A text data recognition processing method and device based on recognition model

Abstract

The present invention discloses a text data recognition processing method and device based on a recognition model. The method comprises: obtaining a job requirement information set and a human resource description information set; the job requirement information set comprises job requirement information; the human resource description information set comprises human resource description information; the job requirement information comprises requirement information of the job in various types of capabilities; the human resource description information comprises evaluation values ​​of personnel in various types of capabilities; pre-processing the job requirement information set and the human resource description information set to obtain a job requirement matrix and a human resource description matrix; obtaining job allocation history information; the job allocation history information represents the allocation correspondence between jobs and personnel; identifying and processing the job requirement matrix, the human resource description matrix and the job allocation history information to obtain a task deployment model; the task deployment model represents the correspondence between personnel types and job types.

Description

Technical Field

[0001] The present invention relates to the field of pattern recognition and data processing, and in particular to a text data recognition processing method and device based on a recognition model. Background Art

[0002] In modern society, the division of labor is becoming more and more refined. How to select talents that match the positions from massive and complex data is a hot technical issue at present. At the same time, with the increasing degree of informatization, all walks of life cannot do without a large amount of text information processing. At present, job assignment and task assignment are mainly recorded through text tables. Especially in task deployment management, deployment tables are a common text form used to record various deployment information, such as equipment configuration, personnel allocation, etc. The traditional way of modifying deployment tables is limited to the form of paper or Word documents. Not only is the modification operation not intuitive, but it is also inconvenient to find deployment information. At the same time, for the text recognition data of task deployment, due to the inaccuracy and inconvenience of operation, it is easy to cause the loss or omission of deployment information, affecting the accuracy and timeliness of deployment management.

[0003] In addition, a large amount of historical data on job deployment is a successful case of job deployment, but there is also a large amount of irrelevant data and noise in the job deployment data. How to effectively mine the job deployment information contained in the historical data and obtain a successful method for job matching is also a problem that needs to be solved urgently.

[0004] How to effectively utilize the historical text data of task deployment and summarize and extract the rules of task deployment from it, so as to improve the efficiency and accuracy of task deployment, is a problem that needs to be solved urgently. Summary of the invention

[0005] The present invention mainly solves the problem of how to effectively utilize the historical text data of task deployment, summarize and extract the law of task deployment, so as to improve the efficiency and accuracy of task deployment. In a first aspect of an embodiment of the present invention, a text data recognition processing method based on a recognition model is disclosed, comprising:

[0006] S1, obtaining a job requirement information set and a human resource description information set; the job requirement information set includes job requirement information; the human resource description information set includes human resource description information;

[0007] S2, preprocessing the job requirement information set and the human resource description information set to obtain a job requirement matrix and a human resource description matrix;

[0008] S3, obtaining historical information of job assignment; the historical information of job assignment indicates the corresponding relationship between job assignment and personnel;

[0009] S4, identifying and processing the job requirement matrix, human resource description matrix and job allocation history information to obtain a task deployment model.

[0010] The preprocessing of the job requirement information set and the human resource description information set to obtain a job requirement matrix and a human resource description matrix includes:

[0011] S21, performing text recognition processing on the job requirement information set and the human resource description information set to obtain natural language text format information;

[0012] S22, preprocessing the natural language text format information to obtain a natural language text set;

[0013] S23, performing text encoding conversion processing on the natural language text set to obtain a job requirement matrix and a human resource description matrix.

[0014] The preprocessing of the natural language text format information to obtain a natural language text set includes:

[0015] S221, performing outlier elimination processing on the natural language text format information to obtain a first text data set;

[0016] S222: Perform boundary checking on the first text data set to obtain a natural language text set.

[0017] The step of performing boundary checking on the first text data set to obtain a natural language text set includes:

[0018] S2221, for each data attribute of the first text data set, preset a corresponding value range; the value range includes an upper limit value of the value range and a lower limit value of the value range;

[0019] S2222, for each data of each data attribute of the first text data set, judging whether the value of the data is within the value range according to the value range of the data attribute; when the data is within the value range, not processing the data; when the data is not within the value range, setting the value of the data to the upper limit value or the lower limit value of the value range that is closest to the data;

[0020] S2223, merging the data of all data attributes obtained in S2222 to obtain a natural language text set.

[0021] The text encoding conversion process is performed on the natural language text set to obtain a job requirement matrix and a human resource description matrix, including:

[0022] S231, performing text encoding conversion processing on each job requirement information in the natural language text set to obtain a corresponding job requirement vector;

[0023] S232, using each job requirement vector as a row vector to construct a job requirement matrix; the row dimension and column dimension of the job requirement matrix are M and N respectively;

[0024] S233, performing text encoding conversion processing on each piece of human resource description information in the natural language text set to obtain a corresponding human resource description vector;

[0025] S234, using each human resource description vector as a row vector, constructing a human resource description matrix; the row dimension and column dimension of the human resource description matrix are P and N respectively.

[0026] The identification and processing of the job requirement matrix, the human resource description matrix and the job allocation history information to obtain a task deployment model includes:

[0027] S41, the job assignment history information is represented as a vector V, the i-th element v of the vector V i Indicates the number of the person assigned to the i-th position;

[0028] S42, classifying the position requirement matrix to obtain a position category information set; the position category information set includes the position category information of each position;

[0029] S43, classifying the human resource description matrix to obtain a human resource category information set; the human resource category information set includes capability range category information of each person;

[0030] S44, using the vector V, establishing a matching correspondence between each element of the position category information set and each element of the human resources category information set;

[0031] S45, based on the matching correspondence, mapping, identifying and calculating the position category information set and the human resource category information set to obtain a task deployment model.

[0032] The classifying process of the job requirement matrix to obtain a job category information set includes:

[0033] S421, obtaining a job category description matrix; the row vector of the job category description matrix is ​​a standard row vector for each job category;

[0034] S422, performing a first distance calculation process on each row vector of the job requirement matrix and the job category description matrix to obtain a first distance vector corresponding to the row vector;

[0035] The expression of the first distance calculation process is:

[0036]

[0037] Among them, a ij is the element in the i-th row and j-th column of the job demand matrix, A1 kj is the element of the kth row and jth column of the job category description matrix, ξ is the preset adjustment factor, d ik represents the kth element of the first distance vector corresponding to the i-th row vector of the job demand matrix;

[0038] S423, determining the serial number value of the minimum element in the first distance vector as the position category value of the row vector;

[0039] S424, merging the job category values ​​of all row vectors of the job requirement matrix to obtain a job category information set.

[0040] The human resource description matrix is ​​classified to obtain a human resource category information set, including:

[0041] S431, obtaining a human resource category description matrix; the row vector of the human resource category description matrix is ​​a standard row vector of each human resource category;

[0042] S432, performing a second distance calculation process on each row vector of the human resource description matrix and the human resource category description matrix to obtain a second distance vector;

[0043] The expression of the second distance calculation process is:

[0044]

[0045] Among them, b ij is the element in the i-th row and j-th column of the human resources description matrix, A2 kj is the element in the kth row and jth column of the human resource category description matrix, ω1 and ω2 are the preset weighting coefficients, c is the preset bias coefficient, and e ik represents the kth element of the second distance vector corresponding to the i-th row vector of the human resource description matrix;

[0046] S433, determining the sequence number value of the maximum element in the second distance vector as the human resources category value of the row vector;

[0047] S434, merging the human resource category values ​​of all row vectors of the human resource description matrix to obtain a human resource category information set.

[0048] In a second aspect of the embodiment of the present application, a text data recognition processing device based on a recognition model is disclosed, the device comprising:

[0049] A memory storing executable program code;

[0050] a processor coupled to the memory;

[0051] The processor calls the executable program code stored in the memory to execute the text data recognition processing method based on the recognition model.

[0052] In a third aspect of an embodiment of the present application, a computer-storable medium is disclosed, wherein the computer-storable medium stores computer instructions, and when the computer instructions are called by a computer, they are used to execute the text data recognition processing method based on the recognition model.

[0053] The beneficial effects of the present invention are:

[0054] The present invention effectively utilizes the historical text data of task deployment, summarizes and extracts the law of task deployment, thereby improving the efficiency and accuracy of task deployment.

[0055] The present invention suppresses a large amount of irrelevant data and noise in the current task post deployment data, thereby effectively mining the post deployment information contained in the historical data, successfully obtaining a partial post model, and improving the efficiency of post deployment. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] Figure 1 It is a flow chart for implementing the method of the present invention. DETAILED DESCRIPTION

[0057] In order to better understand the content of the present invention, an embodiment is given here.

[0058] Figure 1 It is a flow chart for implementing the method of the present invention.

[0059] In a first aspect of an embodiment of the present application, a text data recognition and processing method based on a recognition model is disclosed, comprising:

[0060] S1, obtaining a job requirement information set and a human resource description information set; the job requirement information set includes job requirement information; the human resource description information set includes human resource description information; the job requirement information includes the job requirement information in various types of capabilities; the human resource description information includes the evaluation value of the personnel in various types of capabilities;

[0061] S2, preprocessing the job requirement information set and the human resource description information set to obtain a job requirement matrix and a human resource description matrix;

[0062] S3, obtaining historical information of job assignment; the historical information of job assignment indicates the corresponding relationship between job assignment and personnel;

[0063] S4, identifying and processing the job requirement matrix, the human resource description matrix and the job allocation history information to obtain a task deployment model; the task deployment model represents the correspondence between personnel types and job types.

[0064] The preprocessing of the job requirement information set and the human resource description information set to obtain a job requirement matrix and a human resource description matrix includes:

[0065] S21, performing text recognition processing on the job requirement information set and the human resource description information set to obtain natural language text format information;

[0066] S22, preprocessing the natural language text format information to obtain a natural language text set;

[0067] S23, performing text encoding conversion processing on the natural language text set to obtain a job requirement matrix and a human resource description matrix;

[0068] The preprocessing of the natural language text format information to obtain a natural language text set includes:

[0069] S221, performing outlier elimination processing on the natural language text format information to obtain a first text data set;

[0070] S222: Perform boundary checking on the first text data set to obtain a natural language text set.

[0071] The step of performing boundary checking on the first text data set to obtain a natural language text set includes:

[0072] S2221, for each data attribute of the first text data set, preset a corresponding value range; the value range includes an upper limit value of the value range and a lower limit value of the value range;

[0073] S2222, for each data of each data attribute of the first text data set, judging whether the value of the data is within the value range according to the value range of the data attribute; when the data is within the value range, not processing the data; when the data is not within the value range, setting the value of the data to the upper limit value or the lower limit value of the value range that is closest to the data;

[0074] S2223, after completing S2222 for all data attributes, a natural language text set is obtained.

[0075] The text encoding conversion process is performed on the natural language text set to obtain a job requirement matrix and a human resource description matrix, including:

[0076] S231, performing text encoding conversion processing on each job requirement information in the natural language text set to obtain a corresponding job requirement vector;

[0077] S232, using each job requirement vector as a row vector to construct a job requirement matrix; the row dimension and column dimension of the job requirement matrix are M and N respectively;

[0078] S233, performing text encoding conversion processing on each piece of human resource description information in the natural language text set to obtain a corresponding human resource description vector;

[0079] S234, using each human resource description vector as a row vector, constructing a human resource description matrix; the row dimension and column dimension of the human resource description matrix are P and N respectively.

[0080] The identification and processing of the job requirement matrix, the human resource description matrix and the job allocation history information to obtain a task deployment model includes:

[0081] S41, the job assignment history information is represented as a vector V, the i-th element v of the vector V i Indicates the number of the person assigned to the i-th position;

[0082] S42, classifying the position requirement matrix to obtain a position category information set; the position category information set includes the position category information of each position;

[0083] S43, classifying the human resource description matrix to obtain a human resource category information set; the human resource category information set includes capability range category information of each person;

[0084] S44, using the vector V, establishing a matching correspondence between each element of the position category information set and each element of the human resources category information set;

[0085] S45, based on the matching correspondence, mapping, identifying and calculating the position category information set and the human resource category information set to obtain a task deployment model.

[0086] The classifying process of the job requirement matrix to obtain a job category information set includes:

[0087] S421, obtaining a job category description matrix; the row vector of the job category description matrix is ​​a standard row vector for each job category;

[0088] S422, performing a first distance calculation process on each row vector of the job requirement matrix and the job category description matrix to obtain a first distance vector corresponding to the row vector;

[0089] The expression of the first distance calculation process is:

[0090]

[0091] Among them, a ij is the element in the i-th row and j-th column of the job demand matrix, A1 kj is the element of the kth row and jth column of the job category description matrix, ξ is the preset adjustment factor, d ik represents the kth element of the first distance vector corresponding to the i-th row vector of the job demand matrix;

[0092] S423, determining the serial number value of the minimum element in the first distance vector as the position category value of the row vector;

[0093] S424, merging the job category values ​​of all row vectors of the job requirement matrix to obtain a job category information set;

[0094] The human resource description matrix is ​​classified to obtain a human resource category information set, including:

[0095] S431, obtaining a human resource category description matrix; the row vector of the human resource category description matrix is ​​a standard row vector of each human resource category;

[0096] S432, performing a second distance calculation process on each row vector of the human resource description matrix and the human resource category description matrix to obtain a second distance vector;

[0097] The expression of the second distance calculation process is:

[0098]

[0099] Among them, b ij is the element in the i-th row and j-th column of the human resources description matrix, A2 kj is the element in the kth row and jth column of the human resource category description matrix, ω1 and ω2 are the preset weighting coefficients, c is the preset bias coefficient, and e ik represents the kth element of the second distance vector corresponding to the i-th row vector of the human resource description matrix;

[0100] S433, determining the sequence number value of the maximum element in the second distance vector as the human resources category value of the row vector;

[0101] S434, merging the human resource category values ​​of all row vectors of the human resource description matrix to obtain a human resource category information set.

[0102] Based on the matching correspondence, the job category information set and the human resource category information set are mapped, identified and calculated to obtain a task deployment model, which is to use the job category information set as the dependent variable y and the human resource category information set as the independent variable x, and use the matching correspondence to establish a pairwise matching relationship between the elements of x and y, and perform function fitting on the matching relationship to obtain a fitting function as the task deployment model; the input variable of the task deployment model is the human resource category value, and the output variable of the task deployment model is the job category value. The function fitting can adopt linear fitting or nonlinear fitting.

[0103] The mapping recognition calculation process may be to use a function approximation method to perform curve fitting on x and y, and may use a best consistent linear approximation method. The best consistent approximation polynomial f(x) is expressed as:

[0104] f(x)=α P1 (x) P1 +α P1-1 (x) P1-1 +…+α2(x) 2 +α1(x)+α0,

[0105] Where P1 is the order of the best consistent approximation polynomial f(x), α0, α1, α2, …, α P1 are the coefficients of the best consistent approximation polynomial f(x);

[0106] The job allocation history information includes a vector V, whose i-th element v i Indicates the number of the person assigned to the i-th position;

[0107] The text encoding conversion may be to set corresponding values, such as 8, 6, 4, and 2, for strong, medium, general, and weak in the demand information, or to set corresponding values, such as 10, 8, 6, 4, 2, and 0, for leading, excellent, proficient, general, understood, and not possessed in the evaluation values;

[0108] The text recognition process can be implemented using text recognition software or program.

[0109] The row vector of the human resource description matrix is ​​a human resource description vector;

[0110] The text vector conversion process can be implemented using the word2vector model;

[0111] The outlier elimination includes filling missing values, smoothing noise data, smoothing or deleting outlier points; the smoothed noise data is first determined to obtain noise data, and then the noise data is smoothed according to the data before and after the noise data; the noise data is a value whose value is less than the detection sensitivity of the sensor of the observed data, or greater than the measurement upper limit of the sensor of the observed data. The outlier point can be determined by the Kalman filtering method. The filling value of the missing value can be determined by averaging the measured values ​​within a certain sampling interval before and after the missing value.

[0112] The capability types include electrical professional capability, electronic information professional capability, control professional capability, word processing capability, structural design capability, etc. The requirement information can be strong, medium, general, or weak; the evaluation value can be leading, excellent, proficient, general, understood, or not possessed;

[0113] The method of using the vector V to establish a matching correspondence between each element of the job category information set and each element of the human resource category information set includes: taking the serial number value of each element in the vector V as a first variable and the value of the element as a second variable, taking the element with the serial number of the first variable from the job category information set and taking the element with the serial number of the second variable from the human resource category information set, establishing a corresponding relationship between the two elements, repeating the above operation for each element in the vector V, and establishing a matching correspondence between each element of the job category information set and each element of the human resource category information set.

[0114] In a second aspect of the embodiment of the present application, a text data recognition processing device based on a recognition model is disclosed, the device comprising:

[0115] A memory storing executable program code;

[0116] a processor coupled to the memory;

[0117] The processor calls the executable program code stored in the memory to execute the text data recognition processing method based on the recognition model.

[0118] In a third aspect of an embodiment of the present application, a computer-storable medium is disclosed, wherein the computer-storable medium stores computer instructions, and when the computer instructions are called by a computer, they are used to execute the text data recognition processing method based on the recognition model.

[0119] The above is only an embodiment of the present application and is not intended to limit the present application. For those skilled in the art, the present application may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included in the scope of the claims of the present application.

Claims

1. A text data recognition and processing method based on a recognition model, characterized in that: include: S1, obtaining a job requirement information set and a human resource description information set; the job requirement information set includes job requirement information; the human resource description information set includes human resource description information; S2, preprocessing the job requirement information set and the human resource description information set to obtain a job requirement matrix and a human resource description matrix; S3, obtaining historical information of job assignment; the historical information of job assignment indicates the corresponding relationship between job assignment and personnel; S4, identifying and processing the job requirement matrix, the human resource description matrix and the job allocation history information to obtain a task deployment model; The preprocessing of the job requirement information set and the human resource description information set to obtain a job requirement matrix and a human resource description matrix includes: S21, performing text recognition processing on the job requirement information set and the human resource description information set to obtain natural language text format information; S22, preprocessing the natural language text format information to obtain a natural language text set; S23, performing text encoding conversion processing on the natural language text set to obtain a job requirement matrix and a human resource description matrix; The preprocessing of the natural language text format information to obtain a natural language text set includes: S221, performing outlier elimination processing on the natural language text format information to obtain a first text data set; S222, performing boundary check processing on the first text data set to obtain a natural language text set; The text encoding conversion process is performed on the natural language text set to obtain a job requirement matrix and a human resource description matrix, including: S231, performing text encoding conversion processing on each job requirement information in the natural language text set to obtain a corresponding job requirement vector; S232, using each job requirement vector as a row vector to construct a job requirement matrix; the row dimension and column dimension of the job requirement matrix are M and N respectively; S233, performing text encoding conversion processing on each piece of human resource description information in the natural language text set to obtain a corresponding human resource description vector; S234, using each human resource description vector as a row vector to construct a human resource description matrix; the row dimension and column dimension of the human resource description matrix are P and N respectively; The identification and processing of the job requirement matrix, the human resource description matrix and the job allocation history information to obtain a task deployment model includes: S41, the job assignment history information is represented as a vector V, the i-th element v of the vector V i Indicates the number of the person assigned to the i-th position; S42, classifying the position requirement matrix to obtain a position category information set; the position category information set includes the position category information of each position; S43, classifying the human resource description matrix to obtain a human resource category information set; the human resource category information set includes capability range category information of each person; S44, using the vector V, establishing a matching correspondence between each element of the position category information set and each element of the human resources category information set; S45, based on the matching correspondence, performing mapping, identification and calculation processing on the position category information set and the human resource category information set to obtain a task deployment model; The classifying process of the job requirement matrix to obtain a job category information set includes: S421, obtaining a job category description matrix; the row vector of the job category description matrix is ​​a standard row vector for each job category; S422, performing a first distance calculation process on each row vector of the job requirement matrix and the job category description matrix to obtain a first distance vector corresponding to the row vector; The expression of the first distance calculation process is: Among them, a ij is the element in the i-th row and j-th column of the job demand matrix, A1 kj is the element of the kth row and jth column of the job category description matrix, ξ is the preset adjustment factor, d ik represents the kth element of the first distance vector corresponding to the i-th row vector of the job demand matrix; S423, determining the serial number value of the minimum element in the first distance vector as the position category value of the row vector; S424, merging the job category values ​​of all row vectors of the job requirement matrix to obtain a job category information set.

2. The text data recognition processing method based on the recognition model according to claim 1, characterized in that: The step of performing boundary checking on the first text data set to obtain a natural language text set includes: S2221, for each data attribute of the first text data set, preset a corresponding value range; the value range includes an upper limit value of the value range and a lower limit value of the value range; S2222, for each data of each data attribute of the first text data set, judging whether the value of the data is within the value range according to the value range of the data attribute; when the data is within the value range, not processing the data; when the data is not within the value range, setting the value of the data to the upper limit value or the lower limit value of the value range that is closest to the data; S2223, merging the data of all data attributes obtained in S2222 to obtain a natural language text set.

3. The text data recognition processing method based on the recognition model as claimed in claim 2, characterized in that: The human resource description matrix is ​​classified to obtain a human resource category information set, including: S431, obtaining a human resource category description matrix; the row vector of the human resource category description matrix is ​​a standard row vector of each human resource category; S432, performing a second distance calculation process on each row vector of the human resource description matrix and the human resource category description matrix to obtain a second distance vector; The expression of the second distance calculation process is: Among them, b ij is the element in the i-th row and j-th column of the human resources description matrix, A2 kj is the element in the kth row and jth column of the human resource category description matrix, ω1 and ω2 are the preset weighting coefficients, and e ik represents the kth element of the second distance vector corresponding to the i-th row vector of the human resource description matrix; S433, determining the sequence number value of the maximum element in the second distance vector as the human resources category value of the row vector; S434, merging the human resource category values ​​of all row vectors of the human resource description matrix to obtain a human resource category information set.

4. A text data recognition processing device based on a recognition model, characterized in that: The device comprises: A memory storing executable program code; a processor coupled to the memory; The processor calls the executable program code stored in the memory to execute the text data recognition processing method based on the recognition model as described in any one of claims 1 to 3.

5. A computer storable medium, characterized in that: The computer storable medium stores computer instructions, and when the computer instructions are called by a computer, they are used to execute the text data recognition processing method based on the recognition model as described in any one of claims 1 to 3.

Images