Personnel examination arrangement method and system based on multi-dimensional data analysis

Abstract

The invention relates to the field of intelligent data analysis, in particular to a personnel examination arrangement method and system based on multi-dimensional data analysis. The method comprises the following steps: acquiring personal information of a target examinee object and personal information of other four examinee objects adjacent to the target examinee object; a data processing and semantic understanding algorithm based on artificial intelligence and a natural language processing technology is introduced at the rear end to extract semantic features of the personal information of the target examinee and the personal information of the other four examinees; and automatically learning and capturing an implicit association relationship and a semantic difference measurement coefficient between the personal semantic features of the target examinee object and the personal semantic features of the other four examinee objects, so as to intelligently evaluate the rationality of the arrangement of the examination seats and judge the reasonable grade of the arrangement of the examination seats. Therefore, a powerful guarantee can be provided for the fairness of the examination.

Description

Technical Field

[0001] The present application relates to the field of intelligent data analysis, and specifically to a personnel examination arrangement method and system based on multidimensional data analysis. Background Art

[0002] In the modern education system, cheating in exams has always been one of the main issues affecting the fairness of exams. Exam cheating not only undermines the fairness of exams, but also damages the rights and interests of honest candidates, and has a negative impact on the integrity system of society. With the continuous advancement of information technology, the means of cheating in exams have become more diverse, bringing new challenges to education administrators. Traditional exam arrangement methods are often based on simple random selection or fixed seating arrangements, which are difficult to effectively prevent mutual cheating among candidates.

[0003] Chinese patent CN111274290B discloses a system and method for arranging examinations based on multidimensional data analysis, which can arrange examination seats based on multidimensional data analysis technology by calculating the cluster center of examinee data, data affinity between each dimension label and mathematical models such as chaotic mapping. Although this method can reduce the possibility of cheating in examinations to a certain extent and improve the rationality of examination arrangements, it usually requires manual setting of analysis parameters and thresholds, and relies on pre-defined data dimensions, such as information such as graduating colleges, colleges currently studying, age, gender and place of residence. Such a method is difficult to utilize the semantic association relationship between different examinee data, and it is also difficult to identify potential cheating behavior hidden dangers, resulting in an increase in cheating risk. In addition, this model can only perform one-time data analysis, and it is difficult to make timely adjustments according to updated data or newly added data, thereby reducing the timeliness and effectiveness of examination seat arrangements. In other words, with the passage of time and the development of society, the data information and cheating behavior of examinees may change. The examination arrangement method based on multidimensional data analysis needs to manually update parameters and models regularly, which is less efficient and less reliable.

[0004] Therefore, an optimized personnel examination arrangement scheme based on multidimensional data analysis is desired. Summary of the invention

[0005] This application is made in view of the above problems. One purpose of this application is to provide a method and system for arranging personnel examinations based on multi-dimensional data analysis.

[0006] The embodiment of the present application provides a method for arranging personnel examinations based on multidimensional data analysis, which includes:

[0007] Obtain the personal information of the first candidate object;

[0008] Extracting personal information of second to fifth candidate objects adjacent to the first candidate object from the candidate information database;

[0009] Performing semantic encoding on the personal information of the first examinee object to obtain a semantic encoding feature vector of the personal information of the first examinee object;

[0010] Performing semantic encoding on the personal information of the second to fifth examinee objects to obtain semantic encoding feature vectors of the personal information of the second to fifth examinee objects;

[0011] Jointly encode the first candidate object personal information semantic encoding feature vector and each candidate object personal information semantic encoding feature vector in the second to fifth candidate object personal information semantic encoding feature vectors to obtain four adjacent candidate object personal information joint encoding vectors;

[0012] Calculate the personal information semantic difference measurement coefficient between any two adjacent candidate object personal information joint coding vectors in the four adjacent candidate object personal information joint coding vectors to obtain a candidate object personal information semantic difference topological distribution matrix composed of multiple personal information semantic difference measurement coefficients;

[0013] Based on the semantic difference topological distribution matrix of the candidate object's personal information, reasonable grade labels for the examination seat arrangement are determined.

[0014] For example, according to the personnel examination arrangement method based on multidimensional data analysis of an embodiment of the present application, the first candidate object personal information semantic coding feature vector is jointly encoded with each candidate object personal information semantic coding feature vector in the second to fifth candidate object personal information semantic coding feature vectors to obtain four adjacent candidate object personal information joint coding vectors, including:

[0015] Calculate the semantic association score vectors between the first candidate object personal information semantic encoding feature vector and each candidate object personal information semantic encoding feature vector in the second to fifth candidate object personal information semantic encoding feature vectors to obtain a set of candidate object personal information semantic association score vectors;

[0016] Determine a global representation vector of the semantic relevance of the candidate object's personal information based on the set of semantic relevance score vectors of the candidate object's personal information;

[0017] Based on the global representation vector of the semantic relevance of the personal information of the examinee object, calculating the association optimization factor of each of the semantic encoding feature vectors of the personal information of the examinee object in the second to fifth semantic encoding feature vectors of the personal information of the examinee object to obtain a sequence of semantic association optimization factors of the personal information of the examinee object;

[0018] Inputting the sequence of the semantic association optimization factors of the candidate object personal information into the Softmax activation function to obtain the sequence of the semantic association optimization weight factors of the candidate object personal information;

[0019] Taking each candidate object personal information semantic association optimization weight factor in the sequence of candidate object personal information semantic association optimization weight factors as a weight, each candidate object personal information semantic encoding feature vector in the second to fifth candidate object personal information semantic encoding feature vectors is weighted respectively to obtain the four adjacent candidate object personal information joint encoding vectors.

[0020] For example, according to the personnel examination arrangement method based on multidimensional data analysis of an embodiment of the present application, the semantic association score vectors between the first candidate object personal information semantic encoding feature vector and each candidate object personal information semantic encoding feature vector in the second to fifth candidate object personal information semantic encoding feature vectors are calculated to obtain a set of candidate object personal information semantic association score vectors, including:

[0021] Concatenate the first candidate object personal information semantic encoding feature vector with each candidate object personal information semantic encoding feature vector in the second to fifth candidate object personal information semantic encoding feature vectors to obtain a set of candidate object personal information semantic cascade feature vectors;

[0022] Each semantic cascade feature vector of candidate object personal information in the set of semantic cascade feature vectors of candidate object personal information is multiplied by the candidate object personal information weight matrix and then added positionally with the candidate object personal information bias vector to obtain the set of semantic association score vectors of candidate object personal information.

[0023] For example, according to the personnel examination arrangement method based on multidimensional data analysis of an embodiment of the present application, a global representation vector of the semantic relevance of candidate object personal information is determined based on a set of semantic association score vectors of candidate object personal information, including: calculating the mean vector of the set of semantic association score vectors of candidate object personal information as the global representation vector of the semantic relevance of candidate object personal information.

[0024] For example, according to the personnel examination arrangement method based on multidimensional data analysis of an embodiment of the present application, based on the global representation vector of the semantic relevance of the personal information of the examinee object, the association optimization factor of each of the semantic encoding feature vectors of the personal information of the examinee object in the second to fifth semantic encoding feature vectors of the personal information of the examinee object is calculated to obtain a sequence of semantic association optimization factors of the personal information of the examinee object, including:

[0025] Multiplying the global representation vector of the semantic relevance of the candidate object personal information by the global weight matrix of the candidate object personal information to obtain the global representation weight vector of the candidate object personal information;

[0026] Multiply each candidate object personal information semantic encoding feature vector in the second to fifth candidate object personal information semantic encoding feature vectors by the corresponding candidate object personal information local weight matrix to obtain a set of candidate object personal information local representation weight vectors;

[0027] Calculate each candidate object personal information local representation weight vector in the set of candidate object personal information local representation weight vectors, add them to the candidate object personal information global representation weight vector and the corresponding candidate object personal information bias vector by position to obtain a set of candidate object personal information semantic association vectors;

[0028] The set of semantic association vectors of the candidate object's personal information is input into a Sigmoid function to obtain a sequence of optimization factors of the semantic association of the candidate object's personal information.

[0029] For example, according to the personnel examination arrangement method based on multidimensional data analysis of an embodiment of the present application, the personal information semantic difference measurement coefficient between any two adjacent candidate object personal information joint coding vectors in the four adjacent candidate object personal information joint coding vectors is calculated to obtain a candidate object personal information semantic difference topological distribution matrix composed of multiple personal information semantic difference measurement coefficients, including:

[0030] Calculate the positional difference between any two adjacent candidate object personal information joint encoding vectors among the four adjacent candidate object personal information joint encoding vectors to obtain a sequence of adjacent candidate object personal information semantic difference feature vectors;

[0031] Calculate the covariance matrix between any two adjacent candidate object personal information joint encoding vectors in the four adjacent candidate object personal information joint encoding vectors, and calculate the inverse of each eigenvalue in the covariance matrix to obtain a sequence of adjacent candidate object personal information semantic anti-correlation covariance matrices;

[0032] Calculate the transposed vector of each of the semantic differential feature vectors of personal information of the adjacent examinee objects in the sequence of semantic differential feature vectors of personal information of the adjacent examinee objects to obtain a sequence of transposed representation vectors of semantic differential representation of personal information of the adjacent examinee objects;

[0033] Calculate the product of each group of corresponding neighboring candidate object personal information semantic difference transposed representation vectors, neighboring candidate object personal information semantic anti-correlation covariance matrix and neighboring candidate object personal information semantic difference eigenvectors in the sequence of neighboring candidate object personal information semantic difference transposed representation vectors, the sequence of neighboring candidate object personal information semantic anti-correlation covariance matrix and neighboring candidate object personal information semantic difference eigenvectors to obtain a sequence of neighboring candidate object personal information semantic difference correlation values;

[0034] Calculate the square root of each of the semantic difference correlation values ​​of the personal information of the adjacent examinee objects in the sequence of semantic difference correlation values ​​of the personal information of the adjacent examinee objects to obtain the plurality of semantic difference measurement coefficients of the personal information;

[0035] The multiple personal information semantic difference measurement coefficients are arranged in a matrix to obtain a topological distribution matrix of the personal information semantic differences of the examinee objects.

[0036] For example, according to the personnel examination arrangement method based on multidimensional data analysis of an embodiment of the present application, determining a reasonable grade label for the examination seat arrangement based on the semantic difference topological distribution matrix of the candidate object personal information includes: inputting the semantic difference topological distribution matrix of the candidate object personal information into a classifier-based examination arrangement intelligent evaluator to obtain an evaluation result, and the evaluation result is used to represent a reasonable grade label for the examination seat arrangement.

[0037] The embodiment of the present application also provides a personnel examination arrangement system based on multi-dimensional data analysis, which includes:

[0038] The first candidate personal information acquisition module is used to acquire the personal information of the first candidate object;

[0039] An adjacent examinee information extraction module, used to extract personal information of the second to fifth examinee objects adjacent to the first examinee object from the examinee information database;

[0040] A first semantic encoding module, configured to semantically encode the personal information of the first examinee object to obtain a semantic encoding feature vector of the personal information of the first examinee object;

[0041] A second semantic encoding module, used to semantically encode the personal information of the second to fifth examinee objects to obtain semantic encoding feature vectors of the personal information of the second to fifth examinee objects;

[0042] A joint encoding module, used for jointly encoding the first candidate object personal information semantic encoding feature vector with each candidate object personal information semantic encoding feature vector in the second to fifth candidate object personal information semantic encoding feature vectors to obtain four adjacent candidate object personal information joint encoding vectors;

[0043] A personal information semantic difference measurement coefficient calculation module is used to calculate the personal information semantic difference measurement coefficient between any two adjacent candidate object personal information joint coding vectors in the four adjacent candidate object personal information joint coding vectors to obtain a candidate object personal information semantic difference topological distribution matrix composed of multiple personal information semantic difference measurement coefficients;

[0044] The grade label determination module is used to determine the reasonable grade label of the examination seat arrangement based on the semantic difference topological distribution matrix of the candidate object's personal information.

[0045] For example, according to the personnel examination arrangement system based on multidimensional data analysis of an embodiment of the present application, the joint coding module includes:

[0046] A semantic association score vector calculation unit, used to calculate the semantic association score vectors between the first candidate object personal information semantic coding feature vector and each of the candidate object personal information semantic coding feature vectors from the second to fifth candidate object personal information semantic coding feature vectors to obtain a set of candidate object personal information semantic association score vectors;

[0047] A semantic relevance unit, used to determine a global representation vector of the semantic relevance of the candidate object's personal information based on a set of semantic relevance score vectors of the candidate object's personal information;

[0048] A correlation optimization factor calculation unit, used to calculate the correlation optimization factor of each candidate object personal information semantic encoding feature vector in the second to fifth candidate object personal information semantic encoding feature vectors based on the candidate object personal information semantic relevance global representation vector to obtain a sequence of candidate object personal information semantic correlation optimization factors;

[0049] An activation unit, used for inputting the sequence of the semantic association optimization factors of the candidate object personal information into a Softmax activation function to obtain a sequence of the semantic association optimization weight factors of the candidate object personal information;

[0050] A weighting unit is used to use each candidate object personal information semantic association optimization weight factor in the sequence of candidate object personal information semantic association optimization weight factors as a weight, and weight each candidate object personal information semantic encoding feature vector in the second to fifth candidate object personal information semantic encoding feature vectors respectively to obtain the four adjacent candidate object personal information joint encoding vectors.

[0051] For example, according to the personnel examination arrangement system based on multidimensional data analysis of an embodiment of the present application, the semantic association score vector calculation unit is used to:

[0052] Concatenate the first candidate object personal information semantic encoding feature vector with each candidate object personal information semantic encoding feature vector in the second to fifth candidate object personal information semantic encoding feature vectors to obtain a set of candidate object personal information semantic cascade feature vectors;

[0053] Each semantic cascade feature vector of candidate object personal information in the set of semantic cascade feature vectors of candidate object personal information is multiplied by the candidate object personal information weight matrix and then added positionally with the candidate object personal information bias vector to obtain the set of semantic association score vectors of candidate object personal information.

[0054] According to the personnel examination arrangement method and system based on multidimensional data analysis of the embodiment of the present application, the personal information of the target candidate object and the personal information of the other four candidate objects adjacent to the target candidate object are collected, and the data processing and semantic understanding algorithm based on artificial intelligence and natural language processing technology is introduced in the back end to extract the semantic features of the personal information of the target candidate object and the personal information of the other four candidate objects, and automatically learn and capture the implicit correlation relationship and semantic difference measurement coefficient between the personal semantic features of the target candidate object and the personal semantic features of the other four candidate objects, so as to perform an intelligent evaluation of the rationality of the examination seat arrangement to judge the rationality level of the examination seat arrangement. In this way, a strong guarantee can be provided for the fairness of the examination. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings of the embodiments of the present application are briefly introduced below. Obviously, the drawings described below only relate to some embodiments of the present application, and are not intended to limit the present application.

[0056] Figure 1 A schematic diagram of the application architecture of a personnel examination arrangement method based on multidimensional data analysis in an embodiment of the present application is shown;

[0057] Figure 2 A flow chart of a method for arranging personnel examinations based on multidimensional data analysis in an embodiment of the present application is shown;

[0058] Figure 3 A flowchart of sub-step S550 of the personnel examination arrangement method based on multi-dimensional data analysis in an embodiment of the present application is shown;

[0059] Figure 4 A schematic diagram of the structure of a personnel examination arrangement system based on multidimensional data analysis in an embodiment of the present application is shown; and

[0060] Figure 5An application scenario diagram of a method for arranging personnel examinations based on multidimensional data analysis in an embodiment of the present application is shown. DETAILED DESCRIPTION

[0061] The technical solutions in the embodiments of the present application will be described clearly and completely below in conjunction with the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, not all embodiments. Based on the embodiments of the present application, all other embodiments obtained by ordinary technicians in this field without creative work also fall within the scope of protection of the present application.

[0062] The terms used in this specification are those common terms currently widely used in the art in consideration of the functions of the present application, but these terms may vary according to the intentions of those skilled in the art, precedents, or new technologies in the art. In addition, specific terms may be selected, and in this case, their detailed meanings will be described in the detailed description of the present application. Therefore, the terms used in the specification should not be understood as simple names, but rather as a general description based on the meaning of the terms and the present application.

[0063] Although the present application makes various references to certain modules in the system according to the embodiments of the present application, any number of different modules can be used and run on the user terminal and / or server. The modules are only illustrative, and different aspects of the system and method can use different modules.

[0064] Flowcharts are used in the present application to illustrate the operations performed by the system according to the embodiments of the present application. It should be understood that the preceding or following operations are not necessarily performed accurately in order. On the contrary, various steps may be processed in reverse order or simultaneously as required. At the same time, other operations may also be added to these processes, or a certain step or several steps of operations may be removed from these processes.

[0065] Figure 1 A schematic diagram of the application architecture of a personnel examination arrangement method based on multi-dimensional data analysis in an embodiment of the present application is shown, including a server 100 and a terminal device 200.

[0066] The terminal device 200 and the server 100 can be connected through the Internet to achieve mutual communication. Optionally, the above-mentioned Internet uses standard communication technology and / or protocol. The Internet is usually the Internet, but it can also be any network, including but not limited to any combination of local area network (Locals Area Network, LAN), metropolitan area network (Metropolitan Area Network, MAN), wide area network (Wide Area Network, WAN), mobile, wired or wireless network, dedicated network or virtual private network. In some embodiments, the data exchanged through the network is represented by the technology and / or format including Hypertext Markup Language (HyperText Markup Language, HTML), Extensible Markup Language (Extensible Markup Language, XML), etc. In addition, conventional encryption technologies such as Secure Socket Layer (Secure Socket Layer, SSL), Transport Layer Security (Transport Layer Security, TLS), Virtual Private Network (Virtual Private Network, VPN), Internet Protocol Security (Internet Protocol Security, IPsec) can also be used to encrypt all or some links. In other embodiments, customized and / or dedicated data communication technology can also be used to replace or supplement the above-mentioned data communication technology.

[0067] The server 100 can provide various network services for the terminal device 200, wherein the server 100 can be a single server, a server cluster consisting of several servers, or a cloud computing center. Specifically, the server 100 may include a processor 110 (Center Processing Unit, CPU), a memory 120, an input device 130, and an output device 140, etc. The input device 130 may include a keyboard, a mouse, a touch screen, etc., and the output device 140 may include a display device, such as a liquid crystal display (Liquid Crystal Display, LCD), a cathode ray tube (Cathode Ray Tube, CRT), etc.

[0068] The memory 120 may include a read-only memory (ROM) and a random access memory (RAM), and provides the processor 110 with program instructions and data stored in the memory 120. In an embodiment of the present application, the memory 120 may be used to store the program of the personnel examination arrangement method based on multidimensional data analysis in an embodiment of the present application.

[0069] The processor 110 calls the program instructions stored in the memory 120, and the processor 110 is used to execute the steps of any one of the personnel examination arrangement methods based on multidimensional data analysis in the embodiments of the present application according to the obtained program instructions.

[0070] In addition, the application architecture diagram in the embodiment of the present application is intended to more clearly illustrate the technical solution in the embodiment of the present application, and does not constitute a limitation on the technical solution provided in the embodiment of the present application. Of course, for other application architectures and business applications, the technical solution provided in the embodiment of the present application is also applicable to similar problems.

[0071] The following is a non-restrictive description of the personnel examination arrangement method based on multidimensional data analysis provided according to at least one embodiment of the present application through several examples or embodiments. As described below, different features in these specific examples or embodiments can be combined with each other without conflicting with each other to obtain new examples or embodiments, and these new examples or embodiments also fall within the scope of protection of the present application.

[0072] With the rapid development of information technology, especially the application of big data and artificial intelligence technology, in-depth analysis of candidates' personal information has become possible. Through multi-dimensional data analysis, we can have a more comprehensive understanding of the characteristics of candidates, including but not limited to basic information, educational background, work experience, professional skills, etc. The deep learning model has the ability of self-learning, and can automatically extract features from a large amount of data and learn, so that it may more accurately identify the potential connections between candidates' information. Therefore, through in-depth mining of candidate information and potential correlation analysis, technical support is provided for the optimization of examination seat arrangements.

[0073] Based on this, in the technical solution of the present application, a personnel examination arrangement method based on multidimensional data analysis is proposed, which can collect the personal information of the target candidate object, including but not limited to basic information, educational background, work experience, professional skills, etc., as well as the personal information of the other four candidate objects adjacent to the target candidate object, and introduce data processing and semantic understanding algorithms based on artificial intelligence and natural language processing technology at the back end to extract the semantic features of the personal information of the target candidate object and the personal information of the other four candidate objects, and automatically learn and capture the implicit correlation relationship and semantic difference measurement coefficient between the personal semantic features of the target candidate object and the personal semantic features of the other four candidate objects, so as to perform an intelligent evaluation of the rationality of the examination seat arrangement to judge the rationality of the examination seat arrangement. In this way, a more intelligent personnel examination arrangement can be achieved based on natural language processing technology and multidimensional data analysis to avoid the situation where candidates with close relationships or similar professional skills sit next to each other and cheat in the examination, and provide a strong guarantee for the fairness of personnel examinations.

[0074] Figure 2 The flowchart of the method for arranging personnel examinations based on multidimensional data analysis in an embodiment of the present application is shown. For example, the method for arranging personnel examinations based on multidimensional data analysis can be executed by a server, which can be Figure 1 The server 100 shown in FIG. Figure 2 As shown, according to the personnel examination arrangement method based on multidimensional data analysis of the embodiment of the present application, the steps include: S510, obtaining the personal information of the first candidate object; S520, extracting the personal information of the second to fifth candidate objects adjacent to the first candidate object from the candidate information database; S530, semantically encoding the personal information of the first candidate object to obtain the semantic encoding feature vector of the first candidate object personal information; S540, semantically encoding the personal information of the second to fifth candidate objects to obtain the semantic encoding feature vectors of the second to fifth candidate objects personal information; S550, jointly encoding the semantic encoding feature vector of the first candidate object personal information with each candidate object personal information semantic encoding feature vector in the second to fifth candidate object personal information semantic encoding feature vector to obtain four adjacent candidate object personal information joint encoding vectors; S560, calculating the personal information semantic difference measurement coefficient between any two adjacent candidate object personal information joint encoding vectors in the four adjacent candidate object personal information joint encoding vectors to obtain a candidate object personal information semantic difference topological distribution matrix composed of multiple personal information semantic difference measurement coefficients; S570, based on the candidate object personal information semantic difference topological distribution matrix, determining a reasonable grade label for the examination seat arrangement.

[0075] Specifically, in the technical solution of the present application, first, the personal information of the first candidate object is obtained, and the personal information of the second to fifth candidate objects adjacent to the first candidate object is extracted from the candidate information database, wherein the personal information of the candidate object includes but is not limited to basic information, educational background, work experience, professional skills, etc. Then, the personal information of the first candidate object is semantically encoded to extract the semantic coding features in the personal information of the first candidate object, thereby obtaining the semantic coding feature vector of the personal information of the first candidate object. In addition, the personal information of the second to fifth candidate objects is semantically encoded to extract the semantic coding features in the personal information of the second to fifth candidate objects, respectively, thereby obtaining the semantic coding feature vectors of the personal information of the second to fifth candidate objects.

[0076] It is worth mentioning that in the field of data storage, the format and structure of personal information presents a wide variety, which mainly depends on the design of the storage system, the source of the data and the intended use. For example, relational databases are in tabular form with a fixed row and column structure, and each column has a specific data type. Non-relational databases, or NoSQL, provide more flexible data storage formats, such as document storage, key-value pairs, wide column storage and graph databases, which can store structured, semi-structured or unstructured data. Data in plain text format, such as CSV files, separate data items with commas, each row represents a record, and each column represents a field. It is a simple and widely used data exchange format. XML and JSON are more complex data representation methods. XML uses a hierarchical markup language, while JSON is a lightweight data exchange format based on key-value pairs, which is easy for people to read and write, and easy for machines to parse and generate. Binary files store multimedia data such as images, audio, and video. They usually have a specific encoding format and may contain metadata. Excel spreadsheets provide a more interactive data environment that allows formulas and charts. Of course, it can also exist in text form, which can record information such as timestamp, error level and message. The format may vary depending on the application.

[0077] It should be understood that in the arrangement of examination seats, there may be some social connection or similar professional skills between adjacent candidates, which may increase the risk of cheating. Therefore, in order to be able to perform a correlation analysis on the personal information semantics of the second to fifth candidate objects based on the personal information semantics of the first candidate object, to determine whether the second to fifth candidate objects have such a correlation relationship with the first candidate object, thereby providing a basis for the subsequent examination seat arrangement, in the technical solution of the present application, the first candidate object personal information semantic coding feature vector is further jointly encoded with each candidate object personal information semantic coding feature vector in the second to fifth candidate object personal information semantic coding feature vector to obtain four adjacent candidate object personal information joint coding vectors.

[0078] It is worth mentioning that joint coding can quantify and optimize the semantic association between the semantic coding feature vectors of each candidate object's personal information to improve the quality of the semantic association feature representation of the adjacent candidate object's personal information, and provide a basis for the subsequent seat rationality detection. Specifically, first, by calculating the semantic association score vectors between the first candidate object's personal information semantic coding feature vector and each candidate object's personal information semantic coding feature vector in the second to fifth candidate object's personal information semantic coding feature vectors, the similarity or difference between the first candidate object's personal information semantic feature and the other candidate object's personal information semantic feature can be quantified to understand the inherent semantic connection between the two, and generate a set of candidate object personal information semantic association score vectors. Then, based on the aggregate information of the set of candidate object personal information semantic association score vectors, the global representation vector of the semantic relevance of the candidate object's personal information semantics is determined, providing an overall context for subsequent association optimization. Next, based on the global representation vector of the semantic relevance of the candidate object's personal information semantics, the association optimization factor of the second to fifth candidate object personal information semantic coding feature vectors is calculated. The association optimization factor highlights the consistency or deviation between the second to fifth candidate object personal information semantic coding feature vectors and the overall semantics of the global representation vector of the semantic relevance of the candidate object's personal information semantics, and provides guidance for feature adjustment. In order to enhance the model's ability to distinguish weights and make the weights more adaptable to the needs of feature optimization, the associated optimization factor sequence is further input into the Softmax activation function to convert it into a weight factor to achieve factor normalization and nonlinear mapping. Finally, the semantically associated optimization weight factors of each candidate object's personal information are used to weight the semantically encoded feature vectors of the second to fifth candidate objects' personal information to generate the four adjacent candidate objects' personal information joint encoding vectors. That is, the weighted operation is combined with the original semantic feature information of the second to fifth candidate objects' personal information and the optimization factor to enhance the expressiveness of the features, which helps to improve the performance of the model. In this way, the examination arrangement system can process candidate information more carefully, reduce the possibility of cheating in the examination, and improve the overall management level and fairness of the examination.

[0079] Accordingly, if Figure 3As shown, the first candidate object personal information semantic coding feature vector and each candidate object personal information semantic coding feature vector in the second to fifth candidate object personal information semantic coding feature vectors are jointly encoded to obtain four adjacent candidate object personal information joint coding vectors, including: S551, calculating the semantic association score vectors between the first candidate object personal information semantic coding feature vector and each candidate object personal information semantic coding feature vector in the second to fifth candidate object personal information semantic coding feature vectors respectively to obtain a set of candidate object personal information semantic association score vectors; S552, based on the set of candidate object personal information semantic association score vectors, determining the candidate object personal information semantic correlation global representation vector; S553, based on the candidate object personal information semantic correlation score vectors S554, input the sequence of candidate object personal information semantic association optimization factors into the Softmax activation function to obtain a sequence of candidate object personal information semantic association optimization weight factors; S555, using each candidate object personal information semantic association optimization weight factor in the sequence of candidate object personal information semantic association optimization weight factors as a weight, respectively weight each candidate object personal information semantic encoding feature vector in the second to fifth candidate object personal information semantic encoding feature vectors to obtain the four adjacent candidate object personal information joint encoding vectors.

[0080] Among them, in step S551, the semantic association score vectors between the first candidate object personal information semantic coding feature vector and each candidate object personal information semantic coding feature vector in the second to fifth candidate object personal information semantic coding feature vectors are calculated to obtain a set of candidate object personal information semantic association score vectors, including: cascading the first candidate object personal information semantic coding feature vector with each candidate object personal information semantic coding feature vector in the second to fifth candidate object personal information semantic coding feature vectors to obtain a set of candidate object personal information semantic cascade feature vectors; multiplying each candidate object personal information semantic cascade feature vector in the set of candidate object personal information semantic cascade feature vectors with the candidate object personal information weight matrix and then adding them by position with the candidate object personal information bias vector to obtain the set of candidate object personal information semantic association score vectors.

[0081] Among them, in step S552, based on the set of semantic association score vectors of the candidate object's personal information, a global representation vector of the semantic relevance of the candidate object's personal information is determined, including: calculating the mean vector of the set of semantic association score vectors of the candidate object's personal information as the global representation vector of the semantic relevance of the candidate object's personal information.

[0082] Among them, in step S553, based on the global representation vector of the semantic relevance of the candidate object personal information, the association optimization factor of each candidate object personal information semantic encoding feature vector in the second to fifth candidate object personal information semantic encoding feature vectors is calculated to obtain a sequence of candidate object personal information semantic association optimization factors, including: multiplying the global representation vector of the semantic relevance of the candidate object personal information with the candidate object personal information global weight matrix to obtain the candidate object personal information global representation weight vector; multiplying each candidate object personal information semantic encoding feature vector in the second to fifth candidate object personal information semantic encoding feature vectors with the corresponding candidate object personal information local weight matrix to obtain a set of candidate object personal information local representation weight vectors; calculating each candidate object personal information local representation weight vector in the set of candidate object personal information local representation weight vectors, adding them to the candidate object personal information global representation weight vector and the corresponding candidate object personal information bias vector by position to obtain a set of candidate object personal information semantic association vectors; inputting the set of candidate object personal information semantic association vectors into the Sigmoid function to obtain a sequence of candidate object personal information semantic association optimization factors.

[0083] In one example, jointly encoding the first candidate object personal information semantic encoding feature vector and each candidate object personal information semantic encoding feature vector in the second to fifth candidate object personal information semantic encoding feature vectors to obtain four adjacent candidate object personal information joint encoding vectors, including: jointly encoding the first candidate object personal information semantic encoding feature vector and each candidate object personal information semantic encoding feature vector in the second to fifth candidate object personal information semantic encoding feature vectors using the following joint encoding formula to obtain the four adjacent candidate object personal information joint encoding vectors; wherein the joint encoding formula is:

[0084] X={x1,x2,…,x n}

[0085] r h =W h [x i ;v z ]+b h

[0086]

[0087] G i =Softmax[Sigmoid(M m R+H m x i +b m )]

[0088] {x i ′}={G i ·x i}

[0089] Where X is the semantic encoding feature vector of the personal information of the second to fifth examinee objects, n represents the number of feature vectors in the semantic encoding feature vector of the personal information of the second to fifth examinee objects, and x i is the semantic encoding feature vector of the personal information of the i-th candidate object among the semantic encoding feature vectors of the personal information of the second to fifth candidate objects, v z is the semantic encoding feature vector of the personal information of the first candidate object, [·;·] is the cascade processing operation, W h is the personal information weight matrix of the examinee, b h is the candidate object personal information bias vector, r h is the hth candidate object personal information semantic association score vector in the set of candidate object personal information semantic association score vectors, T is the number of vectors in the set of candidate object personal information semantic association score vectors, R is the candidate object personal information semantic relevance global representation vector, M m and H m are the global weight matrix of the personal information of the examinee object and the local weight matrix of the personal information of the examinee object, respectively. m is the candidate’s personal information bias vector, sigmoid(·) is the sigmoid function, Softmax(·) is the Softmax function, G i is the i-th candidate object personal information semantic association optimization weight factor in the sequence of the candidate object personal information semantic association optimization weight factors, x i ′ is the joint encoding vector of personal information of the i-th neighboring candidate object among the four joint encoding vectors of personal information of the neighboring candidate objects.

[0090] Furthermore, since each adjacent candidate object personal information joint coding vector in the four adjacent candidate object personal information joint coding vectors respectively contains the personal information coding features about the first candidate object and the personal information semantic coding joint features of the other four candidate objects. In order to be able to compare the differences between the personal information semantic joint coding features of these candidate objects, thereby avoiding candidates with relatively large differences from being assigned together, in the technical solution of the present application, the personal information semantic difference measurement coefficient between any two adjacent candidate object personal information joint coding vectors in the four adjacent candidate object personal information joint coding vectors is further calculated to obtain a candidate object personal information semantic difference topological distribution matrix composed of multiple personal information semantic difference measurement coefficients. In particular, the calculation process of the personal information semantic difference measurement coefficient is: first, by calculating the difference vector between any two adjacent candidate objects' personal information joint coding vectors among the four adjacent candidate objects' personal information joint coding vectors to obtain the semantic difference between the two quantities to quantify, and at the same time, calculating the covariance matrix between the any two adjacent candidate objects' personal information joint coding vectors, and based on the difference features and the covariance matrix, reversely obtaining the difference features between the two as the personal information semantic difference measurement coefficient. In particular, the personal information semantic difference measurement coefficient represents the difference between the joint coding features of personal information of any two adjacent candidate objects, thereby constructing a topological distribution matrix of the candidate object personal information semantic differences of the four adjacent candidate objects, which provides a basis for judging the rationality of the arrangement of examination seats.

[0091] Correspondingly, in step S560, the personal information semantic difference measurement coefficient between any two adjacent candidate object personal information joint coding vectors in the four adjacent candidate object personal information joint coding vectors is calculated to obtain a candidate object personal information semantic difference topological distribution matrix composed of multiple personal information semantic difference measurement coefficients, including: calculating the positional difference between any two adjacent candidate object personal information joint coding vectors in the four adjacent candidate object personal information joint coding vectors to obtain a sequence of adjacent candidate object personal information semantic difference feature vectors; calculating the covariance matrix between any two adjacent candidate object personal information joint coding vectors in the four adjacent candidate object personal information joint coding vectors, and calculating the inverse of each eigenvalue in the covariance matrix to obtain a sequence of adjacent candidate object personal information semantic anti-correlation covariance matrices; calculating the positional difference between each adjacent candidate object personal information in the sequence of adjacent candidate object personal information semantic difference feature vectors. The method comprises the following steps: calculating the transpose vector of the personal information semantic difference feature vector of the neighboring candidate object to obtain a sequence of the personal information semantic difference transpose representation vectors of the neighboring candidate object; calculating the product of each group of corresponding neighboring candidate object personal information semantic difference transpose representation vectors, neighboring candidate object personal information semantic anti-correlation covariance matrix and neighboring candidate object personal information semantic difference feature vector in the sequence of the neighboring candidate object personal information semantic difference representation vectors, the sequence of the neighboring candidate object personal information semantic anti-correlation covariance matrix and the neighboring candidate object personal information semantic difference feature vector to obtain a sequence of neighboring candidate object personal information semantic difference correlation values; calculating the square root of each neighboring candidate object personal information semantic difference correlation value in the sequence of neighboring candidate object personal information semantic difference correlation values ​​to obtain the multiple personal information semantic difference measurement coefficients; and arranging the multiple personal information semantic difference measurement coefficients in a matrix to obtain a candidate object personal information semantic difference topological distribution matrix.

[0092] In a specific example, calculating the personal information semantic difference measurement coefficient between any two adjacent candidate object personal information joint coding vectors in the four adjacent candidate object personal information joint coding vectors to obtain the candidate object personal information semantic difference topological distribution matrix composed of multiple personal information semantic difference measurement coefficients, including: calculating the personal information semantic difference measurement coefficient between any two adjacent candidate object personal information joint coding vectors in the four adjacent candidate object personal information joint coding vectors using the following semantic difference measurement formula to obtain the candidate object personal information semantic difference topological distribution matrix composed of multiple personal information semantic difference measurement coefficients; wherein the semantic difference measurement formula is:

[0093]

[0094] Among them, x i ′ and xj ′ are the joint encoding vectors of personal information of any two adjacent examinee objects among the four joint encoding vectors of personal information of adjacent examinee objects, (·) T is the transpose of the vector, S is the covariance matrix between the joint encoding vectors of personal information of any two adjacent examinees, D(x i , v c ) is each personal information semantic difference measurement coefficient among the multiple personal information semantic difference measurement coefficients.

[0095] Then, the semantic difference topological distribution matrix of the personal information of the examinee objects is input into the intelligent evaluator for examination arrangement based on the classifier to obtain the evaluation result, and the evaluation result is used to represent the reasonable grade label of the examination seat arrangement. In other words, the semantic difference topological distribution characteristics of the personal information of the adjacent examinee objects are used for classification processing, so as to carry out the intelligent evaluation of the rationality of the examination seat arrangement, so as to judge the reasonable grade of the examination seat arrangement. In this way, it is possible to realize a more intelligent arrangement of personnel examinations based on natural language processing technology and multidimensional data analysis, so as to avoid the situation where examinees with close relationship or similar professional skills sit next to each other and cheat in the examination, and provide a strong guarantee for the fairness of personnel examinations.

[0096] Accordingly, in step S570, based on the semantic difference topological distribution matrix of the candidate object personal information, a reasonable grade label of the examination seat arrangement is determined, including: inputting the semantic difference topological distribution matrix of the candidate object personal information into a classifier-based examination arrangement intelligent evaluator to obtain an evaluation result, and the evaluation result is used to represent the reasonable grade label of the examination seat arrangement.

[0097] It should be understood that the role of the classifier is to use the given categories and known training data to learn classification rules and classifiers, and then classify (or predict) unknown data. Logistic regression and SVM are often used to solve binary classification problems. For multi-class classification problems, logistic regression or SVM can also be used, but multiple binary classifications are required to form a multi-classification. However, this is prone to errors and is not efficient. Commonly used multi-classification methods include the Softmax classification function.

[0098] In a preferred example, considering that the semantic coding feature vector of the candidate object personal information corresponding to each candidate object represents the semantic coding features of its personal information, and the four adjacent candidate object personal information joint coding vectors respectively represent the joint semantic coding features between the first candidate and the second to fifth candidates, therefore, considering the uneven distribution of the joint semantic coding features caused by the source semantic differences of the personal information, the personal information semantic difference measurement coefficient between any two adjacent candidate object personal information joint coding vectors will also have an aggregate mapping deviation towards the semantic difference measurement domain, thereby affecting the semantic difference measurement domain aggregation regression comprehensibility of the candidate object personal information semantic difference topological distribution matrix composed of the personal information semantic difference measurement coefficients, and affecting the accuracy of the evaluation results obtained by the classifier-based examination arrangement intelligent evaluator whose input is the classifier.

[0099] Based on this, in the preferred example, calculating the personal information semantic difference measurement coefficient between any two adjacent candidate object personal information joint coding vectors in the four adjacent candidate object personal information joint coding vectors to obtain a candidate object personal information semantic difference topological distribution matrix composed of multiple personal information semantic difference measurement coefficients specifically includes: cascading the four adjacent candidate object personal information joint coding vectors into four adjacent candidate object personal information joint coding cascade vectors; subtracting the zero norm of the four adjacent candidate object personal information joint coding cascade vectors from the length of the four adjacent candidate object personal information joint coding cascade vectors to obtain four adjacent candidate object personal information joint coding cascade vectors; Information joint coding cascade isolated zero value; calculate the logarithm of the square of the isolated zero value of the joint coding cascade of the four adjacent candidate objects' personal information and the sum of the isolated zero values ​​of the joint coding cascade of the four adjacent candidate objects' personal information to obtain the superimposed information value of the joint coding cascade of the four adjacent candidate objects' personal information; calculate a power function with each eigenvalue of the joint coding cascade vector of the four adjacent candidate objects' personal information as the base, and the difference of the isolated zero value of the joint coding cascade of the four adjacent candidate objects' personal information minus one as the exponent, and perform point multiplication with the superimposed information value of the joint coding cascade of the four adjacent candidate objects' personal information to obtain the first four adjacent candidate objects' personal information joint coding cascade The intermediate vector of the coding cascade is obtained by multiplying the four adjacent candidate object personal information joint coding cascade vectors by the difference between the four adjacent candidate object personal information joint coding cascade isolated zero values ​​minus one, and then multiplying the difference with the reciprocal of the four adjacent candidate object personal information joint coding cascade isolated zero values ​​to obtain the second four adjacent candidate object personal information joint coding cascade intermediate vectors; calculate an exponential function with the natural constant as the base, with each eigenvalue of the second four adjacent candidate object personal information joint coding cascade intermediate vectors as the exponent to obtain the third four adjacent candidate object personal information joint coding cascade intermediate vectors; multiply the first four adjacent candidate object personal information joint coding cascade intermediate vectors by the reciprocal of the four adjacent candidate object personal information joint coding cascade isolated zero values. The combined coding concatenation intermediate vector is added to the third four adjacent candidate object personal information joint coding cascade intermediate vectors to obtain the optimized four adjacent candidate object personal information joint coding cascade vectors; the optimized four adjacent candidate object personal information joint coding cascade vectors are split into four optimized adjacent candidate object personal information joint coding vectors; and the personal information semantic difference measurement coefficient between any two optimized adjacent candidate object personal information joint coding vectors among the four optimized adjacent candidate object personal information joint coding vectors is calculated to obtain a candidate object personal information semantic difference topological distribution matrix composed of multiple personal information semantic difference measurement coefficients.

[0100] That is, the optimization process of the joint encoding concatenated vector of the personal information of the four adjacent examinees, for example, denoted as V, is expressed by the following optimization formula:

[0101]

[0102] n=L-||V||0

[0103] Wherein, V represents the four adjacent candidate object personal information joint coding cascade vector, ||V||0 represents the 0 norm of the four adjacent candidate object personal information joint coding cascade vector, L represents the length of the four adjacent candidate object personal information joint coding cascade vector, n represents the four adjacent candidate object personal information joint coding cascade isolated zero value, exp[·] represents the exponential operation of the value, the exponential operation of the value represents the calculation of the natural exponential function value with the value as the power, log(·) represents the logarithmic function value with 2 as the base, V′ represents the optimized four adjacent candidate object personal information joint coding cascade vector, ⊙ represents the dot product, Indicates point addition.

[0104] In this way, the high-dimensional feature manifold of the joint encoding cascade vectors of the personal information of the four adjacent candidate objects is represented by the eigenvalues ​​of the feature set as the vector field of the aggregation dimension, and the overlapping values ​​of the vector fields of the joint encoding cascade vectors of the four adjacent candidate objects at the isolated zero position are used as the order information to fix the local position of the eigenvalues ​​of the feature set, and a bias for the reversibility of the feature regression distribution field of the joint encoding cascade vectors of the four adjacent candidate objects is added as a reward to achieve the mapping target tracking of the regression distribution of the joint encoding cascade vectors of the personal information of the four adjacent candidate objects for the eigenvalue position, so that the feature set of the joint encoding cascade vectors of the personal information of the four adjacent candidate objects can perceive the mapping migration to the aggregate distribution, thereby improving the semantic difference measurement domain aggregation regression comprehensibility of the joint encoding cascade vectors of the four adjacent candidate objects, and improving the calculation accuracy of the personal information semantic difference measurement coefficient between any two optimized joint encoding vectors of the four optimized joint encoding vectors of the personal information of the adjacent candidate objects, thereby improving the expression effect of the topological distribution matrix of the semantic difference of the personal information of the candidate objects, and improving the accuracy of the evaluation result obtained by the intelligent evaluator for examination arrangement based on the classifier. In this way, a more accurate and intelligent assessment of the test seat arrangement can be conducted to determine the rationality of the test seat arrangement, thereby avoiding cheating in the test due to candidates with close relationships or similar professional skills sitting next to each other, and providing a strong guarantee for the fairness of personnel examinations.

[0105] Based on the above embodiments, see Figure 4As shown, it is a structural diagram of a personnel examination arrangement system 800 based on multi-dimensional data analysis in an embodiment of the present application. The personnel examination arrangement system 800 based on multi-dimensional data analysis includes: a first candidate personal information acquisition module 810, used to obtain the personal information of the first candidate object; an adjacent candidate information extraction module 820, used to extract the personal information of the second to fifth candidate objects adjacent to the first candidate object from the candidate information database; a first semantic encoding module 830, used to semantically encode the personal information of the first candidate object to obtain the semantic encoding feature vector of the first candidate object's personal information; a second semantic encoding module 840, used to semantically encode the personal information of the second to fifth candidate objects to obtain the semantic encoding feature vector of the second to fifth candidate objects' personal information; a joint encoding module 850, used to semantically encode the personal information of the first candidate object The semantic coding feature vector is jointly encoded with the personal information semantic coding feature vectors of each candidate object in the second to fifth candidate object personal information semantic coding feature vectors to obtain four adjacent candidate object personal information joint coding vectors; a personal information semantic difference measurement coefficient calculation module 860 is used to calculate the personal information semantic difference measurement coefficient between any two adjacent candidate object personal information joint coding vectors in the four adjacent candidate object personal information joint coding vectors to obtain a candidate object personal information semantic difference topological distribution matrix composed of multiple personal information semantic difference measurement coefficients; a grade label determination module 870 is used to determine a reasonable grade label for the examination seat arrangement based on the candidate object personal information semantic difference topological distribution matrix.

[0106] In one example, in the above-mentioned personnel examination arrangement system 800 based on multidimensional data analysis, the joint encoding module 850 includes: a semantic association score vector calculation unit, which is used to calculate the semantic association score vectors between the first candidate object personal information semantic encoding feature vector and each of the candidate object personal information semantic encoding feature vectors in the second to fifth candidate object personal information semantic encoding feature vectors to obtain a set of candidate object personal information semantic association score vectors; a semantic correlation unit, which is used to determine the candidate object personal information semantic correlation global representation vector based on the set of candidate object personal information semantic correlation score vectors; and a correlation optimization factor calculation unit, which is used to calculate the second to fifth candidate object personal information semantic correlation global representation vector based on the candidate object personal information semantic correlation global representation vector. The associated optimization factors of each candidate object personal information semantic coding feature vector in the candidate object personal information semantic coding feature vector are used to obtain a sequence of candidate object personal information semantic association optimization factors; an activation unit is used to input the sequence of candidate object personal information semantic association optimization factors into a Softmax activation function to obtain a sequence of candidate object personal information semantic association optimization weight factors; a weighting unit is used to use each candidate object personal information semantic association optimization weight factor in the sequence of candidate object personal information semantic association optimization weight factors as a weight, and weight each candidate object personal information semantic coding feature vector in the second to fifth candidate object personal information semantic coding feature vectors respectively to obtain the four adjacent candidate object personal information joint coding vectors.

[0107] In one example, in the above-mentioned personnel examination arrangement system 800 based on multidimensional data analysis, the semantic association score vector calculation unit is used to: cascade the first candidate object personal information semantic coding feature vector with each candidate object personal information semantic coding feature vector in the second to fifth candidate object personal information semantic coding feature vectors to obtain a set of candidate object personal information semantic cascade feature vectors; multiply each candidate object personal information semantic cascade feature vector in the set of candidate object personal information semantic cascade feature vectors with the candidate object personal information weight matrix, and then add them by position with the candidate object personal information bias vector to obtain the set of candidate object personal information semantic association score vectors.

[0108] Here, those skilled in the art can understand that the specific functions and operations of each module in the above-mentioned personnel examination arrangement system 800 based on multi-dimensional data analysis have been described in the above reference. Figure 2 to Figure 3 The method for arranging personnel examinations based on multidimensional data analysis has been introduced in detail, and therefore, its repeated description will be omitted.

[0109] Figure 5FIG. 1 is an application scenario diagram of a personnel examination arrangement method based on multidimensional data analysis according to an embodiment of the present application. Figure 5 As shown, in this application scenario, first, the personal information of the first candidate object is obtained (for example, Figure 5 D1 shown in the figure) and extracting personal information of the second to fifth candidate objects adjacent to the first candidate object from the candidate information database (for example, Figure 5 Then, the personal information of the first candidate object and the personal information of the second to fifth candidate objects are input into a server (for example, Figure 5 In S) as shown in , the server is able to use the personnel examination arrangement algorithm based on multidimensional data analysis to process the personal information of the first candidate object and the personal information of the second to fifth candidate objects to obtain an evaluation result for representing a reasonable grade label for the examination seat arrangement.

[0110] Based on the above embodiment, another exemplary embodiment of an electronic device is also provided in the embodiment of the present application. In some possible implementations, the electronic device in the embodiment of the present application may include a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor can implement the steps of the personnel examination arrangement method based on multidimensional data analysis in the above embodiment when executing the program.

[0111] For example, in the case of electronic equipment Figure 1 Taking the server 100 in the example as an example, the processor in the electronic device is the processor 110 in the server 100, and the memory in the electronic device is the memory 120 in the server 100.

[0112] The embodiment of the present application also provides a computer-readable storage medium, on which computer-executable instructions are stored. When the computer-executable instructions are executed by a processor, the personnel examination arrangement method based on multidimensional data analysis according to the embodiment of the present application described with reference to the above figures can be executed. The computer-readable storage medium includes, but is not limited to, for example, volatile memory and / or non-volatile memory. The volatile memory may, for example, include a random access memory (RAM) and / or a cache memory (cache), etc. The non-volatile memory may, for example, include a read-only memory (ROM), a hard disk, a flash memory, etc.

[0113] The embodiment of the present application also provides a computer program product or a computer program, which includes computer executable instructions, and the computer executable instructions are stored in a computer readable storage medium. The processor of the computer device reads the computer executable instructions from the computer readable storage medium, and the processor executes the computer executable instructions, so that the computer device executes the personnel examination arrangement method based on multidimensional data analysis according to the embodiment of the present application.

[0114] Those skilled in the art will appreciate that the contents disclosed in this application may be subject to various variations and improvements. For example, the various devices or components described above may be implemented by hardware, or by software, firmware, or a combination of some or all of the three.

[0115] In addition, although the present application makes various references to certain units in the system according to embodiments of the present application, any number of different units can be used and run on the client and / or server. The units are only illustrative, and different aspects of the system and method can use different units.

[0116] Those skilled in the art will appreciate that all or part of the steps in the above method can be completed by instructing related hardware through a program, and the program can be stored in a computer-readable storage medium, such as a read-only memory, a disk or an optical disk. Optionally, all or part of the steps in the above embodiment can also be implemented using one or more integrated circuits. Accordingly, each module / unit in the above embodiment can be implemented in the form of hardware or in the form of software function modules. The present application is not limited to any particular form of combination of hardware and software.

[0117] Unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by those of ordinary skill in the art to which this application belongs. It should also be understood that terms such as those defined in common dictionaries should be interpreted as having a meaning consistent with their meaning in the context of the relevant technology, and should not be interpreted in an idealized or extremely formal sense, unless explicitly defined as such herein.

[0118] The above is an explanation of the present application and should not be considered as limiting thereof. Although several exemplary embodiments of the present application have been described, those skilled in the art will readily appreciate that many modifications may be made to the exemplary embodiments without departing from the novel teachings and advantages of the present application.

Claims

1. A personnel examination arrangement method based on multidimensional data analysis, characterized in that: include: Obtain the personal information of the first candidate object; Extracting personal information of second to fifth candidate objects adjacent to the first candidate object from the candidate information database; Performing semantic encoding on the personal information of the first examinee object to obtain a semantic encoding feature vector of the personal information of the first examinee object; Performing semantic encoding on the personal information of the second to fifth examinee objects to obtain semantic encoding feature vectors of the personal information of the second to fifth examinee objects; Jointly encode the first candidate object personal information semantic encoding feature vector and each candidate object personal information semantic encoding feature vector in the second to fifth candidate object personal information semantic encoding feature vectors to obtain four adjacent candidate object personal information joint encoding vectors; Calculate the personal information semantic difference measurement coefficient between any two adjacent candidate object personal information joint coding vectors in the four adjacent candidate object personal information joint coding vectors to obtain a candidate object personal information semantic difference topological distribution matrix composed of multiple personal information semantic difference measurement coefficients; Based on the semantic difference topological distribution matrix of the candidate object's personal information, reasonable grade labels for the examination seat arrangement are determined.

2. The method for arranging personnel examinations based on multidimensional data analysis according to claim 1, characterized in that: The first candidate object personal information semantic coding feature vector is jointly encoded with each candidate object personal information semantic coding feature vector in the second to fifth candidate object personal information semantic coding feature vectors to obtain four adjacent candidate object personal information joint coding vectors, including: Calculate the semantic association score vectors between the first candidate object personal information semantic encoding feature vector and each candidate object personal information semantic encoding feature vector in the second to fifth candidate object personal information semantic encoding feature vectors to obtain a set of candidate object personal information semantic association score vectors; Determine a global representation vector of the semantic relevance of the candidate object's personal information based on the set of semantic relevance score vectors of the candidate object's personal information; Based on the global representation vector of the semantic relevance of the personal information of the examinee object, calculating the association optimization factor of each of the semantic encoding feature vectors of the personal information of the examinee object in the second to fifth semantic encoding feature vectors of the personal information of the examinee object to obtain a sequence of semantic association optimization factors of the personal information of the examinee object; Inputting the sequence of the semantic association optimization factors of the candidate object personal information into the Softmax activation function to obtain the sequence of the semantic association optimization weight factors of the candidate object personal information; Taking each candidate object personal information semantic association optimization weight factor in the sequence of candidate object personal information semantic association optimization weight factors as a weight, each candidate object personal information semantic encoding feature vector in the second to fifth candidate object personal information semantic encoding feature vectors is weighted respectively to obtain the four adjacent candidate object personal information joint encoding vectors.

3. The method for arranging personnel examinations based on multidimensional data analysis according to claim 2, characterized in that: Calculating the semantic association score vectors between the first candidate object personal information semantic encoding feature vector and each candidate object personal information semantic encoding feature vector in the second to fifth candidate object personal information semantic encoding feature vectors to obtain a set of candidate object personal information semantic association score vectors, including: Concatenate the first candidate object personal information semantic encoding feature vector with each candidate object personal information semantic encoding feature vector in the second to fifth candidate object personal information semantic encoding feature vectors to obtain a set of candidate object personal information semantic cascade feature vectors; Each semantic cascade feature vector of candidate object personal information in the set of semantic cascade feature vectors of candidate object personal information is multiplied by the candidate object personal information weight matrix and then added positionally with the candidate object personal information bias vector to obtain the set of semantic association score vectors of candidate object personal information.

4. The method for arranging personnel examinations based on multidimensional data analysis according to claim 3 is characterized in that: Based on the set of the candidate object personal information semantic association score vectors, determine the candidate object personal information semantic relevance global representation vector, including: calculating the mean vector of the set of the candidate object personal information semantic association score vectors as the candidate object personal information semantic relevance global representation vector.

5. The method for arranging personnel examinations based on multidimensional data analysis according to claim 4, characterized in that: Based on the global representation vector of the semantic relevance of the personal information of the examinee object, calculating the association optimization factor of each of the semantic encoding feature vectors of the personal information of the examinee object in the second to fifth semantic encoding feature vectors of the personal information of the examinee object to obtain a sequence of semantic association optimization factors of the personal information of the examinee object, including: Multiplying the global representation vector of the semantic relevance of the candidate object personal information by the global weight matrix of the candidate object personal information to obtain the global representation weight vector of the candidate object personal information; Multiply each candidate object personal information semantic encoding feature vector in the second to fifth candidate object personal information semantic encoding feature vectors by the corresponding candidate object personal information local weight matrix to obtain a set of candidate object personal information local representation weight vectors; Calculate each candidate object personal information local representation weight vector in the set of candidate object personal information local representation weight vectors, add them to the candidate object personal information global representation weight vector and the corresponding candidate object personal information bias vector by position to obtain a set of candidate object personal information semantic association vectors; The set of semantic association vectors of the candidate object's personal information is input into a Sigmoid function to obtain a sequence of optimization factors of the semantic association of the candidate object's personal information.

6. The method for arranging personnel examinations based on multidimensional data analysis according to claim 5, characterized in that: Calculate the personal information semantic difference measurement coefficient between any two adjacent candidate object personal information joint coding vectors in the four adjacent candidate object personal information joint coding vectors to obtain a candidate object personal information semantic difference topological distribution matrix composed of multiple personal information semantic difference measurement coefficients, including: Calculate the positional difference between any two adjacent candidate object personal information joint encoding vectors among the four adjacent candidate object personal information joint encoding vectors to obtain a sequence of adjacent candidate object personal information semantic difference feature vectors; Calculate the covariance matrix between any two adjacent candidate object personal information joint encoding vectors in the four adjacent candidate object personal information joint encoding vectors, and calculate the inverse of each eigenvalue in the covariance matrix to obtain a sequence of adjacent candidate object personal information semantic anti-correlation covariance matrices; Calculate the transposed vector of each of the semantic differential feature vectors of personal information of the adjacent examinee objects in the sequence of semantic differential feature vectors of personal information of the adjacent examinee objects to obtain a sequence of transposed representation vectors of semantic differential representation of personal information of the adjacent examinee objects; Calculate the product of each group of corresponding neighboring candidate object personal information semantic difference transposed representation vectors, neighboring candidate object personal information semantic anti-correlation covariance matrix and neighboring candidate object personal information semantic difference eigenvectors in the sequence of neighboring candidate object personal information semantic difference transposed representation vectors, the sequence of neighboring candidate object personal information semantic anti-correlation covariance matrix and neighboring candidate object personal information semantic difference eigenvectors to obtain a sequence of neighboring candidate object personal information semantic difference correlation values; Calculate the square root of each of the semantic difference correlation values ​​of the personal information of the adjacent examinee objects in the sequence of semantic difference correlation values ​​of the personal information of the adjacent examinee objects to obtain the plurality of semantic difference measurement coefficients of the personal information; The multiple personal information semantic difference measurement coefficients are arranged in a matrix to obtain a topological distribution matrix of the personal information semantic differences of the examinee objects.

7. The method for arranging personnel examinations based on multidimensional data analysis according to claim 6, characterized in that: Based on the semantic difference topological distribution matrix of the candidate object personal information, determine the reasonable grade label of the examination seat arrangement, including: inputting the semantic difference topological distribution matrix of the candidate object personal information into a classifier-based examination arrangement intelligent evaluator to obtain an evaluation result, and the evaluation result is used to represent the reasonable grade label of the examination seat arrangement.

8. A personnel examination arrangement system based on multidimensional data analysis, characterized in that: include: The first candidate personal information acquisition module is used to acquire the personal information of the first candidate object; An adjacent examinee information extraction module, used to extract personal information of the second to fifth examinee objects adjacent to the first examinee object from the examinee information database; A first semantic encoding module, configured to semantically encode the personal information of the first examinee object to obtain a semantic encoding feature vector of the personal information of the first examinee object; A second semantic encoding module, used to semantically encode the personal information of the second to fifth examinee objects to obtain semantic encoding feature vectors of the personal information of the second to fifth examinee objects; A joint encoding module, used for jointly encoding the first candidate object personal information semantic encoding feature vector with each candidate object personal information semantic encoding feature vector in the second to fifth candidate object personal information semantic encoding feature vectors to obtain four adjacent candidate object personal information joint encoding vectors; A personal information semantic difference measurement coefficient calculation module is used to calculate the personal information semantic difference measurement coefficient between any two adjacent candidate object personal information joint coding vectors in the four adjacent candidate object personal information joint coding vectors to obtain a candidate object personal information semantic difference topological distribution matrix composed of multiple personal information semantic difference measurement coefficients; The grade label determination module is used to determine the reasonable grade label of the examination seat arrangement based on the semantic difference topological distribution matrix of the candidate object's personal information.

9. The personnel examination arrangement system based on multidimensional data analysis according to claim 8 is characterized in that: The joint encoding module comprises: A semantic association score vector calculation unit, used to calculate the semantic association score vectors between the first candidate object personal information semantic coding feature vector and each of the candidate object personal information semantic coding feature vectors from the second to fifth candidate object personal information semantic coding feature vectors to obtain a set of candidate object personal information semantic association score vectors; A semantic relevance unit, used to determine a global representation vector of the semantic relevance of the candidate object's personal information based on a set of semantic relevance score vectors of the candidate object's personal information; A correlation optimization factor calculation unit, used to calculate the correlation optimization factor of each candidate object personal information semantic encoding feature vector in the second to fifth candidate object personal information semantic encoding feature vectors based on the candidate object personal information semantic relevance global representation vector to obtain a sequence of candidate object personal information semantic correlation optimization factors; An activation unit, used for inputting the sequence of the semantic association optimization factors of the candidate object personal information into a Softmax activation function to obtain a sequence of the semantic association optimization weight factors of the candidate object personal information; A weighting unit is used to use each candidate object personal information semantic association optimization weight factor in the sequence of candidate object personal information semantic association optimization weight factors as a weight, and weight each candidate object personal information semantic encoding feature vector in the second to fifth candidate object personal information semantic encoding feature vectors respectively to obtain the four adjacent candidate object personal information joint encoding vectors.

10. The personnel examination arrangement system based on multidimensional data analysis according to claim 9 is characterized in that: The semantic association score vector calculation unit is used to: Concatenate the first candidate object personal information semantic encoding feature vector with each candidate object personal information semantic encoding feature vector in the second to fifth candidate object personal information semantic encoding feature vectors to obtain a set of candidate object personal information semantic cascade feature vectors; Each semantic cascade feature vector of candidate object personal information in the set of semantic cascade feature vectors of candidate object personal information is multiplied by the candidate object personal information weight matrix and then added positionally with the candidate object personal information bias vector to obtain the set of semantic association score vectors of candidate object personal information.

Images