Design and Implementation of a Policy Recommender System Towards Social Innovation: An Experience With Hybrid Machine Learning

Abstract

Recommender systems are typical software applications that attempt to reduce information overload, while providing recommendations to end users based on their choices and preferences. Conventional structure of recommender system, collaborative filtering and content based filtering are not adequate enough to derive a policy consensus and recommendation from social network based data due to the uncertainty and overlap of data itself. Hence, this paper proposes, a conceptual design and implementation of a recommender system, capable of yielding different social innovation based policies incorporating first level of machine learning on policy data click or access method. The proposed model refers hybrid technology using fuzzy logic led to cumulative prospect theory and machine learning strategies to quantify optimal social innovation based policy recommendation. The repetitive pattern of policy selection and recommendation can also be taken care through learning attribute of the proposed system. The challenge of developing such recomender system is to analyze the pragmatics of contextual multi-optimization behavior of decisions under several social innovation based policy framework.