Talent enabling method and system based on big data

Abstract

The invention discloses a talent enabling method and system based on big data, and relates to the technical field of information management, and the method comprises the steps: S1, obtaining the behavior characteristics of a to-be-distributed person, and constructing a person behavior characteristic set; s2, constructing a first test item corresponding to the personnel behavior feature set, wherein the first test item comprises demand features corresponding to the behavior features; performing one-to-one correspondence on the demand characteristics and the behavior characteristics, and calculating to obtain a first matching coefficient related to the demand characteristics; obtaining a first matching result according to the first matching coefficient; s3, extracting a second test item from the first test items according to the first matching result; s4, matching the second test item with the behavior characteristics, and identifying a second matching coefficient related to the behavior characteristics; obtaining a second matching result according to the second matching coefficient; and the personnel matching precision and efficiency are improved.

Description

Technical Field

[0001] The present invention relates to the field of information management technology, and in particular to a talent empowerment method and system based on big data. Background Art

[0002] In the process of personnel allocation, traditional matching methods often rely on manual experience and simple indicator comparison, which has problems such as low matching accuracy and low efficiency. With the development of big data and artificial intelligence technology, more and more companies are trying to use algorithms and models for personnel matching. However, these algorithms and models are often too complex to be quickly promoted in practical applications.

[0003] For example, Chinese patent publication No. CN115471377A discloses a talent empowerment management method and system, which includes the following steps: pushing multiple types of operation instruction streams to the operator, obtaining the operator's physiological information during the execution of the operation instruction stream, analyzing and determining the operator's indicator information based on the physiological information obtained at different operation times, and determining the operator's talent type based on the indicator information and the talent type data stored in the talent database; the talent empowerment management system includes: a display unit, which is used to provide the operator with operation instructions for training; a data connection unit, which is used to collect physiological information from the wearable device worn by the operator on the one hand, and to send the physiological information and the operation instructions associated with time to a behavior data analysis agency on the other hand; the behavior data analysis agency determines the talent type of the operator based on the data transmitted by the data connection unit, and outputs the matching work of the operator based on the talent type determined by the behavior data analysis agency.

[0004] For example, Chinese patent publication number CN115392804A discloses a talent empowerment method and system based on big data, wherein the talent empowerment system includes: a data verification module, which maintains the unstructured data stored in the general-purpose computer and monitors the equipment associated with the general-purpose computer; a talent management module, which determines the personnel characteristics of the staff and sends them to the data verification module; a task management module, which assigns work item tasks related to the maintenance of the general-purpose computer and its associated equipment to the staff through the data verification module. The present invention also relates to a talent empowerment method based on big data, including a data verification module that maintains the unstructured data stored in the general-purpose computer and monitors the equipment associated with the general-purpose computer; the talent management module is coupled to the operation capability extreme value test module to determine the personnel characteristics of the staff through the operation capability extreme value test module; and the task management module assigns work item tasks related to the maintenance of the general-purpose computer and its associated equipment to the staff through the data verification module.

[0005] In the prior art, personnel allocation tends to verify the push characteristics of personnel at work, and analyzes and processes personnel empowerment based on these data. However, when matching, these methods ignore the corresponding relationship between behavior and demand, which makes it easy to ignore the differences between different behaviors or needs when empowering personnel, thereby reducing the accuracy and efficiency of processing. Summary of the invention

[0006] This application solves the technical problem of low accuracy in personnel empowerment matching in the prior art by providing a talent empowerment method and system based on big data, thereby improving the matching accuracy and efficiency.

[0007] This application provides a talent empowerment method based on big data, including: S1, obtain the behavioral characteristics of the personnel to be assigned and construct a set of personnel behavioral characteristics.

[0008] S2, constructing a first test item corresponding to a set of personnel behavior characteristics, wherein the first test item includes demand characteristics corresponding to the behavior characteristics; matching the demand characteristics with the behavior characteristics one by one, and calculating a first matching coefficient related to the demand characteristics; and obtaining a first matching result based on the first matching coefficient.

[0009] S3: extract a second test item from the first test item according to the first matching result.

[0010] S4, matching the second test item with the behavior feature, identifying a second matching coefficient related to the behavior feature; and obtaining a second matching result according to the second matching coefficient.

[0011] S5, performing a difference comparison between the first matching result and the second matching result, obtaining a difference matching coefficient corresponding to the first matching result and the second matching result, and adjusting the allocation of personnel according to the difference matching coefficient.

[0012] A talent empowerment system based on big data, comprising: The behavior feature collection module is used to obtain the behavior features of the personnel to be assigned and construct a set of personnel behavior features.

[0013] The test item construction module is used to construct a first test item and a second test item corresponding to the personnel behavior feature set according to the personnel behavior feature set.

[0014] The matching calculation module is used to calculate the first matching coefficient, the second matching coefficient, and the final difference matching coefficient.

[0015] The allocation strategy adjustment module is used to adjust the allocation strategy for personnel according to the difference matching coefficient and output the final allocation result.

[0016] One or more technical solutions provided in this application have at least the following technical effects or advantages: By constructing a set of personnel behavior characteristics and designing the first test items and the second test items, the demand characteristics that best match the personnel behavior characteristics are gradually screened out, thereby improving the matching accuracy; using decision trees, cluster analysis and other algorithms, the test items are quickly processed and screened, thereby improving the efficiency of personnel allocation; by calculating the matching coefficient and the difference matching coefficient, the personnel allocation strategy can be dynamically adjusted to enhance the adaptability to different personnel and job requirements; by outputting the first matching result and the second matching result, as well as the final difference matching coefficient, strong data support is provided for decision makers, enabling them to make more reasonable allocation decisions. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 The present invention is a flowchart of a talent empowerment method based on big data.

[0018] Figure 2 This is a flow chart of step S2 of a talent empowerment method based on big data of the present invention.

[0019] Figure 3 This is a flow chart of step S4 of a talent empowerment method based on big data of the present invention.

[0020] Figure 4 This is a flow chart of step S5 of a talent empowerment method based on big data of the present invention.

[0021] Figure 5 This is a flow chart of step S53 of a talent empowerment method based on big data of the present invention.

[0022] Figure 6 This is a system framework diagram of a talent empowerment system based on big data in the present invention. DETAILED DESCRIPTION

[0023] To facilitate the understanding of the present invention, the present application will be described more comprehensively below with reference to the relevant drawings; the drawings show preferred embodiments of the present invention, but the present invention can be implemented in many different forms and is not limited to the embodiments described herein; on the contrary, the purpose of providing these embodiments is to enable a more thorough and comprehensive understanding of the disclosed content of the present invention.

[0024] It should be noted that the terms “vertical”, “horizontal”, “up”, “down”, “left”, “right” and similar expressions used in this document are only for illustrative purposes and do not represent the only implementation method.

[0025] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by technicians in the technical field to which the present invention belongs; the terms used in the specification of the present invention are only for the purpose of describing specific embodiments and are not intended to limit the present invention; the term "and / or" used herein includes any and all combinations of one or more related listed items.

[0026] Generally speaking, talent empowerment is to match the behavioral characteristics of the current personnel to be assigned with the demand characteristics corresponding to the current project, and calculate the personnel score by combining multiple behavioral characteristics with demand characteristics, or use the weighted sum of the ratio of the personnel score to the overall score and the score value for a single characteristic as the basis for allocation at this time; however, this processing method has certain problems. First, it is necessary to make a decision on the demand characteristics, construct relevant analysis results corresponding to the demand characteristics, and analysis results of the behavioral characteristics to ensure that the behavioral characteristics and demand characteristics are one-to-one corresponding at this time, and it is necessary to verify the deviations in the corresponding data to identify whether there are any non-corresponding or over-corresponding deviations in the current decision, so as to prevent the situation where the score is high but the actual match with the demand characteristics of the current project is low.

[0027] In the present invention, the behavioral characteristics of all personnel are first matched, and the behavioral characteristics are used as a data set to construct a decision tree corresponding to the demand characteristics to obtain a shortest decision path. At the same time, according to the content presented on this decision tree, the shortest decision path is matched twice with the behavioral characteristics of the relevant personnel in turn to verify whether the distribution of the shortest decision path is consistent with the behavioral characteristics. At the same time, it is also necessary to verify the correlation between the behavioral characteristics and the demand characteristics to complete the realization of personnel empowerment; the behavioral characteristics described above include: work performance, skill level, learning and development ability, collaboration ability, innovation ability, leadership, adaptability and flexibility, professionalism, and cultural fit. The behavioral characteristics will convert these contents into specific feature values ​​to represent the relevant situation of the current personnel to be assigned, and will also convert the current project or other types of tasks into multiple demand characteristics to describe the currently required feature values, and these two sets of feature values ​​will be verified accordingly to achieve relevant matching and comparison.

[0028] like Figure 1 As shown, the present application provides a talent empowerment method based on big data, including: S1, obtain the behavioral characteristics of the personnel to be assigned and construct a set of personnel behavioral characteristics.

[0029] S2, constructing a first test item corresponding to a set of personnel behavior characteristics, wherein the first test item includes demand characteristics corresponding to the behavior characteristics; matching the demand characteristics with the behavior characteristics one by one, and calculating a first matching coefficient related to the demand characteristics; and obtaining a first matching result based on the first matching coefficient.

[0030] S3: extract a second test item from the first test item according to the first matching result.

[0031] S4, matching the second test item with the behavior feature, identifying a second matching coefficient related to the behavior feature; and obtaining a second matching result according to the second matching coefficient.

[0032] S5, performing a difference comparison between the first matching result and the second matching result, obtaining a difference matching coefficient corresponding to the first matching result and the second matching result, and adjusting the allocation of personnel according to the difference matching coefficient.

[0033] The first matching result mentioned above is essentially to construct a decision tree according to the corresponding features of the personnel behavior feature set and the first demand item, wherein the Gini impurity and partial correlation coefficient of the behavior features and the demand features are calculated at both ends of the tree respectively, and the specific comparison result content at this time is identified according to the output content of the decision tree; at the same time, it is necessary to verify the Gini impurity and partial correlation coefficient calculated at this time, and use these two values ​​to identify the similar items in the behavior features and demand features at this time to obtain the suitability score of the current personnel, and at the same time verify whether there is a correlation between the similar items at this time to complete the identification of the first matching result.

[0034] like Figure 2 As shown, the implementation method of the first matching result includes: S21, using demand characteristics and behavior characteristics as input to construct a decision tree, each node of the decision tree includes at least a set of demand characteristics and behavior characteristics, and a set of demand characteristics and behavior characteristics represented at this time means that there is at least one behavior feature and one demand feature corresponding to the behavior feature.

[0035] S22, calculate the Gini impurity and partial correlation coefficient of each node in the decision tree in turn; determine the division method of the child nodes of the decision tree according to the values ​​of the Gini impurity and partial correlation coefficient on each node.

[0036] S23, continuously iterating the decision tree, outputting the iterated decision tree, and obtaining a first matching result.

[0037] When calculating the Gini impurity and partial correlation coefficient in step S22, the demand characteristics and behavior characteristics in the divided sub-nodes are selected according to the partial correlation coefficients between the current demand characteristics and behavior characteristics, and the Gini impurity of the divided sub-nodes is calculated. When the Gini impurity calculation result is optimal, the average value of the corresponding partial correlation coefficients is used as the first matching coefficient. When the partial correlation coefficients of the demand characteristics and the behavior characteristics are greater than the first matching coefficient, the corresponding behavior characteristics and demand characteristics are used as the data of the sub-nodes in the decision tree to obtain the division method of the sub-nodes of the decision tree.

[0038] The Gini impurity is calculated by subtracting 1 from the sum of the corresponding category proportions in the current node to get a difference to evaluate the Gini impurity of each node. When multiple sub-nodes are divided into multiple categories, the weighted sum of the Gini impurities of all sub-nodes is calculated. The weight used is the ratio of the number of data in the sub-node to the number of data in the current node to calculate the node division at this time. The optimal Gini impurity calculation result selected at this time means that after the current node is divided, the Gini impurity of the current node reaches the maximum value, that is, when the Gini impurities of the root node and the sub-nodes of the root node are both optimal, the biased phase of the behavioral characteristics and demand characteristics of the root node and the sub-nodes of the root node in the decision tree at this time is The average value of the relationship number is used as the first matching coefficient at this time, and the demand characteristics and behavior characteristics in the leaf nodes are selected based on the first matching coefficient, so as to complete the final division of the current decision tree. The result after division can be output as the first matching result after identification; the purpose of subsequent iterations is to determine the specific value of the first matching coefficient, so as to complete the division of the decision tree; at this time, the first matching coefficient is used for division in the form of when the partial correlation coefficient of the behavior characteristics and demand characteristics of the child nodes of the current node is greater than the first matching coefficient, the corresponding behavior characteristics and demand characteristics are divided into the child nodes of the current node, and at the same time, the sum of the Gini impurities of the child nodes of the current node must be guaranteed to be the maximum value of all division methods.

[0039] The partial correlation coefficient is the calculation of two values ​​given an additional value to obtain the corresponding partial correlation coefficient; at this time, the first conditional probability corresponding to the demand characteristics and behavior characteristics is introduced in the calculation, that is, the probability of occurrence of the behavior characteristics under the condition that the current demand characteristics occur; when calculating the partial correlation coefficient, the characteristic value of the demand characteristic, the characteristic value of the behavior characteristic and the first conditional probability are used as input, and the Pearson correlation coefficient between the demand characteristic and the first conditional probability, the Pearson correlation coefficient between the behavior characteristic and the first conditional probability, and the Pearson correlation coefficient between the demand characteristic and the behavior characteristic are calculated in turn; the three values ​​are combined to obtain the partial correlation coefficient. When calculating the Pearson correlation coefficient, the current input value is subtracted from the average value of the corresponding historical data to calculate the three Pearson correlation coefficients at this time.

[0040] The first matching result finally output includes not only the behavior characteristics and demand characteristics, but also the corresponding Gini impurity and partial correlation coefficient values, as well as the first matching coefficient value under specific Gini impurity and partial correlation coefficient.

[0041] In step S3, when extracting the second test item from the first test item, it is necessary to calculate the cosine similarity of the current first test item and use the cosine similarity to select the second test item at this time. The first matching result can be optimized based on these items, making the matching of subsequent project candidates more reasonable.

[0042] For example, the implementation of step S3 may include: obtaining a feature vector of the first test item, and calculating the cosine similarity between the current first test item and other first test items in turn; clustering all the first test items according to the value of the cosine similarity to generate multiple clusters, each cluster is set according to the variance of the cosine similarity of the first test item in the cluster, and the clusters are continuously iterated. At this time, the iteration method is to merge the clusters according to the total variance within the clusters, and each time the two clusters with the smallest increase in the total variance within the cluster are selected to generate multiple clusters of similar size to describe the different clusters. Under the requirements of the same demand characteristics, the current combination of demand characteristics, the total variance within the cluster represents the variance of all data in a cluster, that is, the variance value calculated by the cosine similarity within the current cluster; the cluster with the largest value after iteration is used as the second test item; selecting the cluster with the largest value is essentially selecting a group of data with the largest cosine similarity variance. This group of data can represent the changes in the results brought about by different distributions of demand characteristics, to describe the specific behavioral characteristics and demand characteristics in the currently selected scenario, and finally complete the allocation and adjustment of current personnel.

[0043] When using the first test item and the second test item, in the process of empowering personnel, the personnel will tend to identify the processing speed, reaction speed, problem handling scope, completion time, emotions, memory ability, the interval time for each task processing, and the continuous processing time of tasks in the behavioral characteristics as the main behavioral characteristics to be identified at this time. These behavioral characteristics will indicate the specific capabilities of personnel in the process of processing related projects and tasks, and these capabilities will be converted into multiple indicator values ​​for calculation. At the same time, the demand characteristics will set data corresponding to the behavioral characteristics. These data will indicate the basic requirements for completing the current projects and tasks. For example, the scores of reaction speed and problem handling scope are set within the range of 0 to 1. The current user scores are 0.6 and 0.7. Assuming that the demand characteristics need to be above 0.6 and 0.8, it is necessary to verify the behavioral characteristics of all current personnel, and select the behavioral characteristics that best meet the current demand characteristics based on the comprehensive situation of the behavioral characteristics, and set the personnel corresponding to the behavioral characteristics to corresponding jobs.

[0044] like Figure 3 As shown, in one embodiment of the present invention, the implementation method of step S4 includes: S41, comparing the second test item with the behavioral feature, and calculating the co-occurrence probability of the second test item and the behavioral feature; the co-occurrence probability here is used to identify what form of behavioral feature is likely to appear when the second test item is selected, and the ratio of the amount of data in which the corresponding second test item and behavioral feature appear together in the historical data to the total number is used as the common probability used here.

[0045] S42, determining the correlation between the behavior features under the corresponding values ​​of the common occurrence probability, and calculating the feature matching value between the behavior features.

[0046] S43, calculating a second matching coefficient based on the common occurrence probability and the feature matching value of the second test item and the behavior feature.

[0047] Since the currently set feature matching value is to identify the situation under the co-occurrence probability of the current corresponding second test item and the behavioral feature, it is necessary to set the corresponding feature matching value according to the value of the correlation between the current behavioral features, and correspond this feature matching value to the corresponding co-occurrence probability, so as to obtain a corresponding coefficient value. This coefficient value is used to select the required second matching result, as well as the corresponding behavioral feature and the second test item, so as to assist the obtained first test item and complete the screening of the current item.

[0048] The feature matching value can be expressed as: obtaining the correlation coefficient corresponding to the behavior feature, and calculating the feature matching value.

[0049] ; in, represents the feature matching value between behavior feature i and behavior feature j, It represents the correlation coefficient between behavior feature i and behavior feature j. The correlation coefficient at this time is calculated using the Pearson correlation coefficient. The values ​​of these two behavior features are used as input and compared with the standard values ​​of the behavior feature values ​​in the historical data in turn to obtain the corresponding correlation coefficient. It represents the standard value of the correlation coefficient, which adopts the average value of the correlation coefficients between all behavioral features in the historical data.

[0050] When the feature matching value is 1 at this time, it means that the correlation between the currently acquired behavior features is high and they can be easily regarded as a group of features for calculation. These obtained behavior features are processed according to the currently selected situation, so that the currently required behavior features can be obtained.

[0051] After the feature matching value is obtained, the obtained feature matching value is compared with the co-occurrence probability of the second test item and the behavior feature to obtain a second matching coefficient.

[0052] The second matching coefficient is expressed as: ; in, represents the second matching coefficient of behavior feature i, represents the feature matching value between behavior feature i and behavior feature j, represents the co-occurrence probability of behavior feature i and the second test item t, Indicates the adjustment coefficient, the value range of the adjustment coefficient is 0 to 1; Indicates the number of behavior features, where the number of behavior features i and behavior features j is the same, and the value range of i and j is 1 to n; Indicates the number of test items, The value range of is 1 to m; the second matching coefficient obtained at this time can represent the second matching coefficient corresponding to each behavior feature.

[0053] S44, screening the current behavior feature according to the obtained second matching coefficient, and outputting the screened behavior feature and the demand feature corresponding to the behavior feature as a second matching result; the second matching coefficient of the screened behavior feature needs to be greater than the average value of the second matching coefficient in the historical data.

[0054] Compared with the first matching result, the second matching result tends to select behavioral features whose comprehensive values ​​achieve higher values ​​among similar test items, and outputs these behavioral features to obtain the required second matching result.

[0055] When implementing the current step S5, it is necessary to compare the differences between the corresponding first matching result and the second matching result, and quantify these data contents according to the specific behavioral characteristics, demand characteristics and corresponding calculated values, so as to distinguish whether the personnel and projects divided at this time can be matched accordingly. For example, the mean value, variance, etc. of the corresponding features in the two results can be compared, or the feature recognition coefficient of variation can be used to complete the implementation of the difference results.

[0056] like Figure 4 As shown, step S5 includes: S51, based on the first matching result and the second matching result, respectively obtaining the behavior variation coefficient and the demand variation coefficient related to the first matching result and the second matching result.

[0057] S52, according to the obtained behavior variation coefficient and demand variation coefficient, identify the intersection-and-union ratio of the first matching result and the second matching result under the corresponding values ​​of the behavior variation coefficient and the demand variation coefficient; the intersection-and-union ratio represents the ratio of the intersection and union of the data in the first matching result and the data in the second matching result under the same values ​​of the behavior variation coefficient and the demand variation coefficient, so as to determine the distribution of the data in the currently obtained results when there are corresponding differences in the behavior characteristics and the demand characteristics.

[0058] S53, combining the obtained intersection-and-union ratio, behavior variation coefficient and demand variation coefficient to calculate a difference matching coefficient.

[0059] The corresponding behavior variation coefficient is expressed as the reciprocal mean of the distance between the current behavior feature and the nearest neighbor behavior feature. The ratio of the reciprocal mean corresponding to the current behavior feature to the sum of the reciprocal mean corresponding to the total behavior feature is used as the behavior variation coefficient. At this time, the distance expressed by the calculation is the behavior feature represented by a vector to calculate the corresponding distance; for the demand variation coefficient, it is the ratio of the moving average of the current demand feature and the adjacent demand feature to the standard deviation of the total demand feature. This ratio is used as the demand variation coefficient.

[0060] like Figure 5 As shown, the calculation method of the difference matching coefficient in step S53 is expressed as follows: S531, the behavior characteristics and demand characteristics in the first matching result are matched one by one with the behavior variation coefficient, the demand variation coefficient and the intersection and union ratio, and the difference coefficient corresponding to the current behavior characteristic is calculated.

[0061] The coefficient of variation is expressed as: ; in, represents the coefficient of variation, represents the behavioral coefficient of variation of the first matching result, represents the behavioral coefficient of variation of the second matching result, represents the coefficient of variation of the demand for the first matching result, represents the demand variation coefficient of the second matching result, represents the intersection-and-union ratio of the first matching result and the second matching result under the corresponding values ​​of the behavior variation coefficient and the demand variation coefficient; represents the weight coefficient of the behavior variation coefficient, represents the weight coefficient of the demand variation coefficient, Indicates the weight coefficient of the intersection-and-union ratio; the weight coefficients set here can be set to 0.4, 0.3, 0, and 3 in sequence. The difference coefficient calculated at the same time represents the value corresponding to a set of behavioral characteristics and demand characteristics. When processing all behavioral characteristics and demand characteristics, this value can be adjusted to reach a value that can comprehensively represent all current behavioral characteristics and demand characteristics.

[0062] S532, obtaining a first matching coefficient and a second matching coefficient corresponding to the current behavior feature, and using the first matching coefficient and the second matching coefficient to adjust the difference coefficient of the current behavior feature to obtain a final output difference matching coefficient.

[0063] The difference matching coefficient can be expressed as: ; in, represents the difference matching coefficient, represents the first matching coefficient, represents the second matching coefficient, Represents an exponential constant.

[0064] The current difference matching coefficient can indicate the form in which the current behavioral characteristics need to be adjusted, and ultimately obtain a coefficient value that can better express the current needs. According to this coefficient value, the personnel at this time are selected to complete the allocation of personnel empowerment.

[0065] In one embodiment of the present invention, when setting the difference matching coefficient, the difference matching coefficient may also be represented by a combination of the first matching coefficient and the second matching coefficient.

[0066] For example, the difference matching coefficient ; This form is used to express the specific value of the difference matching coefficient, and this situation mainly strengthens the influence of the second matching result on the overall selection after the second matching result is obtained, and selects the currently suitable coefficient value based on this relevant influence to represent the specific situation of the current staff member.

[0067] Finally, allocation is made according to the value of the difference matching coefficient. When the value of the difference matching coefficient is large, it means that there are significant differences in behavioral characteristics and demand characteristics between the first matching result and the second matching result; this may mean that the gap between the behavioral variation coefficient and the demand variation coefficient is large, and the intersection and union ratio is low; that is, the feature distributions in the two matching results are significantly different, and the data of the two results overlap less, and there are more inconsistent parts; in recruitment or talent allocation, it indicates that there is a large deviation between the candidate's skill combination, experience background, etc. and the job requirements; in project allocation, it reflects the mismatch between the project's demand characteristics and the candidate's capabilities.

[0068] When the value of the difference matching coefficient is small, it means that the first matching result and the second matching result have a high degree of similarity in behavioral characteristics and demand characteristics; this usually means that the behavioral variation coefficient and the demand variation coefficient are close, and the intersection ratio is high; that is, the feature distributions in the two matching results are relatively consistent, the data of the two results overlap more, and the consistency is better; in recruitment or talent allocation, it shows that the candidate's skill combination, experience background, etc. are highly consistent with the job requirements; in project allocation, it reflects a good match between the project's demand characteristics and the candidate's capabilities.

[0069] like Figure 6 As shown, in one embodiment of the present invention, the present invention also provides a talent empowerment system based on big data, including: The behavior feature collection module is used to obtain the behavior features of the personnel to be assigned and construct a set of personnel behavior features.

[0070] The test item construction module is used to construct a first test item and a second test item corresponding to the personnel behavior feature set according to the personnel behavior feature set.

[0071] The matching calculation module is used to calculate the first matching coefficient, the second matching coefficient, and the final difference matching coefficient.

[0072] The allocation strategy adjustment module is used to adjust the allocation strategy for personnel according to the difference matching coefficient and output the final allocation result.

[0073] In the allocation strategy adjustment module, after obtaining the difference matching coefficient, personnel are allocated according to the difference matching coefficient, and the allocated personnel are output as the final allocation result. At this time, the allocation method can compare the current difference matching coefficient with the preset plan in the historical data, and use the calculation method of the Pearson correlation coefficient to compare the difference matching coefficient calculated for these personnel to be allocated with the preset plan, and output the preset plan corresponding to the maximum value of the Pearson correlation coefficient. The preset plan includes the basic value of the difference matching coefficient under the project or task to be assigned at this time, and calculate according to this basic value and the difference matching coefficient calculated at this time to obtain the final allocation result. The basic value can use the average value in the historical data to quantify the content to be allocated.

[0074] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A talent empowerment method based on big data, characterized in that: include: S1, obtain the behavioral characteristics of the personnel to be assigned and construct a set of personnel behavioral characteristics; S2, constructing a first test item corresponding to the personnel behavior feature set, wherein the first test item includes a demand feature corresponding to the behavior feature; Matching the demand characteristics with the behavior characteristics one by one, and calculating a first matching coefficient related to the demand characteristics; and obtaining a first matching result according to the first matching coefficient; S3, extracting a second test item from the first test item according to the first matching result; S4, matching the second test item with the behavior feature, identifying a second matching coefficient related to the behavior feature; and obtaining a second matching result according to the second matching coefficient; S5, performing a difference comparison between the first matching result and the second matching result, obtaining a difference matching coefficient corresponding to the first matching result and the second matching result, and adjusting the allocation of personnel according to the difference matching coefficient.

2. A talent empowerment method based on big data as claimed in claim 1, characterized in that: Step S2 includes: S21, using the demand characteristics and the behavior characteristics as input to construct a decision tree, where each node of the decision tree includes at least one set of demand characteristics and behavior characteristics; S22, calculating the Gini impurity and partial correlation coefficient of each node in the decision tree in turn; determining the division method of the child nodes of the decision tree according to the values ​​of the Gini impurity and partial correlation coefficient on each node; S23, continuously iterating the decision tree, outputting the iterated decision tree, and obtaining a first matching result.

3. A talent empowerment method based on big data as claimed in claim 2, characterized in that: Step S22 includes: selecting demand features and behavior features in the divided sub-nodes according to the partial correlation coefficients between the current demand features and behavior features, and calculating the Gini impurity of the divided sub-nodes. When the Gini impurity calculation result is optimal, the average value of the corresponding partial correlation coefficients is used as the first matching coefficient. When the partial correlation coefficients of the demand features and the behavior features are greater than the first matching coefficient, the corresponding behavior features and demand features are used as the data of the sub-nodes in the decision tree to obtain the division method of the sub-nodes of the decision tree.

4. A talent empowerment method based on big data as claimed in claim 1, characterized in that: The implementation method of step S3 includes: obtaining the feature vector of the first test item, and calculating the cosine similarity between the current first test item and other first test items in turn; clustering all the first test items according to the value of cosine similarity to generate multiple clusters, each cluster is set according to the variance of the cosine similarity of the first test items in the cluster, and the cluster is continuously iterated, and the cluster with the largest value after iteration is used as the second test item.

5. A talent empowerment method based on big data as claimed in claim 1, characterized in that: The implementation of step S4 includes: S41, comparing the second test item with the behavior feature, and calculating the co-occurrence probability of the second test item and the behavior feature; S42, determining the correlation between the behavior features under the corresponding values ​​of the common occurrence probability, and calculating the feature matching value between the behavior features; S43, calculating a second matching coefficient based on the co-occurrence probability and the feature matching value of the second test item and the behavior feature; S44, screening the current behavior features according to the obtained second matching coefficient, and outputting the screened behavior features and the demand features corresponding to the behavior features as a second matching result.

6. A talent empowerment method based on big data as claimed in claim 5, characterized in that: The feature matching value is expressed as, obtaining the correlation coefficient corresponding to the behavior feature, and calculating the feature matching value; ； in, represents the feature matching value between behavior feature i and behavior feature j, represents the correlation coefficient between behavior feature i and behavior feature j, represents the standard value of the correlation coefficient; Comparing the obtained feature matching value with the co-occurrence probability of the second test item and the behavior feature to obtain a second matching coefficient; The second matching coefficient is expressed as: ; in, represents the second matching coefficient of behavior feature i, represents the co-occurrence probability of behavior feature i and the second test item t, represents the adjustment factor, Indicates the number of behavioral features, Indicates the number of test items.

7. A talent empowerment method based on big data as claimed in claim 1, characterized in that: Step S5 includes: S51, based on the first matching result and the second matching result, respectively obtaining a behavior variation coefficient and a demand variation coefficient related to the first matching result and the second matching result; S52, according to the obtained behavior variation coefficient and demand variation coefficient, identifying the intersection-over-union ratio of the first matching result and the second matching result under corresponding values ​​of the behavior variation coefficient and the demand variation coefficient; S53, combining the obtained intersection-and-union ratio, behavior variation coefficient and demand variation coefficient to calculate a difference matching coefficient.

8. A talent empowerment method based on big data as claimed in claim 7, characterized in that: The calculation method of the difference matching coefficient in step S53 is as follows: S531, the behavior characteristics and demand characteristics in the first matching result and the second matching result are matched one by one, and the difference coefficient corresponding to the current behavior characteristic is calculated using the behavior variation coefficient, the demand variation coefficient and the intersection and union ratio; S532, obtaining a first matching coefficient and a second matching coefficient corresponding to the current behavior feature, and using the first matching coefficient and the second matching coefficient to adjust the difference coefficient of the current behavior feature to obtain a final output difference matching coefficient.

9. A talent empowerment method based on big data as claimed in claim 8, characterized in that: The coefficient of variation is expressed as: ； in, represents the coefficient of variation, represents the behavioral coefficient of variation of the first matching result, represents the behavioral coefficient of variation of the second matching result, represents the coefficient of variation of the demand for the first matching result, represents the demand variation coefficient of the second matching result, represents the intersection-and-union ratio of the first matching result and the second matching result under the corresponding values ​​of the behavior variation coefficient and the demand variation coefficient; represents the weight coefficient of the behavior variation coefficient, represents the weight coefficient of the demand variation coefficient, Represents the weight coefficient of the intersection-and-union ratio; The difference matching coefficient is expressed as: ； in, represents the difference matching coefficient, represents the first matching coefficient, represents the second matching coefficient, Represents an exponential constant.

10. A talent empowerment system based on big data, using a talent empowerment method based on big data as described in any one of claims 1 to 9, characterized in that: include: The behavior feature collection module is used to obtain the behavior features of the personnel to be assigned and build a set of personnel behavior features; A test item construction module, used to construct a first test item and a second test item corresponding to the personnel behavior feature set according to the personnel behavior feature set; A matching calculation module, used to calculate a first matching coefficient, a second matching coefficient, and a final difference matching coefficient; The allocation strategy adjustment module is used to adjust the allocation strategy for personnel according to the difference matching coefficient and output the final allocation result.

Images