Functional information classification method and device, electronic equipment and storage medium

Abstract

The invention discloses a function information classification method and device, electronic equipment and a storage medium. The method comprises the steps of obtaining a first heterogeneous graph corresponding to job application behavior information and a second heterogeneous graph corresponding to recruitment behavior information; determining vector information of each node in the first heterogeneous graph and vector information of each node in the second heterogeneous graph; obtaining function information to be classified, and determining position nodes corresponding to the function information; for each piece of function information, obtaining N target position nodes and M target job seeker nodes according to nodes respectively included in the first heterogeneous graph and the second heterogeneous graph; determining an aggregation vector of each piece of function information according to the vector information of the N target position nodes corresponding to each piece of function information and the vector information of the M target job seeker nodes; and performing clustering processing on the plurality of aggregation vectors to generate a target number of target clusters. According to the embodiment of the invention, the flexibility and accuracy of the function information classification process can be improved.

Description

Technical Field

[0001] The present application relates to the field of data processing technology, and in particular to a classification method, device, electronic device and storage medium for functional information. Background Art

[0002] With the popularization of the Internet and the increasing maturity of data processing technology, more and more convenience has been brought to users' lives and work. For example, in the recruitment industry, recruiters can post recruitment information on recruitment software and find the job seekers they need in the recruitment software, and job seekers can post resumes on the recruitment software and find suitable positions in the recruitment software.

[0003] In related technologies, in order to better provide business services, recruitment software needs to classify the job information corresponding to the positions posted by the recruiter, so as to improve the job seekers' and recruiters' satisfaction with the recruitment software. At present, the job information is often classified based on semantic analysis, lexical analysis and other methods. Therefore, the current classification processing method is too dependent on manual classification of job categories. Due to the large amount of recruitment information, relying on manual participation in the classification of job categories not only has problems such as duplication and low differentiation, but also has low classification efficiency. Summary of the invention

[0004] In view of this, the embodiments of the present application provide a method, device, electronic device and storage medium for classifying functional information, which can improve the flexibility and accuracy of the functional information classification process.

[0005] The present application embodiment provides a method for classifying functional information, including:

[0006] Obtaining a first heterogeneous graph corresponding to the job-seeking behavior information within a preset time period and a second heterogeneous graph corresponding to the recruitment behavior information within a preset time period, wherein the first heterogeneous graph includes first relationship information between job seeker nodes and position nodes, and the second heterogeneous graph includes second relationship information between job seeker nodes and position nodes;

[0007] Determine the vector information of each node in the first heterogeneous graph and the vector information of each node in the second heterogeneous graph; obtain the functional information to be classified and determine the position nodes corresponding to the functional information; for each functional information, obtain N target position nodes and M target job seeker nodes according to the nodes respectively included in the first heterogeneous graph and the second heterogeneous graph, wherein the N target position nodes include position nodes corresponding to the position of the functional information, and the M target job seeker nodes include job seeker nodes corresponding to job seekers who have interacted with the position corresponding to the functional information; determine the aggregation vector of each functional information according to the vector information of the N target position nodes and the vector information of the M target job seeker nodes corresponding to each functional information; perform clustering processing on multiple aggregation vectors to generate a target number of target clustering clusters, wherein the target number is determined according to the classification fineness required by the business application, and different target clustering clusters correspond to different functional categories.

[0008] Optionally, according to the method of the embodiment of the present application, the job-seeking behavior information includes job-viewing behavior and resume-submitting behavior; obtaining a first heterogeneous graph corresponding to the job-seeking behavior information within a preset time period includes: determining the first relationship information between the job seeker node and the job node based on the job-viewing behavior and resume-submitting behavior; generating a first heterogeneous graph based on the first relationship information between the job seeker node and the job node.

[0009] Optionally, according to the method of the embodiment of the present application, the recruitment behavior information includes resume viewing behavior and recruitment communication behavior; obtaining the second heterogeneous graph corresponding to the recruitment behavior information within a preset time period includes: determining the second relationship information between the job seeker node and the position node based on the resume viewing behavior and the recruitment communication behavior; generating a second heterogeneous graph based on the second relationship information between the job seeker node and the position node.

[0010] Optionally, according to the method of an embodiment of the present application, the vector information of each node in the first isomerous graph and the vector information of each node in the second isomerous graph are determined, including: generating a training sample set according to the first isomerous graph and the second isomerous graph; training a pre-constructed vector generation model according to the training sample set, and obtaining a target vector generation model after meeting preset training conditions; and determining the vector information of each node in the first isomerous graph and the vector information of each node in the second isomerous graph through the target vector generation model.

[0011] Optionally, according to the method of an embodiment of the present application, a training sample set is generated according to the first heterogeneous graph and the second heterogeneous graph, including: walking the first heterogeneous graph according to the first meta-path, and walking the second heterogeneous graph according to the second meta-path to obtain a walking sequence set, the walking sequence set including multiple walking sequences corresponding to each meta-path; generating a training sample set according to the order of nodes in each walking sequence in the walking sequence set and K nodes adjacent to the nodes, where K is a positive integer.

[0012] Optionally, according to the method of the embodiment of the present application, obtaining the functional information to be classified and processed includes: determining a first node set in a first heterogeneous graph, wherein the first node set includes a first position node and a first job seeker node, the node degree of the first position node is greater than a first threshold, and the node degree of the first job seeker node is greater than a second threshold; determining a second node set in a second heterogeneous graph, wherein the second node set includes a second position node and a second job seeker node, the node degree of the second position node is greater than a second threshold, and the node degree of the second job seeker node is greater than the second threshold; determining the functional information corresponding to each position node in the first node set and the second node set as the functional information to be classified and processed.

[0013] Optionally, according to the method of the embodiment of the present application, for each functional information, N target position nodes and M target job seeker nodes are obtained according to the nodes respectively included in the first heterogeneous graph and the second heterogeneous graph, including: for each functional information, N target position nodes are selected from the position nodes included in the first node set and the second node set, and M target job seeker nodes are selected from the job seeker nodes included in the first node set and the second node set.

[0014] Optionally, according to the method of an embodiment of the present application, multiple aggregation vectors are clustered to generate a target number of target clusters, including: randomly selecting a target number of first aggregation vectors from multiple aggregation vectors as the initial cluster centers of the target number of clusters; according to a preset clustering algorithm, iteratively updating each cluster and the cluster center of the cluster until a preset stop iteration condition is reached to obtain the target cluster; wherein the second aggregation vector is an aggregation vector other than the first aggregation vector among the multiple aggregation vectors.

[0015] The embodiment of the present application provides a classification device for job information, including: an acquisition module, used to acquire a first heterogeneous graph corresponding to job-seeking behavior information within a preset time period and a second heterogeneous graph corresponding to recruitment behavior information within a preset time period, wherein the first heterogeneous graph includes first relationship information between job seeker nodes and position nodes, and the second heterogeneous graph includes second relationship information between job seeker nodes and position nodes; a processing module, used to determine vector information of each node in the first heterogeneous graph and vector information of each node in the second heterogeneous graph; the processing module is also used to acquire job information to be classified and determine the position node corresponding to the job information; the processing module is also used to classify each job information according to the first heterogeneous graph and the second heterogeneous graph. The composition includes nodes respectively, obtaining N target position nodes and M target job seeker nodes, wherein the N target position nodes include position nodes corresponding to the position of the functional information, and the M target job seeker nodes include job seeker nodes corresponding to the job seekers who have interacted with the position corresponding to the functional information; the processing module is also used to determine the aggregation vector of each functional information according to the vector information of the N target position nodes corresponding to each functional information and the vector information of the M target job seeker nodes; the processing module is also used to cluster multiple aggregation vectors to generate a target number of target clustering clusters, wherein the target number is determined according to the classification fineness required by the business application, and different target clustering clusters correspond to different functional categories.

[0016] An embodiment of the present application provides an electronic device, which includes a processor and a memory storing computer program instructions; when the processor executes the computer program instructions, the steps of the above method are implemented.

[0017] An embodiment of the present application provides a computer-readable storage medium, on which computer program instructions are stored. When the computer program instructions are executed by a processor, the steps of the above method are implemented.

[0018] An embodiment of the present application provides a computer program product, which includes computer program instructions. When the computer program instructions are executed by a processor, the steps of the above method are implemented.

[0019] By adopting the embodiment of the present application, a heterogeneous graph is established for the job-seeking behavior information and the released recruitment information acquired within a preset time period, so as to conveniently express the relationship between the job seeker and each position from the perspective of the job seeker, and express the relationship between each position and the job seeker from the perspective of the recruiter. Next, for each job position node to be classified and processed, in combination with its corresponding job position node, N target job position nodes corresponding to the job information and M target job seeker nodes corresponding to the job seeker who has interacted with the position corresponding to the job information are found to generate an aggregation vector for each job information. Finally, according to the classification precision required by the business application, the target number of clustering clusters can be flexibly determined, and multiple aggregation vectors can be automatically processed into a target number of different categories of clustering clusters, thereby improving the efficiency of classifying and processing the job information, and more accurately classifying and merging the job information, and increasing the distinction between different categories of functions. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] In order to more clearly illustrate the technical solution of the embodiments of the present application, the following briefly introduces the drawings in the embodiments of the present application.

[0021] Figure 1 It is a schematic diagram of the system architecture of an embodiment of the present application.

[0022] Figure 2 It is a flowchart of the classification method of functional information in an embodiment of the present application.

[0023] Figure 3 It is a schematic diagram of a first isomorphic graph provided in an embodiment of the present application.

[0024] Figure 4 It is a schematic diagram of a second isomorphic graph provided in an embodiment of the present application.

[0025] Figure 5 This is a schematic diagram of a walking isomorphic graph provided in an embodiment of the present application.

[0026] Figure 6 It is a schematic diagram of another first isomorphic graph provided in an embodiment of the present application.

[0027] Figure 7 It is a schematic diagram of another second isomorphic graph provided in an embodiment of the present application.

[0028] Figure 8 It is a flowchart of another method for classifying functional information provided in an embodiment of the present application.

[0029] Fig. 9 It is a structural block diagram of a device for classifying functional information according to an embodiment of the present application.

[0030] Fig.10It is a schematic diagram of an electronic device used to implement the method for classifying functional information in an embodiment of the present application. DETAILED DESCRIPTION

[0031] The principles and spirit of the present application will be described below with reference to several exemplary embodiments. It should be understood that the purpose of providing these embodiments is to make the principles and spirit of the present application clearer and more thorough, so that those skilled in the art can better understand and implement the principles and spirit of the present application. The exemplary embodiments provided herein are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments herein, all other embodiments obtained by ordinary technicians of the art without creative work are within the scope of protection of this application.

[0032] In this document, terms such as first, second, third, etc. are only used to distinguish one entity (or operation) from another entity (or operation), but not to require or imply any order or relationship between these entities (or operations).

[0033] The embodiments of the present application relate to terminal devices and / or servers. Those skilled in the art will appreciate that the embodiments of the present application may be implemented as a system, apparatus, device, method, computer-readable storage medium, or computer program product. Therefore, the present disclosure may be specifically implemented in at least one of the following forms: complete hardware, complete software, or a combination of hardware and software. According to the embodiments of the present application, the present application requests protection for a method, apparatus, electronic device, and storage medium for classifying functional information. Figure 1 A schematic diagram of a system architecture of an embodiment of the present application is shown. Figure 1 As shown, the system includes a terminal device 102 and a server 104. The terminal device 102 may include at least one of the following: a smart phone, a tablet computer, a laptop computer, a desktop computer, a smart TV, various wearable devices, an augmented reality AR device, a virtual reality VR device, etc. The terminal device 102 may use recruitment software. For example, a client may be installed on the terminal device 102. The client may be a client that specifically performs a specific function (such as an application app), or a client that is embedded with a variety of application applets (with different functions), or a client that logs in through a browser. Job seekers may operate on the terminal device 102. For example, job seekers may open the terminal device 102, use the recruitment software, view positions in the recruitment software, post resumes, and perform other operations. Optionally, the recruiter may also operate on the terminal device, for example, post recruitment information in the recruitment software, view job seekers' resumes, communicate with job seekers, and perform other operations.

[0034] The server 104 can provide storage and retrieval functions, for example, it can store the resume of job seekers, recruitment information of recruiters, etc. After the terminal device 102 receives the instruction input by the job seeker or the recruiter, it sends the request information including the instruction to the server 104. After receiving the request information, the server 104 performs corresponding processing and then returns the processing result information to the terminal device 102. The user instruction is completed through a series of data processing and information interaction. Based on these instructions, the server 104 can obtain the job search behavior information and the convenience of the recruitment software, and at the same time, it is also helpful for the recruiter to find suitable employees.

[0035] Based on this, server 104 can regularly obtain job search behavior information, recruitment behavior information and recruitment information, and analyze the relationship between the information, so as to clearly distinguish the functional information included in the recruitment information, accurately classify and merge the functional information, and improve the flexibility and accuracy of the functional information classification process.

[0036] It should be noted that the acquisition, storage, use, and processing of data in the embodiments of the present application are in compliance with the relevant provisions of national laws and regulations.

[0037] The following is an introduction to the classification method of the functional information provided by the embodiment of the present application in conjunction with the accompanying drawings. Figure 2 A flow chart of a method for classifying functional information according to an embodiment of the present application is shown. The method may include the following steps 201 to 206 .

[0038] Step 201, obtaining a first heterogeneous graph corresponding to job-seeking behavior information within a preset time period and a second heterogeneous graph corresponding to recruitment behavior information within a preset time period, wherein the first heterogeneous graph includes first relationship information between job seeker nodes and position nodes, and the second heterogeneous graph includes second relationship information between job seeker nodes and position nodes.

[0039] Step 202: Determine vector information of each node in the first heterogeneous graph and vector information of each node in the second heterogeneous graph.

[0040] Step 203, obtaining the functional information to be classified and processed, and determining the position node corresponding to the functional information.

[0041] Step 204, for each functional information, obtain N target position nodes and M target job seeker nodes based on the nodes respectively included in the first heterogeneous graph and the second heterogeneous graph, wherein the N target position nodes include position nodes corresponding to the functional information, and the M target job seeker nodes include job seeker nodes corresponding to job seekers who have interacted with the position corresponding to the functional information.

[0042] Step 205 , determining an aggregate vector of each functional information according to the vector information of the N target position nodes and the vector information of the M target job seeker nodes corresponding to each functional information.

[0043] Step 206 , clustering the multiple aggregation vectors to generate a target number of target clusters, wherein the target number is determined according to the classification precision required by the business application, and different target clusters correspond to different functional categories.

[0044] The above steps are described in detail below in conjunction with specific embodiments, as shown below.

[0045] In an embodiment of the present application, the executor of the classification of functional information may be an electronic device with data processing capabilities, such as a server. Optionally, the server may be a physical server, a virtual server, or a cloud server. No specific restrictions are imposed on the form of the server.

[0046] In step 201, the job-seeking behavior information and the recruitment behavior information are obtained, and the job-seeking behavior information and the recruitment behavior information are analyzed to classify the function information. Optionally, the preset time period can be used as the time interval for each execution of the classification method of the function information, and the preset time period can be set according to the application requirements. The preset time period can also be a time period set according to actual application requirements, and the preset time period is not specifically limited here.

[0047] Exemplarily, job-seeking behavior information and recruitment information can be obtained from the service device of the recruitment software respectively. For example, the job seeker's operation behaviors such as viewing positions and submitting resumes in the recruitment software can trigger the service device to provide the job seeker with feedback information corresponding to the operation behavior; the recruiter's operations such as viewing resumes, communicating with job seekers and inviting interviews in the recruitment software can trigger the service device to provide the recruiter with feedback information corresponding to the operation behavior.

[0048] A heterogeneous graph includes multiple nodes and multiple edges, wherein nodes with associated relationships can be connected through the edges.

[0049] In one example, a job seeker can submit a resume to a desired position during the job search process. Thus, the job seeker nodes and positions with job search relationships can be found through job search behavior information, where the job search relationship is also the first relationship information. Next, connections between different job seeker nodes and position nodes are established based on the job search behavior, thereby establishing a first heterogeneous graph. Each resume can correspond to a job seeker node, and each position can correspond to a position node.

[0050] In another example, when a recruiter is looking for a suitable candidate for a position, he can view the resumes of one or more job seekers and can also actively communicate with job seekers. The behavior of actively communicating with job seekers includes but is not limited to asking job seekers for basic information, answering job seekers' questions, and sending interview invitations, job invitations, etc. Therefore, through the recruitment behavior information, nodes and positions with recruitment relationships can be found, where the recruitment relationship is also the second relationship information. Next, connections between different position nodes and job seeker nodes are established according to the recruitment behavior, thereby establishing a second heterogeneous graph.

[0051] By establishing a heterogeneous graph, we can easily express the relationship between job seekers and various positions from the perspective of job seekers, and express the relationship between various positions and job seekers from the perspective of recruiters, thus providing a good data foundation for improving the accuracy of classification results.

[0052] Next, the above step 202 is involved, and the vector information of each node in the first heterogeneous graph and the second heterogeneous graph is determined respectively, that is, the vector information of each job seeker node, the vector information of the position node and the vector information of the function node in the heterogeneous graph is determined. Optionally, the vector generation model trained in advance can output the vector information of each node in each heterogeneous graph.

[0053] Next, the above step 203 is involved, determining the job information to be classified and determining the position node corresponding to the job information. Optionally, the job information to be classified can be determined based on the published recruitment information or based on the existing recruitment behavior information.

[0054] Exemplarily, in recruitment information, positions and functional information are interrelated and one-to-one corresponding, but the functional information corresponding to different positions can be different or the same. Specifically, a position can represent a job role or title. For example, within the same company, there can be one or more of the same positions; functional information can include specific work content, such as the skills or abilities required for the position, where different positions can correspond to the same functional information. For example, when personnel changes, the position may change, but the function may remain unchanged. For another example, different companies offer two different positions, but the functional information corresponding to the two positions may be the same. Thus, the functional information can be found and the position node corresponding to the functional information can be determined.

[0055] In another exemplary embodiment, a recruiter is looking for a suitable candidate for a published position. Thus, the corresponding function information and position node can be obtained for the published position, and the function information can be used as an object to be classified.

[0056] Next, regarding the above steps 204 and 205, for each functional information, N target position nodes and M target job seeker nodes may be obtained from the nodes respectively included in the first heterogeneous graph and the second heterogeneous graph.

[0057] Exemplarily, for each functional information, the position nodes corresponding to the functional information can be selected from the position nodes included in the first heterogeneous graph and the second heterogeneous graph as a selection range, so as to obtain N position nodes from the range as target position nodes. Wherein, N is a preset positive integer, for example, it can be 200, 300, etc. The value of N can be set according to actual application requirements, and the value of N is not specifically limited here.

[0058] In another example, for each job information, the job seeker nodes corresponding to the job seekers who have interacted with the position corresponding to the job information can be selected from the job seeker nodes included in the first heterogeneous graph and the second heterogeneous graph as another selection range, so as to obtain M job seeker nodes from the range as target job seeker nodes. Wherein, M is a preset positive integer, for example, it can be 200, 300, etc. The value of M can be set according to actual application requirements, and the value of M is not specifically limited here.

[0059] Next, the aggregation vector of each functional information can be determined based on the vector information of the N target position nodes and the vector information of the M target job seeker nodes corresponding to each functional information. Exemplarily, the aggregation vector refers to the fusion of multiple vectors into one vector. Optionally, the aggregation method can be selected according to the specific data distribution, and the aggregation method includes but is not limited to calculating the average value, calculating the weighted average value, calculating the maximum or minimum value, calculating the geometric mean, etc. By using the aggregation vector, the complexity of the subsequent functional clustering processing can be reduced.

[0060] Taking the calculation of the average value as an example, for each functional information, the average value of the vector information of N target position nodes and the vector information of M target job seeker nodes can be taken to obtain the aggregate vector of the functional information.

[0061] After obtaining the aggregation vector of each functional information, that is, the aggregation vectors corresponding to the multiple functional information to be classified can be obtained. Next, in step 206, clustering processing can be performed on the multiple aggregation vectors to generate a target number of target clusters.

[0062] Exemplarily, each target cluster includes multiple aggregation vectors, each aggregation vector corresponds to functional information one by one, therefore, each target cluster can correspond to a category, and the functional information corresponding to the multiple aggregation vectors included in the target cluster all belong to this category.

[0063] Based on this, different target clusters can be obtained, and different target clusters correspond to different functional categories, which is conducive to improving the distinction between functional information of different categories.

[0064] Optionally, the target number is determined according to the classification precision of the business application requirements. For example, the higher the classification precision, the larger the value corresponding to the target number can be; the lower the classification precision, the smaller the value corresponding to the target number can be. Therefore, according to the classification method of functional information provided in the embodiment of the present application, different classification requirements can be flexibly responded to.

[0065] According to the embodiments of the present application, the target number of clusters can be flexibly determined according to the classification precision required by the business application, and multiple aggregation vectors can be automatically processed into a target number of clusters of different categories, thereby improving the efficiency of classifying and processing functional information, and more accurately classifying and merging functional information, and improving the distinction between different categories of functional information.

[0066] In some embodiments of the present application, job-seeking behavior information includes job-viewing behavior and resume-submitting behavior; based on this, a first heterogeneous graph corresponding to the job-seeking behavior information within a preset time period is obtained, including: determining the first relationship information between the job seeker node and the job node based on the job-viewing behavior and resume-submitting behavior; generating a first heterogeneous graph based on the first relationship information between the job seeker node and the job node.

[0067] For example, based on the job viewing behavior and resume submission behavior, it can be determined which resumes interact with which jobs, which not only reflects the job seeker's submission history, but also reflects the job seeker's interest preference when selecting jobs. For example, the job seeker's interest in the job he submitted his resume to is higher than the interest in the job he only viewed.

[0068] As a specific example, Figure 3 This is a schematic diagram of a first heterogeneous graph provided by an embodiment of the present application. The job seeker's behavior of viewing positions and submitting resumes is used to associate the job seeker with the position (J). The job seeker (U) and the position (J) are used as nodes, and viewing and submitting are used as edges to construct a first heterogeneous graph of job seekers and positions. Figure 3 As shown, U1, U2, U3, U4, and U5 represent different job seeker nodes, and J1, J2, J3, and J4 represent different positions.

[0069] In the embodiment of the present application, the advantage of such processing is that after establishing connections between nodes with associated relationships, the relationship between job seekers and positions can be conveniently expressed through a heterogeneous graph, so that the functional information corresponding to the positions can be classified and merged more accurately.

[0070] In some embodiments of the present application, recruitment behavior information includes resume viewing behavior and recruitment communication behavior; based on this, a second heterogeneous graph corresponding to the recruitment behavior information within a preset time period is obtained, including: determining the second relationship information between the job seeker node and the position node based on the resume viewing behavior and the recruitment communication behavior; generating a second heterogeneous graph based on the second relationship information between the job seeker node and the position node.

[0071] For example, based on the resume viewing behavior and recruitment communication behavior, it can be determined which resumes the recruiter has viewed for a position and which job seekers he has interacted with. This not only reflects the recruiter's recruitment history, but also reflects the recruiter's interest preferences when selecting job seekers. For example, the recruiter's interest in the job seekers he communicates with is higher than the job seekers whose resumes he has only viewed. As a specific example, Figure 4 1 is a schematic diagram of a second heterogeneous graph provided by an embodiment of the present application, which shows the job seeker's behavior of viewing positions and submitting resumes. The job seeker is associated with the position (J), and the position (J) and the job seeker (U) are used as nodes, and viewing and submitting are used as edges to construct the first heterogeneous graph of the job seeker and the position. Figure 4 As shown, U1, U2, U3, and U4 represent different job seeker nodes, and J1, J2, J3, J4, and J5 represent different positions.

[0072] In the embodiment of the present application, after the nodes with associated relationships are connected, the relationship between job seekers and positions is conveniently expressed through a heterogeneous graph, so that the functional information corresponding to the positions can be classified and merged more accurately.

[0073] In some embodiments of the present application, the steps of determining the vector information of each node in the first heterogeneous graph and the vector information of each node in the second heterogeneous graph may specifically refer to the following steps 301 to 303.

[0074] Step 301: Generate a training sample set according to the first heterogeneous graph and the second heterogeneous graph.

[0075] Step 302, training the pre-built vector generation model according to the training sample set, and obtaining the target vector generation model after reaching the preset training conditions.

[0076] Step 303: Determine vector information of each node in the first heterogeneous graph and vector information of each node in the second heterogeneous graph through a target vector generation model.

[0077] In the embodiment of the present application, the training samples in the training sample set are determined by walking sequences obtained by walking the first heterogeneous graph and the second heterogeneous graph respectively according to different meta paths.

[0078] Exemplarily, a meta-path refers to a specific path pattern between different types of nodes and edges in a heterogeneous graph. For example, the path from the job seeker node to the position node and then to the job seeker node can be used as a meta-path. By setting different meta-paths, multiple random walks can be performed in the heterogeneous graph to generate a walk sequence that reflects the complex relationship between nodes. Each walk sequence actually represents a node relationship under a certain association pattern and can be used as training data for subsequent models. For example, the meta-paths pre-constructed for the first heterogeneous graph are, for example: (1) job seeker → position → job seeker; (2) position → job seeker → position; the meta-paths pre-constructed for the second heterogeneous graph are, for example: (1) job seeker → position → job seeker; (2) position → job seeker → position. Corresponding to each meta-path, multiple walk sequences are found through walks.

[0079] In some embodiments, the pre-built vector generation model includes but is not limited to supporting multiple graph neural network model frameworks (Deep Graph Library, DGL), PyG, NeuGraph and other model frameworks for construction. Optionally, in the embodiment of the present application, a pre-built vector generation model based on the DGL model framework is used to perform image extraction (Graph Embedding) on ​​the relationship between nodes in the heterogeneous graph.

[0080] For example, the pre-built vector generation model may be a model developed based on a Skip-gram model, wherein the Skip-gram model is a prediction model in natural language processing and may be used to learn continuous vector representations of words.

[0081] The preset training condition may be that the function value of the loss function approaches a preset value, or that the training cycle (Epoch) of the model reaches a preset number of times, etc. The specific preset training condition may be set according to actual training needs.

[0082] After the model training is completed, the target vector generation model is obtained. The target vector generation model can output the vector information of each node in the first isomorphic graph and the vector information of each node in the second isomorphic graph, that is, the vector information of the job seeker node and the vector information of the function node in the first isomorphic graph, and the vector information of the job seeker node and the vector information of the function node in the second isomorphic graph.

[0083] According to an embodiment of the present application, by using a heterogeneous graph generated based on job-seeking behavior information and recruitment information to generate training samples and a training vector generation model, it is beneficial to deeply explore the potential relationship between job seekers' job-seeking intentions and functional information, and thus help improve the accuracy of classification processing of functional information corresponding to positions. In addition, by setting different meta-paths, it is convenient to flexibly adjust the roaming strategy according to the needs of actual recruitment scenarios to meet personalized matching needs. The node vectors generated based on the heterogeneous graph and the target vector generation model can easily and efficiently mine useful information from job-seeking behavior information and recruitment information, and convert it into a reliable vectorized representation to support the subsequent steps of classifying and processing functional information.

[0084] In some embodiments of the present application, a training sample set is generated according to the first heterogeneous graph and the second heterogeneous graph, including: walking the first heterogeneous graph according to the first meta-path, and walking the second heterogeneous graph according to the second meta-path to obtain a walking sequence set, the walking sequence set including multiple walking sequences corresponding to each meta-path; generating a training sample set according to the order of nodes in each walking sequence in the walking sequence set and K nodes adjacent to the node, where K is a positive integer.

[0085] For example, Figure 5 is a schematic diagram of a walking heterogeneous graph provided in an embodiment of the present application, combined with Figure 5 As shown. Combined Figure 5 As shown in (a), based on the meta-path (1) job seeker → position → job seeker, a walk in the first heterogeneous graph can obtain the walk sequence: U1-J1-U3. Figure 5 As shown in (b), based on the meta-path (3) position → job seeker → position, a walk can be made in the second heterogeneous graph to obtain a walk sequence: J1-U1-J3. Corresponding to each meta-path set for the first heterogeneous graph, after walking the heterogeneous graph, multiple walk sequences can be obtained. Here, each walk sequence corresponding to each meta-path is not listed one by one.

[0086] In some embodiments, the connection lines between the nodes in the first heterogeneous graph and the second heterogeneous graph are undirected edges. The walking direction can be randomly selected at any node.

[0087] Optionally, in the process of walking the heterogeneous graph, walking parameters can be pre-set, such as the walking depth and the number of times each node in the walking process is used as the starting point of the walking. Continuing with the first heterogeneous graph and meta-path (1) as an example, if the walking depth is 1, the obtained walking sequence can be U1-J1-U3. If the walking depth is 2, after obtaining U1-J1-U3, continue to use U3 as the starting point and continue to walk in the first heterogeneous graph according to meta-path (1). The final walking sequence is, for example, U1-J1-U3-J4-U4. In the embodiment of the present application, optionally, the walking depth and the number of times each node in the walking process is used as the starting point of the walking can be set according to the application requirements, and no specific restrictions are made here.

[0088] Optionally, in the embodiment of the present application, the weight between the job seeker node and the position node can also be set. Figure 6 As shown in the figure, V represents the job seeker's viewing behavior, A represents the job seeker's application behavior, and the number after the colon represents the weight of the edge. Figure 7 As shown in the figure, H represents the recruiter's viewing behavior, and C represents the recruiter's communication delivery behavior. Among them, the greater the weight, the greater the probability of being selected in the wandering process node. The advantage of this is that it is conducive to fully mining the intention information of job seekers' behavior and the intention information of recruiters' behavior, and converting them into vectorized representation to support the subsequent steps of classifying and processing job information.

[0089] In each walk sequence, assuming that a node represents a job seeker node, the K neighboring nodes of the node may include the job nodes that have interacted with it. By using the sequential relationship between these nodes and their neighboring nodes as training samples, the association between resumes and positions can be further analyzed, and training data can be provided for subsequent vector generation models. The value of K can be adjusted according to actual needs. Generally, the larger K is, the more neighboring nodes around the node can participate in the training, thereby generating richer training samples.

[0090] According to the embodiment of the present application, the meta-path-based random walk can capture the multi-dimensional associations between different types of nodes in the heterogeneous graph, so that the training sample set is not limited to the specific relationship of a certain type of node, but can reflect information in multiple dimensions such as resumes and positions. Based on this, it is helpful to mine useful information from on-the-job behavior information and recruitment information, and convert it into a vectorized representation to support the subsequent steps of classifying and processing the functional information.

[0091] In some embodiments of the present application, optionally, the training sample set includes multiple positive training samples and multiple negative training samples. Optionally, a training sample set is generated according to the order of nodes in each walk sequence and the K nodes in the neighborhood of each node, and the positive training samples and negative training samples can be obtained by referring to the following steps. Specifically, for each walk sequence, the current node is combined with each of the K nodes in the neighborhood of the current node to obtain a positive training sample, the current node is any node in the walk sequence, and the current node in each positive training sample is an input sample node, and the other node is an output positive sample node.

[0092] Exemplarily, the walk sequence set includes multiple walk sequences corresponding to each meta-path, and for each walk sequence, a current node is combined with each of K nodes in the neighborhood of the current node to obtain a positive training sample.

[0093] For each input sample node in the positive training sample, L nodes are randomly obtained from multiple walk sequences included in the walk sequence set as output negative sample nodes, and the L output negative sample nodes are respectively combined with the input sample nodes to obtain L negative training samples corresponding to each positive training sample.

[0094] As a specific example, for example, taking the walking sequence U1-J1-U3-J4-U4 as an example, U1, J1, U3, J4 and U4 will be used as the center point, and combined with each of the K nodes in its neighborhood. Taking K=2 as an example, U1 is the center point, and the neighborhood nodes of U1 are J1 and U3 respectively, thus, two positive samples can be obtained: U1-J1, U1-U3. Among them, R2 is the input sample node in the positive training sample.

[0095] J1 is the center point, and its neighboring nodes are U1, U3 and J4, thus, three positive samples can be obtained: J1-U1, J1-U3, and J1-J4. Among them, J1 is the input sample node in the positive training sample.

[0096] U3 is the center point, and the neighboring nodes of U3 are U1, J1, J4 and U4 respectively. Thus, four positive samples can be obtained: U3-U1, U3-J1, U3-J4, U3-U4. Among them, U3 is the input sample node in the positive training sample.

[0097] Based on the same positive sample generation logic, three positive samples can be obtained when J4 is the center point, and two positive samples can be obtained when U4 is the center point, which are not listed here one by one.

[0098] Usually, there are many nodes in heterogeneous graphs. In order to speed up model training, negative samples can be generated by negative sampling. Taking meta-path (1) as an example, U1~U5 are used as starting points respectively. According to the preset depth, the walking paths are obtained, and a total of 5 paths are obtained. Among all the nodes included in the 5 paths, nodes are selected to generate negative samples. For example, for the positive sample U1-J1, 5 nodes are randomly selected from all U and J as negative samples according to probability: U1-F5, U1-U2, U1-U3, U1-J5, and U1-J4.

[0099] It can be seen that in the process of selecting negative samples, the frequency of node appearance is used as the probability of being selected as a negative sample, that is, the higher the frequency of the node, the greater the probability of being selected as a negative sample. The advantage of this setting is to strengthen the training of high-frequency nodes and avoid the model being biased by high-frequency nodes. It is understandable that the number of negative samples corresponding to each positive sample is set according to actual application requirements and performance, and no specific restrictions are made here. According to an embodiment of the present application, a node vector generation method based on a heterogeneous graph and a target vector generation model can effectively mine useful information from job search behavior information and recruitment information, and convert it into a vectorized representation to support subsequent classification processes.

[0100] In some embodiments of the present application, with respect to obtaining the functional information to be classified and processed, the following steps may be specifically included: determining a first node set in a first heterogeneous graph, wherein the first node set includes a first position node and a first job seeker node, the node degree of the first position node is greater than a first threshold, and the node degree of the first job seeker node is greater than a second threshold; determining a second node set in a second heterogeneous graph, wherein the second node set includes a second position node and a second job seeker node, the node degree of the second position node is greater than a second threshold, and the node degree of the second job seeker node is greater than the second threshold; determining the functional information corresponding to each position node in the first node set and the second node set as the functional information to be classified and processed.

[0101] Exemplarily, the degree of a node refers to the number of edges directly connected to a node. In the embodiment of the present application, the edges in the heterogeneous graph are undirected edges, and the degree of a node refers to the total number of edges connected to the node. For example, if a node is connected to three other nodes, then its degree is 3.

[0102] Optionally, in the embodiment of the present application, the first threshold and the second threshold may be the same or different, and may be set specifically according to the application scenario, and no specific limitation is made here. For example, if the first threshold and the second threshold are both 9, then in all vector information, only nodes with node degrees greater than 10 are retained, and nodes with node degrees less than or equal to 9 are screened out, thereby adding the remaining nodes in the first heterogeneous graph to the first node set. Based on the same processing method, the nodes in the second heterogeneous graph that need to be added to the second node set can be found, which will not be repeated here.

[0103] After the first node set and the second node set are determined, the function information corresponding to the position nodes included in the two node sets can be directly determined as the function information to be classified and processed.

[0104] In an optional example, for each functional information, the nodes within the selection range of the target position node include nodes within the first node set and nodes within the second node set. Based on this, the number of target position nodes selected from the first node set and the second node set can be preset, for example, N / 2 nodes are selected for each sub-range; the number of target position nodes selected from the first node set and the second node set can also be determined according to the ratio of the number of position nodes included in the first node set and the second node set respectively. For example, the more the number of position nodes, the greater the number of target position nodes selected. No specific limitation is made here.

[0105] In some embodiments of the present application, for each functional information, N target position nodes and M target job seeker nodes are obtained according to the nodes respectively included in the first heterogeneous graph and the second heterogeneous graph, including: for each functional information, N target position nodes are selected from the position nodes included in the first node set and the second node set, and M target job seeker nodes are selected from the job seeker nodes included in the first node set and the second node set.

[0106] For each job function information, the nodes within the selection range of the target job seeker node include the nodes in the first node set and the nodes in the second node set. Based on this, the number of target job seeker nodes selected from the first node set and the second node set can be preset, for example, N / 2 nodes are selected for each sub-range; the number of target job seeker nodes selected from the first node set and the second node set can also be determined according to the ratio of the number of job seeker nodes included in the first node set and the second node set respectively. For example, the more job seeker nodes there are, the larger the number of target job seeker nodes selected. No specific limitation is made here.

[0107] Based on the processing method of the embodiment of the present application, the functional information to be classified is determined by combining nodes with larger node degrees, wherein, since the larger the node degree, the more associations there are between the node and other nodes (for example, job seeker nodes, position nodes), the relationship between the nodes is more reference-based, which is conducive to mining more useful information therefrom. After converting the nodes with larger node degrees into vectorized representations, it is helpful to improve the accuracy of subsequent classification processing of the functional information.

[0108] In the embodiment of the present application, the aggregate vector of each functional information can be determined according to the vector information of N target position nodes and the vector information of M target job seeker nodes corresponding to each functional information. Exemplarily, the aggregate vector refers to the fusion of multiple vectors into one vector. Compared with the method of directly using the functional information as a node and using the vector generation model to output the vector, after aggregating multiple types of nodes of positions and job seekers, the aggregate vector is used to represent the functional information, which can summarize the features included in the functional information in multiple dimensions and more accurately.

[0109] In some embodiments of the present application, clustering is performed on multiple aggregation vectors to generate a target number of target clusters, and the following steps may be referred to: randomly selecting a target number of first aggregation vectors from multiple aggregation vectors as initial cluster centers of the target number of clusters; iteratively updating each cluster and the cluster center of the cluster according to a preset clustering algorithm until a preset stop iteration condition is reached to obtain the target cluster; wherein the second aggregation vector is an aggregation vector other than the first aggregation vector from the multiple aggregation vectors.

[0110] Optionally, a preset clustering algorithm may be used to perform clustering processing on multiple aggregation vectors to generate a target number of target clusters. Exemplarily, the preset clustering algorithm may include but is not limited to any one of a variety of clustering methods such as Mini Batch K-means clustering, hierarchical clustering, and density clustering.

[0111] Exemplarily, taking Mini Batch K-means clustering as an example, generating a target number of clusters is used as an example to introduce the clustering process.

[0112] The target number is determined according to the classification precision required by the business application. Taking the target number Z as an example, first, Z clustering vectors are randomly selected from multiple clustering vectors, and the Z first clustering vectors are respectively determined as Z clustering centers. At this time, the Z clustering centers are the initial clustering centers.

[0113] For the convenience of description, Z randomly selected aggregation vectors are described as first aggregation vectors, and the remaining aggregation vectors in the multiple aggregation vectors are described as second aggregation vectors.

[0114] Next, in each iterative calculation, Y second clustering vectors are randomly selected from all the second clustering vectors, and for each of the Y second clustering vectors, the cluster center associated with each second clustering vector is determined, wherein the cluster center associated with each second clustering vector is the cluster center closest to the second clustering vector among the Z clustering centers. Afterwards, the position of the cluster center associated with the second clustering vector is updated according to the second clustering vector, thereby completing one iterative calculation.

[0115] It can be understood that in the entire clustering process, each second aggregation vector is selected only once, and each iterative calculation only selects the second aggregation vector that has not been calculated. After reaching the preset stop iteration condition, Z target cluster centers are obtained.

[0116] Exemplarily, the preset condition for stopping the iteration may be that all second aggregation vectors participate in the iterative calculation, or the number of iterative calculations reaches a preset number of iterations, or the position change of each cluster center is less than a threshold, etc.

[0117] Based on this, the functional information can be accurately classified and merged to obtain Z different target clusters with high discrimination, and different target clusters correspond to different functional categories, so as to flexibly respond to different classification needs.

[0118] In an embodiment of the present application, the target number Z can be determined based on business application requirements. For example, the target number can be determined with reference to the following business application requirements: (1) the proportion of functions in the cluster delivered within the cluster; (2) the proportion of correlations between functions within the cluster; and (3) the balance of traffic distribution between clusters.

[0119] For example, business application requirement (1) can be calculated from the perspective of job seekers. Each job seeker node and position node has corresponding functional information. Specifically, when each position is posted, the recruitment information may include not only position information but also functional information. By analyzing the recruitment information, the correlation between the position and functional information can be found. For example, different positions may correspond to different functional information or the same functional information.

[0120] During the job search process, job seekers can submit resumes to the positions they like, which can be simplified into a submission relationship between functions. Based on this, the calculation method can be: for the function information associated with the current job seeker and the function information associated with the position to which the job seeker submits his resume, if they are still in the current cluster, then count them, and finally use the count / all submission numbers, which is the function information in the cluster that is submitted within the cluster.

[0121] For example, the business application requirement (2) may be calculated from the perspective of a job seeker.

[0122] The calculation formula can be shown as formula (1).

[0123]

[0124] Among them, C is all functions in the current cluster, and F is the function.

[0125] For the functions in each cluster, assuming that the cluster includes n functions, there are n*n combinations. Within a certain time range (such as 30 days), there are n*n combinations in each cluster, and there is a delivery relationship between functions. The number of combinations that can exist / (n*n) is the proportion of correlation between functions in the cluster.

[0126] In another example, business application requirements can also be calculated from the perspective of recruiters. For example, the ratio of job seekers who chat with positions in the cluster to those who have the same functions in the cluster is calculated. Each job seeker node and position node has associated function information. A recruiter who has a favorite resume for a certain position can actively communicate with the job seeker corresponding to the favorite resume. Therefore, based on the function information associated with the above-mentioned position, and the function information associated with the job seeker corresponding to the resume, the communication behavior of the recruiter can be simplified to the recruitment relationship between functions. Based on this, the calculation method can be: for the function information associated with the current position, and the function information associated with the job seeker communicated based on the position is still in the current cluster, then count, and finally use the count / all submissions, which is the ratio of job seekers who chat with positions in the cluster to those who have the same functions in the cluster.

[0127] In addition, from the perspective of recruiters, the proportion of chat correlations between functions within a cluster can be calculated. The calculation method is the same as the calculation logic of the proportion of correlations between functions within a cluster from the perspective of job seekers, and will not be repeated here.

[0128] In some embodiments, the business application requirement (3) can be to focus on recommending a certain function to job seekers and increase the exposure of the function, and the size of the function can be determined from the perspective of job seekers; when it is necessary to increase the amount of communication between the recruiter and the job seeker, the size of the function can be determined from the perspective of the recruiter.

[0129] According to the embodiments of the present application, it is possible to determine the number of positions that need to be covered within each category of functions based on an accurate estimate of the required traffic, flexibly determine the target number of clusters, and automatically process multiple aggregation vectors into a target number of clusters of different categories with high discrimination, which not only improves the efficiency of classifying functions, but also allows for more accurate classification and merging of functions.

[0130] The above describes the implementation methods and advantages of the embodiments of the present application through multiple embodiments. In order to more clearly introduce the embodiments of the present application, the specific processing process of the embodiments of the present application is described in detail below with reference to specific examples.

[0131] Figure 8 is a flow chart of another method for classifying functional information provided by an embodiment of the present application, combined with Figure 8 As shown, the solution may include the following steps 801 to 810.

[0132] Step 801, obtaining job search behavior data and recruitment behavior information, and constructing a first heterogeneous graph and a second heterogeneous graph.

[0133] Step 802, obtaining the meta-path.

[0134] Step 803, traversing the job seeker nodes and position nodes in the first heterogeneous graph and the second heterogeneous graph respectively, and walking in the first heterogeneous graph and the second heterogeneous graph respectively according to a preset meta-path, and generating a plurality of walking sequences corresponding to each meta-path.

[0135] Step 804: Sample each walking sequence to generate multiple positive samples and multiple negative samples associated with each positive sample.

[0136] Step 805: construct a vector generation model.

[0137] Step 806: training the vector generation model according to the positive training samples and the negative training samples.

[0138] Step 807: After training for multiple cycles, if the Loss of the vector generation model no longer decreases, stop training to obtain the target vector generation model.

[0139] After stopping the training, the model parameters of the current model are saved to obtain the target vector generation model.

[0140] Step 808, obtaining the functional information to be classified and processed.

[0141] Optionally, the job function information to be classified and processed can be determined based on published recruitment information or based on existing recruitment behavior information.

[0142] Step 809: for each of the job function information, obtain N target position nodes and M target job seeker nodes according to the nodes respectively included in the first heterogeneous graph and the second heterogeneous graph.

[0143] Step 810: Determine an aggregate vector of each of the function information according to the vector information of the N target position nodes and the vector information of the M target job seeker nodes corresponding to each of the function information.

[0144] Step 811: cluster the plurality of aggregation vectors to generate a target number of target clusters.

[0145] Optionally, the number of targets is determined according to the classification precision of business application requirements, and different target clusters correspond to different functional categories.

[0146] Exemplarily, among the multiple aggregation vectors, a target number of first aggregation vectors are randomly selected as the initial cluster centers of the target number of cluster clusters; according to a preset clustering algorithm, each of the cluster clusters and the cluster centers of the cluster clusters are iteratively updated until a preset stop iteration condition is reached to obtain the target cluster cluster; wherein the second aggregation vector is an aggregation vector among the multiple aggregation vectors except the first aggregation vector.

[0147] Corresponding to the method embodiment of the present application, the present application also provides a classification device for functional information, such as Fig. 9 As shown, the classification device 900 of the job information includes:

[0148] The acquisition module 901 is used to obtain a first heterogeneous graph corresponding to the job-seeking behavior information within a preset time period and a second heterogeneous graph corresponding to the recruitment behavior information within a preset time period, wherein the first heterogeneous graph includes first relationship information between job seeker nodes and position nodes, and the second heterogeneous graph includes second relationship information between job seeker nodes and position nodes.

[0149] The processing module 902 is used to determine the vector information of each node in the first heterogeneous graph and the vector information of each node in the second heterogeneous graph.

[0150] The processing module 902 is also used to obtain the functional information to be classified and processed, and determine the position node corresponding to the functional information.

[0151] Processing module 902 is also used to obtain N target position nodes and M target job seeker nodes for each functional information based on the nodes respectively included in the first heterogeneous graph and the second heterogeneous graph, wherein the N target position nodes include position nodes corresponding to the positions corresponding to the functional information, and the M target job seeker nodes include job seeker nodes corresponding to job seekers who have interacted with the positions corresponding to the functional information.

[0152] The processing module 902 is further used to determine the aggregation vector of each functional information according to the vector information of the N target position nodes and the vector information of the M target job seeker nodes corresponding to each functional information.

[0153] The processing module 902 is further used to perform clustering processing on the multiple aggregation vectors to generate a target number of target clusters, wherein the target number is determined according to the classification precision required by the business application, and different target clusters correspond to different functional categories.

[0154] In some embodiments of the present application, the job-seeking behavior information may include a job-viewing behavior and a resume-submitting behavior. The processing module 902 is further configured to determine the first relationship information between the job seeker node and the job node according to the job-viewing behavior and the resume-submitting behavior.

[0155] The processing module 902 is further configured to generate a first heterogeneous graph according to the first relationship information between the job seeker nodes and the position nodes.

[0156] In some embodiments of the present application, the recruitment behavior information may include resume viewing behavior and recruitment communication behavior. The processing module 902 is further configured to determine the second relationship information between the job seeker node and the position node according to the resume viewing behavior and the recruitment communication behavior.

[0157] The processing module 902 is further configured to generate a second heterogeneous graph according to the second relationship information between the job seeker nodes and the position nodes.

[0158] In some embodiments of the present application, optionally, the processing module 902 is further configured to generate a training sample set according to the first heterogeneous graph and the second heterogeneous graph.

[0159] The processing module 902 is further used to train the pre-constructed vector generation model according to the training sample set, and obtain the target vector generation model after the preset training conditions are met.

[0160] The processing module 902 is further used to determine the vector information of each node in the first heterogeneous graph and the vector information of each node in the second heterogeneous graph through the target vector generation model.

[0161] In some embodiments of the present application, optionally, the processing module 902 is further used to wander the first heterogeneous graph according to the first meta-path, and wander the second heterogeneous graph according to the second meta-path, to obtain a wandering sequence set, wherein the wandering sequence set includes multiple wandering sequences corresponding to each meta-path.

[0162] The processing module 902 is further configured to generate a training sample set according to the order of nodes in each walking sequence in the walking sequence set and K nodes adjacent to the node, where K is a positive integer.

[0163] In some embodiments of the present application, optionally, the processing module 902 is also used to determine a first node set in the first heterogeneous graph, wherein the first node set includes a first position node and a first job seeker node, the node degree of the first position node is greater than a first threshold, and the node degree of the first job seeker node is greater than a second threshold.

[0164] The processing module 902 is further used to determine a second node set in the second heterogeneous graph, wherein the second node set includes a second position node and a second job seeker node, the node degree of the second position node is greater than a second threshold, and the node degree of the second job seeker node is greater than the second threshold.

[0165] The processing module 902 is further used to determine the function information corresponding to each position node in the first node set and the second node set as the function information to be classified and processed.

[0166] In some embodiments of the present application, optionally, the processing module 902 is also used to select N target position nodes from the position nodes included in the first node set and the second node set, and select M target job seeker nodes from the job seeker nodes included in the first node set and the second node set for each functional information.

[0167] In some embodiments of the present application, optionally, the processing module 902 is further configured to randomly select a target number of first aggregation vectors from a plurality of aggregation vectors as initial cluster centers of a target number of cluster clusters.

[0168] The processing module 902 is further used to iteratively update each cluster and the cluster center of each cluster according to a preset clustering algorithm until a preset stop iteration condition is reached to obtain a target cluster.

[0169] The second aggregation vector is an aggregation vector other than the first aggregation vector among the multiple aggregation vectors.

[0170] It can be understood that the classification device of functional information in the embodiment of the present application can correspond to the execution subject of the classification method of functional information provided in the embodiment of the present application. The specific details of the operation and / or function of each module / unit of the classification device of functional information can be found in the description of the corresponding parts of the classification method of functional information provided in the above-mentioned embodiment of the present application. For the sake of brevity, they will not be repeated here.

[0171] The electronic device in the embodiment of the present application may be a user terminal device, a server, other computing devices, or a cloud server. Fig.10 A schematic diagram of the hardware structure of an electronic device according to an embodiment of the present application is shown. The electronic device may include a processor 1001 and a memory 1002 storing computer program instructions. When the processor 1001 executes the computer program instructions, the process or function of any of the above-mentioned embodiments of the method is implemented.

[0172] Specifically, the processor 1001 may include a central processing unit (CPU), or an application-specific integrated circuit (ASIC), or may be configured to implement one or more integrated circuits of the embodiments of the present application. The memory 1002 may include a large-capacity memory for data or instructions. For example, the memory 1002 may be at least one of the following: a hard disk drive (HDD), a read-only memory (ROM), a random access memory (RAM), a floppy disk drive, a flash memory, an optical disk, a magneto-optical disk, a tape, a universal serial bus (USB) drive, or other physical / tangible memory storage devices. For another example, the memory 1002 may include a removable or non-removable (or fixed) medium. For another example, the memory 1002 may be inside or outside the integrated gateway disaster recovery device. The memory 1002 may be a non-volatile solid-state memory. In other words, the memory 1002 generally includes a tangible (non-transitory) computer-readable storage medium (such as a memory device) encoded with computer-executable instructions, and when the software is executed (such as executed by one or more processors), the operations described in the method of the embodiments of the present application may be performed. The processor 1001 implements the process or function of any method in the above embodiments by reading and executing the computer program instructions stored in the memory 1002.

[0173] In one example, Fig.10 The electronic device shown may also include a communication interface 1003 and a bus 1010. Among them, the processor 1001, the memory 1002, and the communication interface 1003 are connected through the bus 1010 and complete the communication between each other. The communication interface 1003 is mainly used to realize the communication between each module, device, unit and / or device in the embodiment of the present application. The bus 1010 includes hardware, software or both, and can couple the components of the online data traffic billing device to each other. For example, the bus may include at least one of the following: an accelerated graphics port (AGP) or other graphics bus, an enhanced industrial standard architecture (EISA) bus, a front-end bus (FSB), a hypertransport (HT) interconnect, an industrial standard architecture (ISA) bus, an infinite bandwidth interconnect, a low pin count (LPC) bus, a memory bus, a micro channel architecture (MCA) bus, a peripheral component interconnect (PCI) bus, a PCI-Express (PCI-X) bus, a serial advanced technology attachment (SATA) bus, a video electronics standard association local (VLB) bus or other suitable buses. The bus 1010 may include one or more buses. Although the embodiments of the present application describe or illustrate a specific bus, the embodiments of the present application may consider any suitable bus or interconnection method.

[0174] In combination with the method in the above embodiments, an embodiment of the present application also provides a computer-readable storage medium, on which computer program instructions are stored. When the computer program instructions are executed by a processor, the process or function of any method in the above embodiments is implemented.

[0175] In addition, an embodiment of the present application further provides a computer program product, on which computer program instructions are stored. When the computer program instructions are executed by a processor, the process or function of any one of the methods in the above embodiments is implemented.

[0176] The above exemplarily describes the flowcharts and / or block diagrams of the methods, devices, systems and computer program products of the embodiments of the present application, and describes the relevant various aspects. It should be understood that each box or combination thereof in the flowchart and / or block diagram can be implemented by computer program instructions, or by dedicated hardware that performs a specified function or action, or by a combination of dedicated hardware and computer instructions. For example, these computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer or other programmable data processing device to form a machine that enables these instructions executed by such a processor to enable the implementation of the functions / actions specified in each box or combination thereof in the flowchart and / or block diagram. Such a processor can be a general-purpose processor, a special-purpose processor, a special application processor or a field programmable logic circuit.

[0177] The functional blocks shown in the structured block diagram of the embodiment of the present application can be implemented as hardware, software, firmware or a combination thereof. When implemented in hardware, it can be, for example, an electronic circuit, an application-specific integrated circuit (ASIC), appropriate firmware, a plug-in, a function card, etc.; when implemented in software, it is a program or code segment used to perform the required task. The program or code segment can be stored in a memory, or transmitted on a transmission medium or a communication link via a data signal carried in a carrier. The code segment can be downloaded via a computer network such as the Internet, an intranet, etc.

[0178] It should be noted that the present application is not limited to the specific configurations and processes described above or shown in the figures. The above is only a specific implementation mode of the present application. Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working process of the described system, device, module or unit can refer to the corresponding process in the method embodiment without further description. It should be understood that the scope of protection of the present application is not limited to this. Any technician familiar with the technical field can think of various equivalent modifications or substitutions within the technical scope disclosed in the present application, and these modifications or substitutions should be included in the scope of protection of the present application.

Claims

1. A method for classifying job information, characterized in that: include: Obtaining a first heterogeneous graph corresponding to the job-seeking behavior information within a preset time period and a second heterogeneous graph corresponding to the recruitment behavior information within the preset time period, wherein the first heterogeneous graph includes first relationship information between job seeker nodes and position nodes, and the second heterogeneous graph includes second relationship information between job seeker nodes and position nodes; Determine vector information of each node in the first heterogeneous graph and vector information of each node in the second heterogeneous graph; Obtaining the functional information to be classified and processed, and determining the position node corresponding to the functional information; For each of the functional information, according to the nodes respectively included in the first heterogeneous graph and the second heterogeneous graph, N target position nodes and M target job seeker nodes are obtained, wherein the N target position nodes include position nodes corresponding to the functional information, and the M target job seeker nodes include job seeker nodes corresponding to job seekers who have interacted with the position corresponding to the functional information; Determine an aggregate vector of each of the functional information according to the vector information of the N target position nodes and the vector information of the M target job seeker nodes corresponding to each of the functional information; Clustering is performed on the plurality of aggregation vectors to generate a target number of target clusters, wherein the target number is determined according to the classification precision of the business application requirements, and different target clusters correspond to different functional categories.

2. The method according to claim 1, characterized in that: The job-seeking behavior information includes job-viewing behavior and resume-submitting behavior; The step of obtaining a first heterogeneous graph corresponding to the job-seeking behavior information within a preset time period includes: Determine first relationship information between the job seeker node and the job position node according to the job viewing behavior and the resume submitting behavior; The first heterogeneous graph is generated according to the first relationship information between the job seeker nodes and the position nodes.

3. The method according to claim 1, characterized in that The recruitment behavior information includes resume viewing behavior and recruitment communication behavior; Obtaining the second heterogeneous graph corresponding to the recruitment behavior information within the preset time period includes: Determining second relationship information between the job seeker node and the position node according to the resume viewing behavior and the recruitment communication behavior; The second heterogeneous graph is generated according to the second relationship information between the job seeker nodes and the position nodes.

4. The method according to claim 1, characterized in that The determining of the vector information of each node in the first heterogeneous graph and the vector information of each node in the second heterogeneous graph includes: Generate a training sample set according to the first heterogeneous graph and the second heterogeneous graph; Training the pre-constructed vector generation model according to the training sample set, and obtaining the target vector generation model after reaching a preset training condition; The target vector generation model is used to determine vector information of each node in the first heterogeneous graph and vector information of each node in the second heterogeneous graph.

5. The method according to claim 4, characterized in that The step of generating a training sample set according to the first heterogeneous graph and the second heterogeneous graph includes: According to the first meta-path, wander the first heterogeneous graph, and according to the second meta-path, wander the second heterogeneous graph to obtain a wandering sequence set, wherein the wandering sequence set includes a plurality of wandering sequences corresponding to each meta-path; The training sample set is generated according to the order of nodes in each walking sequence in the walking sequence set and K nodes adjacent to the node, where K is a positive integer.

6. The method according to claim 1, characterized in that The obtaining of the functional information to be classified and processed includes: Determining a first node set in the first heterogeneous graph, wherein the first node set includes a first position node and a first job seeker node, a node degree of the first position node is greater than a first threshold, and a node degree of the first job seeker node is greater than a second threshold; Determine a second node set in the second heterogeneous graph, wherein the second node set includes a second position node and a second job seeker node, the node degree of the second position node is greater than a second threshold, and the node degree of the second job seeker node is greater than the second threshold; The function information corresponding to each position node in the first node set and the second node set is determined as the function information to be classified.

7. The method according to claim 6, characterized in that For each of the job function information, obtaining N target position nodes and M target job seeker nodes according to the nodes respectively included in the first heterogeneous graph and the second heterogeneous graph includes: For each of the functional information, N target position nodes are selected from the position nodes included in the first node set and the second node set, and M target job seeker nodes are selected from the job seeker nodes included in the first node set and the second node set.

8. The method according to claim 1, characterized in that The clustering process is performed on the plurality of the aggregation vectors to generate a target number of target clusters, including: Randomly selecting a target number of first aggregation vectors from the multiple aggregation vectors as initial cluster centers of the target number of cluster clusters; According to a preset clustering algorithm, each of the clusters and the cluster centers of the clusters are iteratively updated until a preset stop iteration condition is reached to obtain the target cluster; The second aggregation vector is an aggregation vector among the multiple aggregation vectors except the first aggregation vector.

9. A classification device for job information, characterized in that: include: An acquisition module, used to acquire a first heterogeneous graph corresponding to the job-seeking behavior information within a preset time period and a second heterogeneous graph corresponding to the recruitment behavior information within the preset time period, wherein the first heterogeneous graph includes first relationship information between job seeker nodes and position nodes, and the second heterogeneous graph includes second relationship information between job seeker nodes and position nodes; A processing module, used to determine vector information of each node in the first heterogeneous graph and vector information of each node in the second heterogeneous graph; The processing module is further used to obtain the functional information to be classified and processed, and determine the position node corresponding to the functional information; The processing module is further used to obtain, for each of the functional information, N target position nodes and M target job seeker nodes according to the nodes respectively included in the first heterogeneous graph and the second heterogeneous graph, wherein the N target position nodes include position nodes corresponding to the functional information, and the M target job seeker nodes include job seeker nodes corresponding to job seekers who have interacted with the position corresponding to the functional information; The processing module is further used to determine the aggregate vector of each of the functional information according to the vector information of the N target position nodes and the vector information of the M target job seeker nodes corresponding to each of the functional information; The processing module is further used to perform clustering processing on the multiple aggregation vectors to generate a target number of target clusters, wherein the target number is determined according to the classification precision required by the business application, and different target clusters correspond to different functional categories.

10. An electronic device, characterized in that: The electronic device comprises: a processor and a memory storing computer program instructions; when the electronic device executes the computer program instructions, the method according to any one of claims 1 to 8 is implemented.

11. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer program instructions, and when the computer program instructions are executed by a processor, the method according to any one of claims 1 to 8 is implemented.

12. A computer program product, characterized in that It comprises computer program instructions, which, when executed by a processor, implement the method according to any one of claims 1 to 8.

Images