A data-driven approach to objective evaluation of urban low carbon development performance

An effective evaluation of a city's low carbon development plays an essential role in promoting low carbon development strategies for achieving the city's sustainable development. This paper proposes a data-driven approach to objectively evaluating the low carbon development level of cities. The approach formulates the low carbon development evaluation problem as a multi-criteria decision analysis problem and incorporates the merits of bibliometric analysis, text mining and optimal weighting to evaluating the urban low carbon devel-opment performance. The bibliometric analysis is applied to systematically identify evaluation criteria and associated indicators and establish an evaluation system for measuring low carbon development levels of urban cities. Equipped with an objective weighting method based on text mining, the approach determines the local weights of the evaluation criteria and indicators for each city by extracting subjective preferential information from the policy documents available on the local government's websites. Two optimal weighting models are developed to determine the optimal global weights of the indicators and criteria by maximizing the low carbon development performance of all cities. The obtained criteria weighting thus can reflect both the preferences of local city governments and the best common interest of all cities involved in the evaluation. The approach is then illustrated with a case study on three cities involved in urban agglomeration planning in China. The results compare the low carbon development performance of the cities, identify the disparities between the cities and reveal each city's obstacle factors that hinder its development. Policy recommendations are then suggested for developing effective low carbon development policies.