Strategy for Evaluation the Data in the Context of Smart Cities: Case Study of Transport System

Abstract

The concept “Smart Cities” knows as a form of sustainable development. It represents an urban model and performance of urban service. This concept is based on distributed and autonomous information infrastructure contains millions of information sources that will be expected more than 50 billion devices connected by using IoT or other similar technologies. In Information Technology, we often need to process and reason with information coming from various sources. Information is almost always tainted with various kinds of imperfection: imprecision, uncertainty, ambiguity, we need a strategy to allow for the representation, propagation and combination of all kinds of imperfect information. One of them is the theory of belief functions. Real-time data generated from autonomous and distributed sources can contain all sorts of imperfections regarding on the quality of data as the imprecision, the uncertainty, the ignorance and/or the incompleteness. For example any imperfection in data within smart city can have an adverse effect over the performance of urban services and decision making. This article is addressed in this context, and we address in this article the problem of imperfection in smart city data. We will focus on handling imperfection during the process of information retrieval and data integration and we will create an evidential database by using the evidence theory in order to improve the efficiency of smart city. The expected outcomes from this paper are 1) to focus on handling imperfection during the process of information retrieval and data integration 2) to create an evidential database by using the evidence theory in order to improve the efficiency of smart city. As experimentation we present a special case of modelling imperfect data in the Transport field. An evidential database will be built which will contain all the perfect and imperfect data. These data come from several Heterogeneous sources. Imperfect aspects in the evidential database expressed by the theory of beliefs.