Towards developing 3rd Generation Intelligent Synthetic Composite Indicators

Synthetic Composite Indicators (SCIs) are assessment tools, usually constructed to evaluate and contrast entities performance, by aggregating abstract issues in many areas such as economy, education, technology and innovation. Most SCIs are built using statistical measures, but the jury is still out regarding their accuracy and effectiveness. This paper is proposing an alternative approach to build a new breed or 3rd Generation Intelligent Synthetic Composite Indicators (3G iSCi) based on computational intelligent methods. The suggested approach utilizes Fuzzy Proximity Knowledge Mining to build the qualitative taxonomy, and Fuzzy C-Means will be used to form the 3G iSCi. It is suggested to cluster related dataset for every single country and detect a cluster centre to act as the aggregated index for that country, which would identify natural homogeneously grouped data, allowing for concise representation of the relationships embedded within the variables. To illustrate, this research uses real variables and data to build a new Unified Intelligent ICT Index (U3I) based on the suggested soft computing methods. The results obtained so far are compared against Principal Components/Factor Analysis (PC/FA), and the Geometric Aggregation (GME), which are widely used statistical methods to weight and aggregate SCIs. The robustness of both techniques are evaluated using Monte Carlo simulation. The results obtained from this case study suggest a novel and intelligent way to build future synthetic composite indicators to better serve international organizations, public officials, decision makers and business leaders.