Presenter: Prof. Eric Tsui (Hong Kong)

Abstract: This tutorial will introduce the concept of taxonomy and outline the methodologies and challenges in the creation and maintenance of corporate taxonomies. It will also explore the relationship between taxonomies and search strategies. Whenever, appropriate local and overseas case studies will be used to help reinforce the concepts

Presenter: Prof. Dr. Horst Bunke (Switzerland)

Abstract: The aim of this tutorial is twofold. First, we want to provide a basic introduction to the use of graphs for object representation in intelligent information processing. Secondly, we want to make the attendees familiar with recent trends and novel methods in this domain. While the keynote speech at the conference on ‘Intelligent Information Processing Using Graph Based Object Representation’ will give a broad introduction to and on overview of the field, this tutorial will go into some of the underlying details and will introduce many key concepts and methods in greater depth. The following topics will be addressed:

• Basic concepts from graph theory
• Basic algorithms for the computation of graph and subgraph isomorphism and maximum common subgraph
• Graph edit distance and other distance functions on graphs
• Suboptimal algorithms for graph distance computation
• Introduction to spectral graph theory
• The median graph problem
• Graph kernels and graph embedding
• Graph classification and clustering
• Hypergraphs
• Graphs with unique node labels
• Applications
• Graph databases for experimental work

Presenter: Dr. Robert Kremer (Canada)

Abstract: This tutorial examines agent communication paradigms starting with simple ad-hoc protocol descriptions, through the BDI (Believe, Desire, Intension) model, to the more recent Social Commitment paradigm. Ad-hoc protocol specification languages, such as Petri Nets, are easy to use for small protocols, but quickly become unwieldy as protocols grow larger and they become increasingly brittle as systems need to to account for failure conditions, unexpected events, and alternative actions. The BDI model, as implemented in the FIPA agent communication act library specification, addresses some of these issues by modeling agents’ internal beliefs (world model/state), desires (states the agent wants to bring about), and intensions (states the agent plans to attempt to bring about in order to fulfill chosen desires). One of the major problems with the BDI model is that an observer must have an omniscient knowledge of the internal workings of all agents if the observer is to understand and verify the correct functioning of the observed agent community. To address this and other problems with the BDI model the Social Commitment model, as implemented in the CASA (Collaborative Agent System Architecture) infrastructure, avoids any modeling of the inner workings of agents. Instead the Social Commitment model looks only at the observable communicative acts of agents to infer social (non-)compliance by the (non-)fulfillment of social commitments either implied (by accepted social norms) or negotiated (among the agents). Handling of instantiation and fulfillment of social commitments will be explained in terms of translating messages into social commitments operators via explicitly declared policies.

Presenter: Dr. Sheng-Chuan Wu (USA)

Abstract: Semantic Technology shares most of the same standards and tools for Semantic Web, but is geared for ICT production today. It is a paradigm shift in information modeling, in which much of the business logics is embedded in the model itself rather than in application code. Such application architecture is extremely flexible, easy to evolve and share. It is the most
robust and practical way for knowledge representation and engineering today.

In this tutorial, we will briefly review what Semantic Technology is, what it is good for and why it is better than traditional technology for certain class of ICT applications. We will give an overview of the basics of Semantic Technology, such as RDF, RDFS and OWL, and commercial tools for ontology editing, semantic database management and reasoning. The tutorial will conclude with some example ontology and applications employing Semantic
Technology.

Presenter: Prof. Norman Foo (Australia)

Abstract: A survey of an emerging area of multi-agent interactions. Topics include the following. What is so different about social agents? How does altruism arise? In what sense is a dysfunctional society a rational outcome? From where do social norms come? How do norms change? How much of these can be rigorously modeled? Why is the area interesting and important?

Presumed background: an interest in computational models of agents with beliefs; knowledge of some psychology, philosophy and social science will be helpful.

Presenter: Duc Nghia Pham, John Thornton, Prof. Abdul Sattar (Australia)

Abstract:A constraint is a powerful representational tool that defines a relationship over variables. These relationships can take many forms, such as predicate relations, algebraic equations, fuzzy relationships, etc. While knowledge about real world domains can often be naturally expressed as a set of constraints, finding satisfying variable assignments is a significant challenge. This tutorial will first provide an overview of and a background to the constraint satisfaction paradigm. We will describe the two most widely used approaches to solving general constraint problems: backtracking and stochastic local search. We will particularly focus on the propositional
satisfiability (SAT) problem, and examine DPLL and Local Search methods that can solve SAT problems efficiently. We will then introduce G12 platform, NICTA's largest project on constraint solving. We will conclude with an overview of the current challenges in the area.

Presenter: Prof. John Sowa (USA)

Abstract: A controlled natural language (CNL) is a precisely defined subset of a natural language that has an unambiguous mapping to a computable notation. The first CNL was Aristotle’s subset of Greek for stating a version of logic and the rules, called syllogisms, for reasoning about it. That subset is still useful for defining a hierarchy of concept types for an ontology, but much more is needed to support all aspects of semantic technology. With suitable extensions, CNLs can be used in all stages from system design to implementation and use.

The most important advantage of a controlled natural language is human readability. Unfortunately, an early CNL callled COBOL had limited readability because its ontology was based on computer hardware, not on the subject matter of the applications. Aristotle’s syllogisms were much more readable because their vocabulary was derived directly from the ontology (concepts and relations) of the subject matter. Modern CNLs follow Aristotle’s tradition of using the terminology familiar to anyone who understands the subject matter. They enable designers, implementers, and end users to read and understand the definitions, rules, facts, and commands processed by the computer.

Although anyone can read a controlled natural language, learning to write it effectively requires some training. This tutorial surveys the range of applications of CNLs, their use in communication among people and machines, and the methodologies, conventions, and tools for using, designing, and implementing CNLs.

Presenter: Dr. Edmund Ng (Malaysia)

Abstract: Augmented Reality (AR) is an old idea that is right now on the verge of success. This is due to the fact that until lately there have not been advanced enough technologies to make feasible AR applications. They have been lacking either in computing power, user tracking accuracy or ease of use and comfortableness, which all are necessary to produce a satisfying AR experience. AR is still in a reasonably early stage of research and development at various universities and high-tech companies, if measured against the vast application areas and potential it has. The greatest change for the AR technology is not necessarily in some specific high-end innovation, but rather in penetrating the mass markets in various forms of pervasive computing. Augmented reality is taking its first steps to the everyday lives of consumers where a range of practical applications can be envisaged beyond military application. These include education and training in general, architecture, construction, devices for the disabled, entertainment, sports broadcasts and et cetera.

Presenter: Prof. Narayanan Kulathuramaiyer (Malaysia)

Abstract: Semantic-aware systems refer to emerging applications that explore a deeper semantic modeling via the integration of a variety of available knowledge sources in enhancing the quality of decision making. The aim here is to provide a more human-like judgment by overcoming problems of current systems, which tend to be largely based on a shallow abstraction of low-level media features and syntactic attributes. There is great promise in enhancing the capability of applications via the assimilation of semantic technologies. This tutorial will provide an insight into semantic technologies as a basis for the development of semantic-aware applications. The topics covered in this workshop include:

• Overview of Semantic Technologies
• Introduction to Semantic-Aware Applications
• Semantic Similarity as a basis for Semantic Awareness
• Multimodal Semantic Integration
• Assimilative Intelligence within user context

Presenter: Prof. Andreas Dengel (Germany)

Abstract: A document is the principal means of preserving and transporting knowledge through time and space. The knowledge they contain may be the foundation for satisfying potential customers, for defining a common understanding, for driving decision making processes, for assessing facts and interrelationships, for establishing valuable contacts, for planning new products or for education in an as yet unknown area. The demand-oriented coverage and bundling of documents in the context of workflows is an essential requirement in order to build organizational memories helping enterprises to save their collective intelligence.

In this tutorial he will give an overview of the current situation of knowledge workers discussing workload, recent developments, and problems they have to meet organizational demands or guidelines. Based on that, I like to introduce the principles of semantic technologies, their strengths and potential, their risks and limits. Further, I will give examples of how they may be used to design and implement an information butler for personal information management that, in combination with a social sharing platform, may foster people to participate and contribute their knowledge to an organizational memory.

Presenter: Prof. Klaus Tochtermann, Vedran Sabol (Austria)

Abstract: Semantic technologies, and in particular the role and impact of semantic services in the future Internet are the main topics of this tutorial. After giving an overview of the field trends and recent developments will be introduced, whereby both practical and theoretical aspects will be addressed. A discussion on specific applications based on the Internet of Services should make the participants more familiar with the topic. A more hands-on experience shall be provided by examples such as semantic-enhanced retrieval and visualization, or automated orchestration of semantic services leading to a goal-oriented usage of the Web. To conclude, an outline of major challenges and a summary of experiences shall provide an outlook on future research.

Presenter: Prof. Paolo Traverso (Italy)

Abstract: Web services provide a universal basis for the integration of business processes that are distributed among the most disparate entities, both within an organization and across organizational borders.

One of the major challenges for industry-wide adoption of Web services is the automated composition of distributed business processes, i.e., the development of techniques and tools supporting an effective, reliable, low-cost, and time-efficient composition of distributed business processes. Such tools should provide an automated, transparent, and user centered support to the entire business process life-cycle, from analysis to execution. They should automatically perform the time consuming and error prone task of analyzing business processes in detail, selecting and composing suitable Web services, detecting problems in the interactions, and monitoring execution step by step. They should operate in a transparent and user centered way by suggesting solutions that can be adopted, refused, or refined by business analysts, designers, and programmers.

The tutorial will discuss some of the main research challenges in this scenario, and how AI can provide solutions to these challenges. We will focus on supporting tasks such as verification, synthesis and monitoring of distributed business processes. The tutorial will be structured in the following main parts:

• Introduction: Software Services as a new development paradigm
• Service Level Agreements
• Automated Service Composition
• Service Adaptation
• Service Run-Time Monitoring
• Semantic Descriptions of Services
• The Future Internet of Services