Identification and Prioritization of Public-Private Partnership Indicators in Iran’s Water and Wastewater Industry via Data Mining Algorithms

Abstract

The restrictions of government resources and the recent alterations in the economy have prompted government agencies to employ the capacities of private sector in all infrastructures. In this regard, a variety of financing methods, including the participatory models, have been applied for many years in the water and wastewater industry of Iran. The aim of this study is to identify and prioritize the Public-Private Partnership (PPP) indicators in the water and wastewater industry of Iran via machine learning techniques. To this end, after collecting, preparing and preprocessing the data, weighted indexing techniques including information gain and Gini index were utilized to prioritize the PPP factors. The results indicated that 93% of the indicators were effective in predicting the success of the projects. To compare the two methods, the precision of Naïve Bayes and Random Forest classifiers were taken into account and the information gain method yielded more reasonable findings with one percent difference. The evaluation of indicators elucidated that "complaints about service quality," "contract type," and "Conventional tariffs" revealed a huge impact on the success of collaborative projects. Among the 15 indicators, eight were directly pertinent to the project financing which is the main concern in this industry.