The influence of knowledge in the design of a recommender system to facilitate industrial symbiosis markets. (December 2018)

• Record Type: Journal Article
• Authors: van Capelleveen, Guido; Amrit, Chintan; Yazan, Devrim Murat; Zijm, Henk
• Is Part Of: Environmental modelling & software. Volume 110(2018)
• Page Start: 139
• Page End: 152
• Publication Date: 2018-12
• Subjects: Industrial symbiosis -- Recommender systems -- Decision support systems -- Input-output matching -- Association-rule mining; Environmental monitoring -- Computer programs -- Periodicals; Ecology -- Computer simulation -- Periodicals; Digital computer simulation -- Periodicals; Computer software -- Periodicals; Environmental Monitoring -- Periodicals; Computer Simulation -- Periodicals; Environnement -- Surveillance -- Logiciels -- Périodiques; Écologie -- Simulation, Méthodes de -- Périodiques; Simulation par ordinateur -- Périodiques; Logiciels -- Périodiques; Computer software; Digital computer simulation; Ecology -- Computer simulation; Environmental monitoring -- Computer programs; Periodicals; Electronic journals; 363.70015118
• DOI: 10.1016/j.envsoft.2018.04.004
• Languages: English
• ISSNs: 1364-8152

Abstract

Abstract: Industrial symbiosis aims to stimulate or enhance cooperation between industrial firms to utilize industrial waste streams from other industries and to share related knowledge, in order to achieve sustainable production. Recommenders can support industries through the identification of item opportunities in waste marketplaces, enhancing activities that may lead to the development of an active waste exchange network. To build effective recommendation, we study the role of knowledge in the design of a recommender that suggests waste materials to be used in process industries. This paper compares the performance of a knowledge based input-output recommender with a recommender based on association rules. The two recommenders are evaluated with real-world data collected through deploying surveys in a workshop setting. Our research shows that many data challenges arise when creating recommendations from explicit knowledge and suggests that techniques based on the concept of implicit knowledge may be preferable in the design of an industrial symbiosis recommender. Highlights: This is a first attempt to implement recommenders in industrial symbiosis networks. This research tests the role of implicit vs. explicit knowledge in recommender design. The recommender design performs better when using implicit knowledge. The major data quality issues for the recommenders are presented.