Performance Ranking Based on Bézier Ranking Principal Curve.

As for the problem of subjectivity and lack of scientificity in performance evaluation, the paper introduces a RPC model with cube Bézier curve being principal curve, and designs the learning algorithm of Bézier curve. First, evaluation indicators are chosen using Pearson correlation coefficient, RPC two-dimensional projection is conducted on relevant data and the ranking result of multiple indicator observation data is obtained. The experiment is done with the performance data of 12345 hotline telephone operators in one month of 2019 in one district of Beijing, and the ranking result is in line with the popular perception. The experimental result proves that as for the ranking in performance evaluation, RPC method is more scientific and interpretable than the method of weighted average and Elmap model.