Slides available
Date and Location
Date: Sunday July 10th
Time: 1.45 - 5.45
Location: Room 3 (Logic/KR) @ New York Hilton Midtown, 1335 Avenue of the Americas, New York, New York 10019, USA
Scope
Cities take more and more advantages of information and communications technology (ICT) to better manage their resources and improve the quality of life of its citizens. ICT spans many departments of the cities, from transportation, water, energy to building management and social-care services. AI techniques are getting more and more attraction from cities to represent and organize information, maintain sustainable networks, predict incidents, optimize distribution, diagnose faults, plan routes and organize their infrastructure. Managing traffic efficiently, among many other domains in cities, is one of the key issues in large cities. In this tutorial we describe the domains of applications which could benefit from AI techniques, along with introducing the necessary background knowledge. Then we focus on traffic applications, which make use of recent AI research in knowledge representation, logic programming, machine learning, planning, reasoning and optimization. Specifically we go through the next version of scalable AI driven traffic related application where (1) data from a variety of sources is collected, (2) knowledge about traffic, vehicles, citizens, events is represented and (ii) deductive and inductive reasoning is combined for diagnosing and predicting road traffic congestion. Based on these principles, a real-time, publicly available AI system named STAR-CITY was developed. We discuss the results of deploying STAR-CITY, and its related AI technologies in cities such as Dublin, Bologna, Miami, Rio and the lessons learned. The final part of the tutorial aims at discussing future AI opportunities including scalability issues for large cities.
1 Goal of the Tutorial
1.1 Aims
The goal of the tutorial is to provide answers to the following questions.
1.2 Target Audience
The target audience is the general IJCAI audience who is interested in the application of AI to solve practical real-world problems. The beginning and the end of the tutorial is deliberately constructed for both a large AI audience and application-driven research audience.
2 Tutorial description
2.1 Summary
Cities take more and more advantages of Information and Communications Technology (ICT) to better manage their resources and improve the quality of life of its citizens. ICT spans many departments of the cities, from transportation, water, energy to building management and social-care services. AI techniques are getting more and more attraction from cities to represent and organize information, maintain sustainable networks, predict incidents, optimize distribution, diagnose faults, plan routes and organize their infrastructure.
Managing traffic efficiently, among many other domains in cities, is one of the key issues in large cities.
In the first part of the tutorial, we describe the domains of applications which could benefit from AI techniques. Then we focus the tutorial on traffic applications, which ingested a lot of recent AI research in Knowledge Representation, Logic Programming, Machine Learning, Planning, Reasoning and Optimization. Specifically we go through the next version of scalable AI driven traffic related application where (i) data from a variety of sources is collected, (ii) knowledge about traffic, vehicles, citizens, events is represented and (ii) deductive and inductive reasoning is combined for diagnosing and predicting road traffic congestion. Based on these principles, a real-time, publicly available AI system named STAR-CITY (award winning system of the ISWC - International Semantic Web Conference - Semantic Web Challenge 2013 / 2014) was developed. We discuss the results of deploying STAR-CITY, and its related AI technologies in cities such as Dublin, Bologna, Miami, Rio and the lessons learned.
The final part of the tutorial aims at discussing future AI opportunities including scalability issues for large cities.
2.2 Duration
4 hours
3 Content
3.1 Detailed outline
Introduction The vision of AI in smarter cities will be introduced. Main challenges and domains where AI (Machine Learning, Operation research, Planning, Knowledge Representation and Reasoning) could be applied will be also introduced.
Focus Domain A focus on traffic and related AI systems, where descriptions and limitations will be exposed. A deep analysis of variety of data (of different format, volume, velocity and veracity) exposed by cities of Dublin, Bologna, Miami and Rio will be also provided e.g., journey times of vehicles, weather information, events occurring in the city, vehicle activity (GPS location, line number etc.). Traffic congestion leads to a wastage of time, fuel and money. AI techniques can be used to diagnose and predict traffic congestion, and provide better planning for city managers.
Main Part This part introduces the basics of AI techniques in use. Specifically we will focus on (i) one commonly used representation language for scalable AI systems, (ii) techniques for collecting and transforming raw data in suitable knowledge graph, (iii) deductive reasoning with dynamic knowledge for maintaining knowledge up-to-date, (iv) inductive reasoning for learning and prediction, and (v) abductive reasoning for diagnosing. An overview of related ontology reasoning tasks such as consistency checking and classification will also be provided. All advantages and limitations of representation choices, design and AI techniques presented will be carefully motivated and discussed. The following topics will be discussed in detail:
Representation A representation language from the Description Logic community, i.e., EL++, the description logic underlying the W3C standard OWL 2 EL.
Access and Transformation Techniques for scalable transformation of city data in consumable knowledge graphs.
Diagnosis Reasoning Identification of the nature and cause of traffic congestion is known as traffic diagnosis. Semantic matchmaking and concept abduction is used to generate diagnosis report. In this part of the tutorial, we discuss how deductive and abductive reasoning are combined.
Predictive Reasoning Estimating traffic congestions in the future using current traffic data and historical events is called traffic prediction. It is done by considering data from multiple sources, reasoning and ranking multiple association of streams. In this part we discuss how deductive and inductive reasoning are combined.
Application STAR-CITY (Semantic Traffic Analytics and Reasoning for CITY) is deployed in Dublin, Bologna, Miami and Rio. Live demo and videos of how the system is working in these cities will be shown.
Future Work Since real-world stream data is obtained from sensors, data can quickly accumulate. Scalable approach to traffic analysis is discussed.
This final part of the tutorial also aims at discussing future AI opportunities for large cities.
3.3 Materials
The material we plan to use in order to cover the topics in this tutorial are mentioned below. Most of them are presenters¢ current or previous work.
4 Prerequisite knowledge
Optional: Semantic web OWL, deductive, inductive, abductive reasoning would be helpful to the attendees but it is definitely not required. We would give a brief overview of the required background in the tutorial.
5 History
AAAI 2015 (+130 registrations)
6 Presenters
Pascal Hitzler is (full) Professor and Director of Data Science at the Department of Computer Science and Engineering at Wright State University in Dayton, Ohio, U.S.A. From 2004 to 2009, he was Akademischer Rat at the Institute for Applied Informatics and Formal Description Methods (AIFB) at the University of Karlsruhe in Germany, and from 2001 to 2004 he was postdoctoral researcher at the Artificial Intelligence institute at TU Dresden in Germany. In 2001 he obtained a PhD in Mathematics from the National University of Ireland, University College Cork, and in 1998 a Diplom (Master equivalent) in Mathematics from the University of Tübingen in Germany. His research record lists over 300 publications in such diverse areas as semantic web, neural-symbolic integration, knowledge representation and reasoning, machine learning, denotational semantics, and set-theoretic topology. He is Editor-in-chief of the Semantic Web journal by IOS Press, and of the IOS Press book series Studies on the Semantic Web. He is co-author of the W3C Recommendation OWL 2 Primer, and of the book Foundations of Semantic Web Technologies by CRC Press, 2010 which was named as one out of seven Outstanding Academic Titles 2010 in Information and Computer Science by the American Library Association's Choice Magazine, and has translations into German and Chinese. He is on the editorial board of several journals and book series and is a founding steering committee member of the Web Reasoning and Rule Systems (RR) conference series, of the Neural-Symbolic Learning and Reasoning (NeSy) workshop series, and of the Association for Ontology Design and Patterns (ODPA). He also frequently acts as conference chair in various functions. For more information, see http://www.pascal-hitzler.de.
Freddy Lécué (PhD 2008, Habilitation 2015) is a principal scientist and research manager in large scale reasoning systems in Accenture Technology Labs, Dublin - Ireland. He is also a research associate at INRIA, in WIMMICS, Sophia Antipolis - France. Before joining Accenture as a principal scientist and research manager in large scale reasoning system in Junuary 2016, he was a research scientist and lead investigator in large scale reasoning systems at IBM Research - Ireland. His research has received IBM internal recognition: IBM research division award in 2015 and IBM Technical Accomplishment award in 2014. His research received external recognition: best paper awards from ISWC (International Semantic Web Conference) in 2014, and ESWC (Extended Semantic Web Conference) in 2014, as well as semantic Web challenge awards from ISWC in 2013 and 2012. Prior to joining IBM Research he was Research Fellow at The University of Manchester from 2008 to 2011 and Research Engineer at Orange Labs (formerly France Telecom R&D) from 2005 to 2008. In particular, he is interested in exploiting and advancing Knowledge Representation and Reasoning methods for representing and inferring actionable insight from large, noisy, heterogeneous and big data. He has over 40 publications in refereed journals and conferences related to Artificial Intelligence (AAAI, ECAI, IJCAI, IUI) and Semantic Web (ESWC, ISWC), all describing new system to handle expressive semantic representation and reasoning. He co-organized the first three workshops on semantic cities (AAAI 2012, 2014, 2015, IJCAI 2013). His research area is at the frontier of intelligent / reasoning systems, and Internet of Things. In particular he is interested in: Cognitive Computing, Knowledge Representation and Reasoning, Internet of Everything, Large Scale Processing, Software Engineering, Service-Oriented Computing, Information Extraction and Integration, Recommendation System, Cloud and Mobile Computing.
Raghava Mutharaju is a PhD student in the Computer Science department of Wright State University, USA and is supervised by Prof. Pascal Hitzler. His dissertation work involves investigating various approaches to distributed reasoning of OWL ontologies. His research interests are in Knowledge Representation and Reasoning, Distributed Computing, Semantic Web and its applications, Scalable SPARQL query processing. His work was presented at ISWC 2012, SSWS 2013 and DL 2010. He is the PC member of ESWC 2014 (posters and demos), SSWS 2014 and served as an external reviewer for several conferences such as ISWC, WWW, ECAI, AAAI, KR, ESWC, Hypertext, ICSC, JIST, ICBO, FOIKS. He will be presenting a tutorial on Large Scale Reasoning over Semantic Data at ISWC 2014. His publication list can be found at http://scholar.google.com/citations?user=69pEM_YAAAAJ.
Jeff Z. Pan is a Reader in the Department of Computing Science at University of Aberdeen, where he is the Deputy Director of Research of the department. He has over 100 refereed publications. He is an editor of the International Journal on Semantic Web and Information Systems (IJSWIS) and serves on the Editorial Board of the Journal of Web Semantics (JoWS) and the Journal of Big Data Resarch. He served as a program chair of JIST 2011, RR 2007, and of the Doctoral Consortiums in ISWC 2010 and ESWC 2011. He is a general chair of JIST 2014 and CSWS 2014. He is a key contributor to the W3C OWL2 standard and is widely recognised for his work on scalable and efficient ontology reasoning and query answering (see e.g. the TrOWL Tractable OWL 2 reasoning infrastructure that he leads, http://trowl.eu/). He has given a number of invited talks, including two keynotes in international conferences, on ontology reasoning in general and scalable ontology query answering in particular. He has rich experience in giving tutorials at leading international conferences and summer schools. He gave a tutorial on OWL 2 as part of the Advanced SIKS 2009 Course on the Semantic Web, a tutorial on OWL 2 at the CSWS 2009 Summer School, and a tutorial on Semantic Web Rule Languages and OWL 2 at CSWS 2009. He gave a tutorial on Scalable OWL Reasoning for Linked Data at ESWC 2010 and a tutorial on Large-Scale Ontology Reasoning and Querying at the AAAI 2010 Tutorial Forum. He gave a tutorial on Efficient and Scalable DL Reasoning at the Reasoning Web Summer School 2010 and 2011. He gave tutorials on Stream LOD and Querying Ontological Linked Data, and on Linked Data Enabled Software Engineering in IASLOD 2012. He also led the ISWC 2013 tutorial on Stream Reasoning for Linked Data. He will be giving a tutorial on Large Scale Reasoning over Semantic Data at ISWC 2014. His publications list is available at http://homepages.abdn.ac.uk/jeff.z.pan/pages/pub/.
Jiewen Wu is a lead research scientist at Accenture Technology Labs in Ireland. Before joining Accenture he was a Postdoctoral Researcher at IBM Research, Smarter Cities Technology Center (SCTC) in Dublin, Ireland. His main research interests include Knowledge Representation and Reasoning, the Semantic Web, and query processing over knowledge/data bases. He is currently working on optimization techniques for reasoning with large semantic data in the Dublin lab. He has published in refereed journals and conferences related to Artificial Intelligence (IJCAI, KR, JAR) on efficient reasoning and query processing techniques over ontologies. He gave several lectures to undergraduate students on Logic and Computation at the University of Waterloo in 2010/2011.
7 Presentation style
Slides based presentation along with a live demo and video of the case studies.