Improved interactive inference approach for constructing a complex multi-industrial symbiosis network.

The development of industrial parks is an important avenue to promote economic growth in the context of modernization. Industrial symbiosis (IS), mainly characterized by the physical exchange of product, by-product, and waste, has been established worldwide in industrial parks for better economic, environmental, and resource-consuming performances. The suitable design of input-output matching builds the base of IS practice. This paper aims to table an improved constructing approach based on interactive inference processes for a complex multi-industrial symbiosis network (MISN). Interactive reasoning rules are simplified, material group reduction is introduced, and the multi-industrial stepwise constructing process is established. The proposed approach has been applied to an assumed district in Qaidam Circular Economy Experiment Area (QCEEA). Three new node information databases are collected. A MISN which is across four industries and includes 79 nodes is built through four steps of constructing and 22 rounds of interaction. Results and comparisons confirm its superiority on designing complex network with numerous resource- and non-resource-based nodes over previous methods. Specifically, this new approach reduces the amount of inference calculation, improves the construction efficiency, and can be utilized in non-resource industries. Preliminary discussions are made to the sequences of target industries when constructing a MISN and implicate which would lower manual workload of extra adjustment and supplement.