An Iterative Transient Rank Aggregation Technique for Mitigation of Rank Reversal

An Iterative Transient Rank Aggregation Technique for Mitigation of Rank Reversal

Bikash Bepari (Haldia Institute of Technology, Haldia, India), Shubham Kumar (Haldia Institute of Technology, Haldia, India), Awanish Tiwari (Haldia Institute of Technology, Haldia, India), Divyam (Haldia Institute of Technology, Haldia, India) and Sharjil Ahmar (Haldia Institute of Technology, Haldia, India)
Copyright: © 2018 |Pages: 11
DOI: 10.4018/IJSE.2018010104


With the advent of decision science, significant elucidation has been sought in the literature of multi criteria decision making. Often, it is observed that for the same MCDM problem, different methods fetch way-apart ranks and the phenomenon leads to rank reversal. To alleviate this problem, different methodologies like the Borda rule, the Copeland method, the Condorcet method, the statistical Thurstone scaling, and linear programming methods are readily available in the literature. In connection with the same, the authors proposed a novel technique to aggregate the ranks laid by different methods. The algorithm initially assigns equal weights to the methods involved to avoid biasness to a particular method and a simple average rank was obtained. Then, after the separation measures of individual methods with respect to average rank were calculated. Considering the separation measure the higher the weightage, the dynamic weights are ascertained to declare the weighted aggregate rank subjected to the terminal condition which include whether the previous rank equals to the current rank or not. To substantiate the proposed algorithm, a materials selection problem was taken into consideration and solved with the proposed technique. Moreover, the same problem was solved by existing voting techniques like the Borda and the Copeland-Condoract methods. The authors found a correlation of more than 85% between the proposed and existing methodologies.
Article Preview


During past decades, multiple criteria decision-making (MCDM) has advocated itself as a vital strategy for decision support systems which essentially require numerous alternatives, to deem their aptness subjected to a set of attributes as revealed by Yue (2011), Hwang and Yoon (1981) and Zeleny (1982). Multi-Criteria Decision Making methods like, TOPSIS, VIKOR, SAW, MOORA, PSI, COPRAS, ELECTRE, EVAMIX, EXPROM, MAUT, QFD, UTA, OCRA, MCDM-BOD, etc. have been successfully engaged to explain Multi criteria choice problem for diverse engineering application. The usage of above mentioned methods are authenticated with exclusive logical and mathematical algorithms which made them yardstick in the arena of decision support system. But no existence proof guideline is still present in the literature to opt for the most appropriate decision-making method as adopted by Cicek and Celik (2010). It has also been reported by Yeh (2002) that, different MCDM methods generate different ranks for a set of alternatives and leads to rank reversal. Voogd (1983) discovered that there seems to be a minimum of 40% probability of divergence of rank acquired by one method with respect to the others. Furthermore, Hajkowicz & Collins K. (2007) affirmed the non-existence of any rule in implementing a particular technique which is essentially superior to others. To dispose of this paucity, researchers have recommended implementing a number of MCDM methods concurrently as a substitute to judge the feasibility by mapping the decisions fetched by the different methods as discussed by Hwang and Lin (1987). Bairagi et al. (2014) & (2015) also encountered the problem of rank reversal while selecting the robots which certainly questions the suitability of their approached fuzzy MCDM techniques. Rao and Padmanabhan (2006) and Agrawal et al. (1991) encountered rank reversal problem during Industrial robot selection. Even in material selection problems, contradictory ranking order problem has been sought in the investigations by Rao and Patel (2010), Manshadi et al. (2007) and Chatterjee et al. (2009). To mitigate these problems, concepts of averaging function for integration of ranks achieved by different MCDM techniques was initiated by Ataei (2010). Moreover, to alleviate this requirement, BORDA and COPELAND voting rules are comparatively more prevalent. As per BORDA rule, elements higher ranks are provided more weightage than the lower one and accumulate up these points over all individual MCDM techniques. Unfortunately, both BORDA and COPELAND voting rules may obtain tied situation in ranking, whereas Jahan et al. (2011) introduced a software-based rank aggregation technique, but at the cost of more computational time. Amalgamating relative merits and demerits of the rank aggregation methods, the authors proposes an initial floating rank aggregation technique which goes on changing in an iterative way and ultimately saturates to determine the final aggregate rank.

The present investigation focuses on the introduction of a novel rank aggregation technique capable of determining accurate and an exclusive order preference to each alternative without the intervention of the group decision maker. This new approach in aggregation essentially eradicates rank reversal and perplexity in decision making. Motivation behind the present research seems the rank reversal phenomenon noticed in the literature which leads the group decision makers in quandary to choose, evaluate and differentiate alternatives. The originality of the current paper can be pointed as follows.

  • 1.

    This study explores a new avenue for accurate order preference and assortment of alternative in MCDM.

  • 2.

    This study introduces a separation-based rank aggregation technique without intervention of the personal choice of the group decision maker.

  • 3.

    This approach eliminates rank reversal in multi-criteria decision-making aggregation.

  • 4.

    The investigation uses the upshots of ranks laid by group of conventional methods.

Complete Article List

Search this Journal:
Open Access Articles
Volume 12: 2 Issues (2021): Forthcoming, Available for Pre-Order
Volume 11: 2 Issues (2020): 1 Released, 1 Forthcoming
Volume 10: 2 Issues (2019)
Volume 9: 2 Issues (2018)
Volume 8: 2 Issues (2017)
Volume 7: 2 Issues (2016)
Volume 6: 2 Issues (2015)
Volume 5: 2 Issues (2014)
Volume 4: 2 Issues (2013)
Volume 3: 2 Issues (2012)
Volume 2: 2 Issues (2011)
Volume 1: 2 Issues (2010)
View Complete Journal Contents Listing