Crispyn : a Python library for determining criteria significance with objective weighting methods

Objective weighting methods play a significant role in multi-criteria decision analysis (MCDA). Determining weights by decision-makers using subjective methods is widely used. However, subjective weighting is often problematic due to the requirement of expert knowledge to correctly assign significance values to the criteria. Thus, objective weighting techniques show high potential in determining criteria weights. This paper presents a Python library providing eleven objective weighting methods that can be applied with MCDA methods. The presented library contains supporting methods to perform the complete multi-criteria evaluation of any decision problem, including analytical methods based on the Stochastic Multicriteria Acceptability Analysis method (SMAA), which may be applied alternatively to objective weighting methods. The practical usefulness of the developed library is demonstrated in an illustrative example of electric cars assessment. The implemented library allows for comparative analysis and indicates the most frequently top-rated and reliable alternatives. Another advantage of the developed library is the ability to automate the multi-criteria decision-making procedure by eliminating the decision-makers’ involvement.