the number of dormitory recommendations; obtaining the occupancy information of the occupants of the occupied dormitories that match the basic intended dormitory information, and based on the occupancy information and the personal information of the dormitory candidates, determining a set of matching occupants that match the dormitory candidates; sorting the occupants in the set of matching occupants and the dormitory candidates, and based on the sorting serial number of the dormitory candidates and the recommendation sampling step, determining the sampled occupants; based on the dormitory attribute information of the sampled occupants, determining the proportion of each type of room, and based on the proportion of each type of room, the room type information of the current available rooms and the number of dormitory recommendations, selecting the recommended dormitories to be recommended to the dormitory candidates from the current available rooms.

[0104] In another aspect, the present invention also provides a non-transitory computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, it is implemented to execute the employee dormitory recommendation method provided by each of the above. The method includes: determining the personal information and basic intended dormitory information of the dormitory candidates, as well as the recommendation sampling step and the number of dormitory recommendations; obtaining the occupancy information of the occupants of the occupied dormitories that match the basic intended dormitory information, and based on the occupancy information and the personal information of the dormitory candidates, determining a set of matching occupants that match the dormitory candidates; sorting the occupants in the set of matching occupants and the dormitory candidates, and based on the sorting serial number of the dormitory candidates and the recommendation sampling step, determining the sampled occupants; based on the dormitory attribute information of the sampled occupants, determining the proportion of each type of room, and based on the proportion of each type of room, the room type information of the current available rooms and the number of dormitory recommendations, selecting the recommended dormitories to be recommended to the dormitory candidates from the current available rooms.

[0105] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place, or may be distributed to multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment. Those of ordinary skill in the art can understand and implement it without creative efforts.

[0106] 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, and of course, it can also be implemented by hardware. Based on such an understanding, the essence of the above technical solution, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0107] 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 them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements for some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for recommending staff dormitories, characterized in that: include: Determine the personal information and basic dormitory information of the candidates, as well as the recommended sampling step length and number of recommended dormitories; Acquire the occupant information of the occupants who have checked into the dormitory and match the basic intended dormitory information, and determine a set of matching occupants who match the dormitory candidate based on the occupant information and the personal information of the dormitory candidate; Sorting the occupants in the set of matched occupants and the dormitory candidates, and determining the sampled occupants based on the sorting sequence numbers of the dormitory candidates and the recommended sampling step length; Based on the dormitory attribute information of the sampled occupants, the proportion of each room type is determined, and based on the proportion of each room type, the room type information of the current vacant rooms and the recommended number of dormitories, a recommended dormitory is selected from the current vacant rooms to be recommended to the dormitory candidate; The step of determining a set of matching occupants matching the dormitory candidate based on the occupant information and the personal information of the dormitory candidate includes: Based on the occupant information of each occupant and the personal information of the dormitory candidate, constructing an occupant vector of each occupant and a personal vector of the dormitory candidate respectively; Creating an auxiliary staff vector set based on the occupant vector of each occupant; the auxiliary staff vector set includes a plurality of auxiliary staff vectors; Fusing the occupant vectors of each occupant and the auxiliary personnel vector set to obtain a fused vector set, and performing density estimation on the fused vector set to obtain a density value of each vector in the fused vector set; Obtaining a preset number of vectors whose density values ​​are higher than a preset threshold as central vectors, and performing clustering based on the central vectors to obtain multiple clusters; Constructing the set of matched occupants based on the occupants corresponding to the occupant vectors contained in the cluster where the personal vectors of the dormitory candidates are located; The step of creating a set of auxiliary personnel vectors based on the occupant vectors of each occupant includes: Generate multiple virtual personnel vectors based on the generator of the generative adversarial network, classify the multiple virtual personnel vectors and the occupant vectors of each occupant based on the discriminator of the generative adversarial network, obtain the classification probability of each virtual personnel vector and the occupant vector, and update the generator and discriminator of the generative adversarial network based on the classification probability of each virtual personnel vector and the occupant vector; A plurality of auxiliary staff vectors are generated using the updated generator of the generative adversarial network, and the auxiliary staff vector set is constructed based on the plurality of auxiliary staff vectors.

2. The staff dormitory recommendation method according to claim 1, characterized in that: The updating of the generator and the discriminator of the generative adversarial network based on the classification probabilities of each virtual person vector and the occupant person vector includes: Based on the classification probability of each virtual person vector and the difference between the classification probability of each occupant vector and 1, updating the model parameters of the discriminator; Based on the difference between the classification probability of each virtual person vector and 1, the model parameters of the generator are updated.

3. The staff dormitory recommendation method according to claim 1, characterized in that: The step of obtaining a preset number of vectors having density values ​​higher than a preset threshold as the center vector includes: Determine a virtual personnel vector having a density value higher than the preset threshold as a reference personnel vector; Determine a resident person vector whose density value is higher than the preset threshold as a candidate person vector; The distance between each candidate person vector and each reference person vector is calculated, and a preset number of candidate person vectors with the closest distance are selected as the center vector.

4. The staff dormitory recommendation method according to any one of claims 1 to 3, characterized in that: The step of sorting the occupants in the set of matched occupants and the dormitory candidates, and determining the sampled occupants based on the sorting sequence numbers of the dormitory candidates and the recommended sampling step length, includes: Sorting the occupants in the set of matched occupants and the dormitory candidates based on their job levels, tenure, total working time, children's education level, and support priority information; Based on the ranking sequence numbers of the dormitory candidates and the recommended sampling step, the occupants whose ranking sequence numbers before and after the dormitory candidates and whose difference with the ranking sequence number of the dormitory candidates is less than the recommended sampling step are determined as sampling occupants.

5. A staff dormitory recommendation device, characterized in that: include: An information determination unit, used to determine the personal information and basic intended dormitory information of the dormitory candidate, as well as the recommended sampling step length and the number of recommended dormitories; a similar person acquisition unit, configured to acquire the occupant information of the occupants who have moved into the dormitory and match the basic intended dormitory information, and determine a set of matching occupants who match the dormitory candidate based on the occupant information and the personal information of the dormitory candidate; A sampling unit, used for sorting the occupants in the set of matched occupants and the dormitory candidates, and determining the sampled occupants based on the sorting sequence numbers of the dormitory candidates and the recommended sampling step length; A recommended dormitory determination unit is used to determine the proportion of each room type based on the dormitory attribute information of the sampled occupants, and select a recommended dormitory from the current vacant rooms to be recommended to the dormitory candidate based on the proportion of each room type, the room type information of the current vacant rooms and the recommended number of dormitories; The step of determining a set of matching occupants matching the dormitory candidate based on the occupant information and the personal information of the dormitory candidate includes: Based on the occupant information of each occupant and the personal information of the dormitory candidate, constructing an occupant vector of each occupant and a personal vector of the dormitory candidate respectively; Creating an auxiliary staff vector set based on the occupant vector of each occupant; the auxiliary staff vector set includes a plurality of auxiliary staff vectors; Fusing the occupant vectors of each occupant and the auxiliary personnel vector set to obtain a fused vector set, and performing density estimation on the fused vector set to obtain a density value of each vector in the fused vector set; Obtaining a preset number of vectors whose density values ​​are higher than a preset threshold as central vectors, and performing clustering based on the central vectors to obtain multiple clusters; Constructing the set of matched occupants based on the occupants corresponding to the occupant vectors contained in the cluster where the personal vectors of the dormitory candidates are located; The step of creating a set of auxiliary personnel vectors based on the occupant vectors of each occupant includes: Generate multiple virtual personnel vectors based on the generator of the generative adversarial network, classify the multiple virtual personnel vectors and the occupant vectors of each occupant based on the discriminator of the generative adversarial network, obtain the classification probability of each virtual personnel vector and the occupant vector, and update the generator and discriminator of the generative adversarial network based on the classification probability of each virtual personnel vector and the occupant vector; A plurality of auxiliary staff vectors are generated using the updated generator of the generative adversarial network, and the auxiliary staff vector set is constructed based on the plurality of auxiliary staff vectors.

6. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the program, the staff dormitory recommendation method as described in any one of claims 1 to 4 is implemented.

7. 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 staff dormitory recommendation method as described in any one of claims 1 to 4 is implemented.

Images