Proceedings of the 2017 Industrial and Systems Engineering Conference K. Coperich, E. Cudney, H. Nembhard, eds. The PolisGnosis Project Enabling the Computational Analysis of City Performance Mark S. Fox Enterprise Integration Laboratory, University of Toronto msf@eil.utoronto.ca; www.eil.utoronto.ca Abstract Cities use a variety of metrics to evaluate and compare their performance. With the introduction of ISO 37120, which contains over 100 indicators for measuring a city’s quality of life and sustainability, it is now possible to consistently measure and compare cities, assuming they adhere to the standard. The goal of this research is to develop theories, embodied in software, to perform longitudinal analysis (i.e., how and why a city’s indicators change over time) and transversal analysis (i.e., how and why cities differ from each other), in order to discover the root causes of differences. The first phase of this project focuses on the creation of standard representations of city knowledge (i.e., Vocabularies and Ontologies) that can be used to represent indicators and their supporting data and publish them on the Semantic Web. The second phase focuses on the development of consistency axioms that automate the determination of whether a city's indicators and supporting data are consistent with the ISO 37120 definitions, and whether they are longitudinally and transversally consistent. The third phase focuses on the development of diagnostic algorithms that identify the root causes of longitudinal and transversal differences. Due to the heterogeneity of the supporting data, the applicability of classical diagnostic techniques is limited. Keywords: City Indicators, city ontology, consistency analysis, root cause analysis. 1. Introduction Open Data, part of the broader Open Government movement, is based on the belief that making data publicly available will lead to more effective public oversight. With more “eyes on the data”, waste and inefficiencies can be detected and crowd-based solutions suggested. But there is a problem with this belief, it assumes that citizens have the ability to read, understand and analyse open data. But its sheer volume and complexity exceeds the typical citizen’s abilities. Consequently, software applications need to be created that read, understand and analyze open data. The PolisGnosis project is exploring one such tool in the domain of city performance analysis. Cities use a variety of metrics to evaluate their performance. With the introduction of ISO 37120 (2014), which contains over 100 indicators for measuring a city’s quality of life and sustainability, it is now possible to consistently measure and compare cities, assuming they adhere to the standard and the data is openly published (Fox, 2015). But the volume of the data required to compute these indicators makes it impossible for citizens to analyse them. The goal of this research is to construct an intelligent agent that can diagnose a city’s performance. It will automate the longitudinal analysis, i.e., how and why a city’s indicators change over time, and transversal analysis, i.e., how and why cities differ from each other, in order to discover the root causes of differences. In the following we describe the PolisGnosis vision and the progress to date. 2. PolisGnosis Agent Architecture Our goal is to create a “universal” agent that is not tailored to specific indicators nor cities. Therefore the design of the PolisGnosis agent must satisfy the following requirements: 1. Indicator Independence. Since there are a vast number of indicators used by cities, beyond those defined in the ISO 37120 standard, and ISO standards evolve over time, we do not want our agent to have any knowledge of indicator definitions “hardwired” into its code. An indicator’s definition must be an input to the agent. 2. City Independence. Cities openly publish vast amounts of data that that our agent would like to use. But the data lacks any standard models or vocabularies - every dataset differs in structure, attributes and