كنفرانس بين المللی مهندسی صنايع و سيستم ها

صفحه اصلی / دوازدهمین كنفرانس بين المللی مهندسی صنايع و سيستم ها

A Combined Data Mining and MCDM Framework for Supplier Evaluation and Ranking Using the BWM-ERASA Method

نویسندگان :

Mohammad Sanjari-Parizi¹ Emad Roghanian²

1- دانشگاه خواجه نصیر الدین طوسی 2- دانشگاه خواجه نصیر الدین طوسی

کلمات کلیدی :

Data mining،Multi-criteria decision-making،Supplier evaluation،Best-Worst Method،ERASA،MCDM

چکیده :

Choosing which suppliers to work with is rarely a single-criterion decision. A buyer typically has to weigh price against delivery speed, quality against payment terms, and proximity against experience. criteria that pull in different directions and rarely agree on a single best option. This paper reports on a case study in which 87 suppliers of an industrial company were evaluated across eight such criteria using a hybrid approach that integrates exploratory data mining with a recently published multi-criteria decision-making (MCDM) method called ERASA. This integrated process is structured into four stages: data cleaning, directional normalization, K-means clustering to identify natural supplier segments; and Apriori-based association rule mining to surface hidden co-occurrence patterns. Criteria weights were then obtained through the Best-Worst Method (BWM), with price competitiveness and operating experience selected as the most and least important criteria, respectively, and the ERASA method was applied to produce a final ranking that accounts simultaneously for the ideal, anti-ideal, and average solutions. The clustering stage separated suppliers into three interpretable groups. Overall, the results suggest that pairing pattern-discovery techniques with a modern MCDM ranking stage produces conclusions that are both easier to justify to decision-makers and more robust to how the weighting question is answered.