CN-121998511-A - Labor affair outsourcing team performance intelligent evaluation system based on big data

CN121998511ACN 121998511 ACN121998511 ACN 121998511ACN-121998511-A

Abstract

The invention provides a labor service outsourcing team performance intelligent evaluation system based on big data, relates to the technical field of performance intelligent evaluation, and aims to solve the technical problems of single dimension, strong subjectivity, poor dynamic suitability and low evaluation accuracy of the conventional labor service outsourcing team performance evaluation. The system comprises a multi-source data acquisition module, a data preprocessing module, a characteristic extraction module, an intelligent evaluation module, a dynamic adjustment module, a result output module and a data safety module, wherein various data of a labor service outsourcing team are acquired through multiple sources, core performance characteristics are extracted after self-adaptive preprocessing, a self-created comprehensive performance evaluation formula and a team-first party adaptation coefficient formula are combined, a BP neural network and a random forest fusion model are adopted to calculate comprehensive performance scores, parameter weights are optimized through the dynamic adjustment module, multidimensional, intelligent and accurate evaluation of performance is achieved, and performance prediction, risk early warning and data safety guarantee functions are achieved.

Inventors

ZHOU YU
WANG ZHANLONG
WANG LONG

Assignees

王龙

Dates

Publication Date: 20260508
Application Date: 20260210

Claims (10)

1. The utility model provides a labor affair outsourcing team performance intelligent evaluation system based on big data which characterized in that includes: the data acquisition module is used for acquiring basic information data, task execution data, first side feedback data, cost control data and environmental impact data of the labor service outsourcing team in a multi-source mode, wherein the multi-source data comprises structured data, semi-structured data and unstructured data; the data preprocessing module is used for carrying out cleaning, denoising, normalization and characteristic alignment processing on the acquired multi-source data, removing invalid data by adopting a self-adaptive outlier rejection algorithm and outputting a standardized data set; The feature extraction module is used for extracting performance evaluation core features from the standardized data set, wherein the core features comprise team fitness features, task completion quality features, cooperation efficiency features, cost optimization features and risk early warning features; The intelligent evaluation module is internally provided with a performance evaluation model, and the model is used for calculating the comprehensive performance score of the outsourcing team based on the extracted core characteristics and combining with a self-created comprehensive performance evaluation formula so as to realize multi-dimensional intelligent evaluation of the performance of the labor outsourcing team; The dynamic adjustment module is used for dynamically optimizing parameters and characteristic weights of the performance evaluation model according to historical evaluation data, real-time task changes and first party demand adjustment; the result output module is used for outputting the performance score, the grade assessment, the short-board analysis and the optimization suggestion output by the intelligent evaluation module to the user terminal in a visual mode; the comprehensive performance evaluation formula adopted by the intelligent evaluation module is a self-created formula (1): (1); P is the labor affair outsourcing team comprehensive performance score, and the value range is 0, 100; t is a dimension score of task completion, a value range [0,100] represents timeliness and integrity of task delivery of outsourcing team; F is the fitness coefficient of the team and the first party, and the value range is 0.6 and 1.2, and is calculated by a self-created formula (2); C is a dimension score of the collaboration efficiency, the value range is 0,100, and the collaboration smoothness of the interior of the team and the first party is represented; S is a cost optimization dimension score, a value range [0,100] represents team cost management and control and resource utilization efficiency; D is a dynamic adjustment factor, and the value range [0.9,1.1] is dynamically output by a dynamic adjustment module according to a real-time scene; alpha, beta, gamma and delta are respectively the weight coefficients of each dimension, and satisfy the following conditions Alpha epsilon [0.3,0.4], beta epsilon [0.2,0.3], gamma epsilon [0.2,0.3], delta epsilon [0.1,0.2], and can be optimized in a self-adaptive way through a dynamic adjustment module; The self-created calculation formula of the fitness coefficient F of the team and the first party is formula (2): (2); M is the matching degree of the skills of outsourcing team personnel and the task requirements of the first party, the value range is 0,100, and the matching degree is obtained by comparing the skills labels of the team members with the task skills of the first party through big data; N is the degree of agreement between the service response speed of the outsourcing team and the expected value range [0,100] of the first party, and the value range is calculated by the feedback data of the first party and the response time length data; L is the mobility of the outsourcing team members, the value range is 0,1, and l=the number of current-period departure persons/the average number of current-period team members; k 1 、k 2 、k 3 is the fitness influence weight, which satisfies k 1 +k 2 +k 3 =2, and k 1 ∈[0.8,1.0]、k 2 ∈[0.6,0.8]、k 3 e [0.2,0.4]; epsilon is a correction factor and takes a value of 0.01, so that abnormal situations that denominator is 0 are avoided, and the calculation effectiveness of a formula is ensured.
2. The system of claim 1, wherein the multi-source data of the data acquisition module comprises: basic information data, including outsourcing age, academic, skill certificate, service life and post allocation data of team members; Task execution data, namely task allocation records, task starting/ending time, task completion progress, task reworking times and task acceptance result data; the first party feedback data comprises scoring of the first party on the task quality, evaluation on the team service attitude, complaint recording and demand change response satisfaction degree data; cost management and control data, namely team labor cost, consumable cost, management cost, task delivery cost and cost saving rate data; And the environmental impact data comprises industry benchmark performance data, similar outsourcing team performance data, industry environment change data where the first party is located and policy adjustment data.
3. The system of claim 1, wherein the adaptive outlier rejection algorithm of the data preprocessing module is specifically configured to calculate a mean μ and a standard deviation σ of each data dimension based on a big data statistical analysis, mark data exceeding a [ μ -3σ, μ+3σ ] range as outliers, set an industry threshold for task completion duration and cost fluctuation in combination with a labor outsourcing industry characteristic, perform a secondary check on the marked outliers, reject invalid outliers, retain reasonable outliers, and correct.
4. The system of claim 1, wherein the risk early warning features include an outsourcing team personnel loss risk value, a task delay risk value, and a cost overstock risk value, wherein the personnel loss risk value is predicted by a machine learning model trained by historical loss data, current personnel on-duty time, salary satisfaction data, and the task delay risk value is calculated in combination with task remaining time, current completion progress, and historical delay data.
5. The system of claim 1, wherein the performance assessment model of the intelligent assessment module adopts a fusion model trained by big data, the fusion model is composed of a BP neural network model and a random forest model, wherein the BP neural network model is used for performing fitting calculation on nonlinear features, the random forest model is used for performing weight distribution and score calculation on the linear features, and final performance scores are obtained by fusing output results of the two.
6. The system of claim 1 wherein the parameter optimization logic of the dynamic adjustment module is configured to periodically collect historical evaluation data and actual performance feedback data to calculate an evaluation error When E >5, triggering a parameter optimization flow, adjusting the weight coefficients of alpha, beta, gamma and delta and the value of k 1 、k 2 、k 3 in the formula (2) based on a gradient descent algorithm, and simultaneously adjusting the extraction weight of each core feature by combining the latest requirements and task type changes of the A party, so as to ensure the accuracy and suitability of an evaluation result.
7. The system of claim 1, wherein the task completion dimension score T is calculated by logic that: T=100-5×number of reworks-2×delay time length/standard time length×10, when T <0, t=0; The reworking times are accumulated reworking times of a single task, the delay time is a difference value between the actual completion time of the task and the standard completion time, and the standard time is obtained by analyzing the historical completion time of the similar task through big data.
8. The system of claim 1, further comprising a data security module for performing encryption storage, access control and data desensitization processing on the collected multi-source data, wherein sensitive data is encrypted by using an irreversible encryption algorithm, access control is managed by using character authority classification, and only authorized users can access the evaluation data and the original data of the corresponding level.
9. The system of claim 1, wherein the visual form of the result output module includes a line graph, a radar graph, a thermodynamic diagram and a text report, and supports the derivation of the evaluation result into PDF and Excel formats, and can be automatically pushed to the first party terminal, so as to realize bidirectional synchronization of the evaluation result.
10. The system of claim 1, wherein the intelligent assessment module further has a performance prediction function, iteratively calculates an outsourcing team performance prediction score of 1-3 months in the future through a formula (1) based on historical performance data, current task progress and a dynamic adjustment factor D, marks risk teams with the prediction score lower than 60 points, pushes early warning information to a user terminal, and avoids performance substandard risks in advance.

Description

Labor affair outsourcing team performance intelligent evaluation system based on big data Technical Field The invention belongs to the technical field of performance intelligent assessment, and particularly relates to a labor service outsourcing team performance intelligent assessment system based on big data. Background Along with the rapid development of labor service outsourcing industry, the scale of outsourcing team is continuously enlarged, the types are gradually diversified, and performance evaluation is taken as a core link of outsourcing management, so that the outsourcing service quality, cost management and control and the cooperation stability of both A and B are directly influenced. At present, the existing labor service outsourcing team performance evaluation technology still has a plurality of defects, and is difficult to meet the development requirements of the industry: firstly, the evaluation dimension is single, the prior art only carries out simple quantitative evaluation around the task completion condition, does not integrate multidimensional data such as first side feedback, cost control, team adaptation degree, personnel mobility and the like, and the evaluation result is one-sided and cannot comprehensively reflect the comprehensive capability of an outsourcing team; Secondly, subjectivity is strong, most evaluation modes depend on manual scoring, a scientific quantization formula and intelligent model support are lacked, human error is large, and evaluation results are not fair enough; Thirdly, the data processing capability is weak, an adaptive preprocessing scheme is not designed for multi-source heterogeneous data (structured, semi-structured and unstructured) of the labor outsourcing, and the abnormal value is removed by adopting a fixed threshold value, so that effective data are easy to delete by mistake, the data quality is low, and the evaluation accuracy is influenced; fourth, the dynamic adaptability is poor, the evaluation model parameters and weights are fixed, dynamic optimization cannot be performed according to the scene of the first party demand change, task type adjustment, industry environment fluctuation and the like, and the core demands of the labor service outsourcing team with strong mobility and the multi-first party service are difficult to adapt; Fifthly, a full-flow closed-loop design is lacking, an evaluation-feedback-optimized closed loop is not realized, and an effective performance prediction and risk early warning function is not available, so that risks of performance failure reaching standards cannot be avoided in advance, meanwhile, a data security guarantee mechanism is imperfect, and sensitive data leakage problems are easy to occur. Therefore, aiming at the defects of the prior art, a labor service outsourcing team performance intelligent evaluation system which can integrate multi-source big data, realize accurate quantitative evaluation and dynamic adaptation scene change and has full-flow closed loop and data security guarantee is developed, and the technical problem to be solved currently urgently. Disclosure of Invention In order to solve the technical problems, the invention provides an intelligent evaluation system for performance of labor affair outsourcing team based on big data, which solves the technical problems of single dimension, strong subjectivity, poor dynamic suitability, low evaluation accuracy, weak data processing capability, lack of full-flow closed loop and imperfect data security in the performance evaluation of the existing labor affair outsourcing team. An intelligent assessment system for performance of labor affair outsourcing team based on big data, comprising: the data acquisition module is used for acquiring basic information data, task execution data, first side feedback data, cost control data and environmental impact data of the labor service outsourcing team in a multi-source mode, wherein the multi-source data comprises structured data, semi-structured data and unstructured data; the data preprocessing module is used for carrying out cleaning, denoising, normalization and characteristic alignment processing on the acquired multi-source data, removing invalid data by adopting a self-adaptive outlier rejection algorithm and outputting a standardized data set; The feature extraction module is used for extracting performance evaluation core features from the standardized data set, wherein the core features comprise team fitness features, task completion quality features, cooperation efficiency features, cost optimization features and risk early warning features; The intelligent evaluation module is internally provided with a performance evaluation model, and the model is used for calculating the comprehensive performance score of the outsourcing team based on the extracted core characteristics and combining with a self-created comprehensive performance evaluation formula so as to realize multi-dimensional intel