A Weighted PK Graph Based Ranking Aggregation Method for Incomplete Lists

Ranking aggregation is an effective method to solve the ranking problem of employee performance. PK graph is very effective in dealing with incomplete list ranking aggregation. However, the impact of malicious scoring, incomplete ranking and subjective scoring cannot be solved by it at the same time. Thus, a weighted PK graph based on the rank aggregation method is proposed in this paper, which ultimately makes the ranking results more fair, just and scientific. Firstly, a logarithmic and an exponential functions are employed as the weight to treat the difference occupied by malicious and reasonable ranking lists. Secondly, the weight is introduced to the PK graph to get the ranking result. Thirdly, a series of numerical tests are implemented to compare the proposed method with five traditional ranking aggregation methods, including PK graph, Borda count method, average ranking method, Downdall method and MVR method. The experimental results show that the proposed weighted PK graph based ranking aggregation method has significant advantages over the five traditional ranking aggregation methods in terms of its ability to cope with incomplete and subjective lists.