Introduction to the special issue on engineering semantic agent systems

Editorial ___________________________ DOI: 10.1111/j.1468-0394.2011.00585.x Introduction to the special issue on engineering semantic agent systems Atilla Elçi,1 Mamadou T. Koné2 and Mehmet A. Orgun3 (1) Software Engineering Program, Toros University, Mersin, Turkey Email: atilla.elci@gmail.com (2) Independent Computing Research, Montreal, The Province of Québec, Canada (3) Department of Computing, Macquarie University, Sydney, NSW 2109, Australia 1. Introduction Integrating agent and semantic web technologies in order to extend the ability of programmes have long been hailed as the panacea to serving users more efficiently and with less human intervention (Hendler, 2001). The idea was to significantly affect the performance of communicating multiagent systems through utilization of web services (WS) and ontologies. The World Wide Web has become prevalent and the semantic web has matured to a great extent evolving through indigenous data and protocol standards. Ontology-based applications have been on the rise and major enterprises are starting to take up the semantic web technologies. However, the pace of development in the semantic agent arena early in this decade was nowhere near desirable. We organized a workshop on the topic in conjunction with the 30th COMPSAC (2006) Annual International Computer Software and Applications Conference that took place in Chicago, Illinois, USA (ESAS, 2006). Thus the series of IEEE International Workshop on Engineering Semantic Agent Systems (ESAS) was born in 2006. Within months the ‘Special Issue on AI, Agents, and the Web’ appeared, paving the way for recognition of interplay and for better integration of the three branches (Hendler, 2006). Following the major success of the inaugural workshop, the 2nd IEEE International Work410 Expert Systems, November 2011, Vol. 28, No. 5 shop on ESAS took place in the bustling capital of China (ESAS, 2007). All in proper perspective, ESAS 2007 was a tribute to its main theme, namely, ‘Applying Semantic Web Technologies in Research and Development of Software Agents, Mobile Agents and Multi-Agent Systems’, contributing to the refinement of the concepts of semantic agent systems (SAS) and helping to bring forward better practices of engineering them. This was followed up by The 3rd IEEE International Workshop on ESAS which took place in Turku, the oldest provincial capital city of Finland (ESAS, 2008). The interplay of semantic web technologies and agent systems is promising a new dimension of capability in developing semantic-based multiagent information systems. The ESAS Workshop series aims at garnering the synergy of both worlds by taking up the common research issue. The topics covered span a wide spectrum of both theory and practice in the semantic web, agents, and AI, with due emphasis on the practical aspects of certain specific topics, such as autonomous semantic agents, context-aware intelligent agents, agents as semantic WS, software agents, mobile agents, agent architectures, multi-agent systems, agent communities, cooperation and goal seeking through sharing policy and ontology, safety and security in systems, and QoS issues. Effecting knowledge engineering as a practice and expert systems in general are the primary c 2011 Blackwell Publishing Ltd interests of ESAS. This is where ESAS relates to the Expert Systems: The Journal of Knowledge Engineering by Wiley=Blackwell Publishing and its select community of professionals. The authors of the accepted papers from ESAS (2007, 2008) were invited to submit substantially extended version of their work for consideration towards publication in the Special Issue on Engineering Semantic Agent Systems of the Expert Systems: The Journal of Knowledge Engineering. The special issue was also open to all potential authors. In the end, we received eight high-quality submissions involving researchers from Australia, Belgium, China, Germany, India, North Cyprus, Pakistan, Spain, Ukraine, the United Kingdom, and the United States of America. The submitted papers were reviewed by a large body of independent referees. Each submission was sent to two reviewers; additional reviews were also sought for several papers so that an informed decision could be made about them. Although we judged more submissions to be publishable, we could only include three papers in the special issue due to time and space limitations. 2. The contents of the special issue Three outstanding papers are selected for this special issue. Two of them deal with preparing platforms for semantic agents, and the other is an example case of agents in service. The paper by Stijn Verstichel, Femke Ongenae, Bruno Volckaert, Filip De Turck, Bart Dhoedt, and Piet Demeester is titled as An Autonomous Service Platform to Support Distributed Ontologybased Context-aware Agents. One of the areas in which semantic technology is being used more often is for effecting context-aware services. Utilizing ontologies in general and reasoning in particular can rapidly become resource intensive; especially so, as the dataset used by such services grows over time. Verstichel and colleagues suggest that, in order to address the tendency to outgrow resource requirements, and to take better advantage of the ubiquitous nature of services, there is a need to evolve a distributed approach by c 2011 Blackwell Publishing Ltd means of a service platform for agents collaborating in a context-aware environment. The article discusses the design and implementation of a service platform architecture for semantic ontology-based context-aware collaboration. The authors also discuss the way in which the platform can operate autonomously also taking into account the changing context of services. The paper titled Multi-Agent Collaboration based on Enhanced Cognitive Awareness: An Architecture for Agents’ Profiling on the Semantic Web by Amna Basharat and Gabriella Spinelli takes another look at collaboration of agents. Most agent-based applications assume predefined knowledge of agents’ capability, overlooking application requirements for interaction and adaptability. Often, basic cognitive and interactional requirements in multi-agent collaboration are neglected. This paper applies cognitive models in coming up with architectural and knowledge-based requirements for cognitive profiling of agents. The term ‘cognitive awareness’ alludes to agent’s ability to diagnose their own processing limitations and to establish interactions with their environment. The notion of cognitive profiling presented in this paper considers the semantic web as an action-mediating space, where shared knowledge base in the form of ontological models provides for improving cognitive awareness. Cognitive awareness capability results in enhanced flexibility, reusability, and predictability of agent behavior; thus, contributes towards minimizing human cognitive overload. The paper titled Applying Intelligent Agents and Semantic Web Services in eGovernment Environments by Francisco Garcı́a-Sánchez, Luı́s Álvarez Sabucedo, Rodrigo Martı́nezBéjar, Luı́s Anido Rifón, Rafael Valencia-Garcı́a, and Juan M. Gómez looks into how semantic web services (SWS) and agents may be seamlessly integrated using ontologies. As a case in mind, they inquire into how increasing volume of eGovernment-related services demand new approaches for service integration and different levels of interoperability. In such an environment, improving user experience would entail automating some tasks through software agents exploiting the semantic content. This paper Expert Systems, November 2011, Vol. 28, No. 5 411 offers a framework providing seamless integration of SWS and intelligent agent technologies interoperating through ontologies. The paper offers a proof-of-concept implementation in order to test the breakthrough of integrating disparate eGovernment services. 3. Prospects in SAS Extensive research and technology development for agents need be complemented by means of producing tangible semantic systems. The need for bringing about agent systems with shared semantics at recognizable magnitude was further pointed at in Hendler (2007). The way to the top is through software engineering in the joint venture of the two major fields of semantic web and agent-based systems. ESAS workshops have addressed some of those concerns. The ESAS main themes call for the ways and means of applying the semantic web technologies in researching and developing semantic MAS. The development of ‘semantic-based multi-agent information systems’ is promising new dimensions of capability in interaction of semantic web technologies and agent systems, over and above those of agentbased software engineering (Wooldridge, 1997). SAS, where MAS are equipped with properties and facilities of semantic web, are being used recently in a wide range of fields ranging from ecommerce (Khalili et al., 2008) to Robotics (Elçi & Rahnama, 2007). Today, SAS are able to integrate the semantically retrieved knowledge in decision-making systems. It is a challenging affair to produce context-aware decision-making systems using conventional AI, where an intelligent programme had no understanding of the context but only processes inputs according to the rules provided in advance. Research on SAS also requires deeper understanding of the underlying architecture of WS and agents. There are a few key differences between the conventional WS and the agent approach. A web service essentially is not aware of its user, and therefore cannot adapt itself. WS, unlike agents, are not designed to use and perhaps reconcile different ontologies. Agents are naturally interactive and communicative, 412 Expert Systems, November 2011, Vol. 28, No. 5 whereas WS expect a client to initiate a request. WS are not autonomous, whereas autonomy is one of the main characteristics of agents. Agents are cooperative, and by forming teams, tasks can be executed collaboratively providing more comprehensive services (Huhns, 2002). With SWS on the other hand, there is the OWL-S which is instrumental for semantic description of a service. SWS merge the abilities of conventional agents with facilities provided by WS; therefore producing web intelligence for use by SAS. The discrepancy between agents and WS is likely to shrink further through use of SWS. By the capability gained through OWL-S, built-in or external functionality providing ontology reconciliation, and possibly through discovery, matching, and composition, SWSs are likely to grab a bigger share of the joint functionality in relation to agents (Çelik & Elçi, 2009). Semantics of web and intelligence relies on the information provided by ontologies. While ontological description of services is important, presenting a well-structured ontology helps in retrieving more accurate information in a shorter time. Ontologies can be used for communication and interoperability (Uschold & Gruninger, 1996), that are based on the shared vocabulary to facilitate knowledge classification, reasoning, and navigation which may be deemed as being of paramount importance for semantic agents. As the field of SAS matures further in the next few years, we expect to see ground-breaking advancements throughout the spectrum of the topics covered at ESAS series and others. Acknowledgements This special issue represents the culmination of efforts expanded by numerous fine people. We are greatly indebted to Lucia Rapanotti and Jon G. Hall, the then Editor-in-Chiefs, for their enthusiasm and continuing support for the special issue. Before we could complete this special issue, Lucia had left and Jon had become the Editor-in-Chief. We also whole-heartedly thank our authors who contributed their papers and the reviewers who contributed their expert c 2011 Blackwell Publishing Ltd opinions. The quality of the selected papers was further enhanced by their authors; thanks to detailed comments by the following reviewers of this special issue to whom we are grateful: Ali Orhan Aydin, Costin Badica, Susmit Bagchi, Anne Cregan, R. Cenk Erdur, Manolis Gergatsoulis, M. Adeeb Ghonaimy, Dennis Hooijmaijers, Natalya Keberle, Kevin Lee, Zaki Malik, Constantine Mantratzis, Lars Moench, Maurice Pagnucco, Rolf Schwitter, Michael Stollberg, Vijayan Sugumaran, Markus Stumptner, Murat S- ensoy, Laurent Lefort, Kerry Taylor, Dirk Vermeir, S- ule Yildirim. HENDLER, J.A. (2007) J. Hendler’s paradox: Where are all the agents, Intelligent Systems, IEEE, 22, 2–3. HUHNS, M.N. (2002) Agents as web services, Internet Computing, IEEE, 6, 93–95. KHALILI, A., A.H. BADRABADI and F. KHOSHALHAN (2008) A Framework for Distributed Market Place Based on Intelligent Software Agents and Semantic Web Services. IEEE Congress on Services Part II. IEEE, 23–26 September 2008, pp. 141–148. USCHOLD, M. and M. GRUNINGER (1996) Ontologies: principles, Methods, and Applications, Knowledge Engineering Review, 11, 93–155. WOOLDRIDGE, M. (1997) Agent-based software engineering, IEEE Proceedings of Software Engineering, 144, 26–37. References ÇELİK, D. and A. ELÇİ (2009) Semantic web enabled composition of semantic web services, in Proceedings of the 4th IEEE International Workshop on Engineering Semantic Agent Systems (ESAS 2009) in Proc. of the 33rd COMPSAC, S.I. Ahamed, E. Bertino, C.K. Chan, V. Getov, L. Liu, Hua Ming and R. Subramanyan (eds), Seattle, WA: IEEE Computer Society Press, V.II, 46–51. ELÇİ, A. and B. RAHNAMA (2007) Human-robot interactive communication using semantic web technologies in design and implementation of collaboratively working robots, in Proceedings of the Robot Human Interactive Communication 2007, Jeju Island, Korea: IEEE Computer Society Press, 273–278. ESAS (2006) Proceedings of the IEEE International Workshop on Engineering Semantic Agent Systems, in Proceedings of the Computer Software & Applications 2008, COMPSAC’08, 30th Annual IEEE International, Vol. II, Chicago, Illinois: IEEE Computer Society Press, 271–338. ESAS (2007) Proceedings of the Second IEEE International Workshop on Engineering Semantic Agent Systems, in Proceedings of the Computer Software & Applications 2007, COMPSAC’07, 31st Annual IEEE International, Vol. II, Beijing, China: IEEE Computer Society Press, 125–187. ESAS (2008) Proceedings of the Third IEEE International Workshop on Engineering Semantic Agent Systems in : Proc Computer Software and Applications, 2008. COMPSAC ‘08. 32nd Annual IEEE International, Vol. II, Turku, Finland: IEEE Computer Society Press, 502–598. HENDLER, J.A. (2001) Agents and the semantic web, Intelligent Systems, IEEE, 16, 30–37. HENDLER, J.A. (2006) Introduction to the special issue: AI, agents, and the web, Intelligent Systems, IEEE, 21, 11. c 2011 Blackwell Publishing Ltd The authors Atilla Elçi Atilla Elçi is full professor of software engineering at Toros University, Mersin, Turkey, since July 2010. He served with the Computer Engineering Program, Middle East Technical University (METU NCC, Spring 2010); Eastern Mediterranean University where he established the Internet Technologies Research Center (2003–2009); founder and managing partner of IT&T Inc., Ankara, Turkey (2008–2003); Haliç University, Istanbul, Turkey, where he founded and chaired the Computer Engineering Department (2000–2003); the International Telecommunication Union, Geneva, Switzerland, where he was chief technical advisor (1985–1997) responsible for IT project development and execution; METU Ankara, Turkey, where he was Expert Systems, November 2011, Vol. 28, No. 5 413 assistant chair and chair of Computer Engineering Department (1976–1985); Purdue University, W. Lafayette, Indiana, USA, where he was research assistant (1974–1975). His research and experience encompass web semantics, agent-based systems, robotics, machine learning, knowledge representation and ontology, information security, software engineering, and natural language translation. He has published over a hundred journal and conference papers. He has organized or served in the committees of numerous international conferences and journals. He has been organizing IEEE Engineering Semantic Agent Systems Workshops since 2006, Security of Information and Networks Conferences since 2007, and Intl Joint Robotics Competition and Symposiums since 2007. He obtained B.Sc. in Computer=Control (w=honors) at METU, Ankara, Turkey (1970), MSc and PhD (5.63=6.00) in computer sciences at Purdue University, W. Lafayette, Indiana, USA (1973, 1975). Website: http://toros.academia.edu/AtillaElçi/. Koné Mamadou Tadiou at the National=Panasonic, Tokyo Information and Communications Research Laboratory in Tokyo and then as a research associate at the Japan Advanced Institute of Science and Technology (JAIST) where he received a PhD degree in information science in 2000. The Tokyo Institute of Technology (TIT) granted him a master degree in mathematical and computing science in 1990. Before that, he earned a Licence de mathe´matiques at the National University of Côte d’Ivoire, West Africa in 1985. His research interests are in the semantic web, web services and software agents technology (mobile agents and multi agents systems) and its applications. He has been a member of a number of conferences, symposia, workshops programme committees, scientific societies like IEEE, ACM, Canadian Semantic Web and the International Network of Chairs of UNESCO in Communication (ORBICOM). In addition, from 2002 to 2007, he has contributed to Université Laval as a member of the Counsel of Université Laval, the Counsel of Faculty of Graduate Studies (FÉS), the Electoral College (election of the president of Université Laval, 2007), the selection committee of the Dean of the Faculty of Literature. Mehmet A. Orgun Koné Mamadou Tadiou is currently a researcher at Independent Computing Research (IRC) located in Quebec, Canada. From 2002 to 2007, as an assistant professor, he was involved with the Department of Computer Science and Engineering, Faculty of Science and Engineering, Université Laval in Québec, Canada. Between 1990 and 2002, he worked as a research scientist 414 Expert Systems, November 2011, Vol. 28, No. 5 Mehmet A. Orgun is a professor at Macquarie University, Sydney, Australia. He received his BSc and MSc degrees in computer science and engineering from Hacettepe University, Ankara, Turkey in 1982 and 1985, respectively; c 2011 Blackwell Publishing Ltd and his PhD degree in computer science from the University of Victoria, Canada in 1991. Before joining Macquarie University in September 1992, he worked as a post-doctoral fellow at the University of Victoria. He was elevated to the grade of a Senior Member of IEEE (SMIEEE) in 1996. He researches in the broad area of intelligent systems, with specific research interests in knowledge discovery, trusted systems, multi-agent systems, and industry applications of these research areas. He recently served as the workshop co-chair of the 1st and 2nd Australasian Ontology Workshops (AOW 2005, 2006); Knowledge Representation Ontology Workshop (KROW 2008); the 1st and 2nd Workshops on Logics for Intelligent Agents c 2011 Blackwell Publishing Ltd and Multi-Agent Systems (WLIAMAS 2008, 2009); and The 2nd, 3rd and 4th IEEE International Workshops on Engineering Semantic Agent Systems (ESAS 2007, 2008, 2009). He was the Program Committee Co-Chair of The 20th Australian Joint Conference on Artificial Intelligence (AI’2007) and the 14th Pacific-Rim International Conference on Artificial Intelligence (PRICAI’2010). He was the Conference Co-Chair of the 2nd International Conference on Security of Information and Networks (SIN 2009) and the 3rd International Conference on Security of Information and Networks (SIN 2010). He serves on the editorial boards of The Journal of Universal Computer Science and Expert Systems: The Journal of Knowledge Engineering. Expert Systems, November 2011, Vol. 28, No. 5 415