摘要
The diffusion of municipal wastewater treatment technology is vital for urban environment in developing countries. China has built more than 3000 municipal wastewater treatment plants in the past three decades, which is a good chance to understand how technologies diffused in reality. We used a data-driven approach to explore the relationship between the diffusion of wastewater treatment technologies and collaborations between organizations. A database of 3136 municipal wastewater treatment plants and 4634 collaborating organizations was built and transformed into networks for analysis. We have found that： 1）.the diffusion. networks are assortative, arid the .pattems of diffusion vary across technologies; while the collaboration networks are fragmented, and have an assortativity around zero since the 2000s. 2） Important projects in technology diffusion usually involve central organizations in collaboration networks, but organizations become more central in collaboration by doing circumstantial projects in diffusion. 3）The importance of projects in diffusion can be predicted with a Random Forest-model at a good accuracy and precision level. Our findings provide a quantitative understanding of the technology diffusion processes, which could be used for waterrelevant policy-making and business decisions.
The diffusion of municipal wastewater treatment technology is vital for urban environment in developing countries. China has built more than 3000 municipal wastewater treatment plants in the past three decades, which is a good chance to understand how technologies diffused in reality. We used a data-driven approach to explore the relationship between the diffusion of wastewater treatment technologies and collaborations between organizations. A database of 3136 municipal wastewater treatment plants and 4634 collaborating organizations was built and transformed into networks for analysis. We have found that： 1）.the diffusion. networks are assortative, arid the .pattems of diffusion vary across technologies; while the collaboration networks are fragmented, and have an assortativity around zero since the 2000s. 2） Important projects in technology diffusion usually involve central organizations in collaboration networks, but organizations become more central in collaboration by doing circumstantial projects in diffusion. 3）The importance of projects in diffusion can be predicted with a Random Forest-model at a good accuracy and precision level. Our findings provide a quantitative understanding of the technology diffusion processes, which could be used for waterrelevant policy-making and business decisions.
