Argument Service Platform With Integrated Components

• Home
• Goals
• Results
• Publications
• News

PROJECT

OVERVIEW

ASPIC DELIVERABLES

ASPIC SOFTWARE

ASPIC CONFERENCE COMMA '06

ASPIC NETWORK

IMPRINT

Partners

REsults

The deliverables of the project included contributions to basic research, sound and scalable software and innovative applications. The main body of the work is centred on the Priority Thematic Area "Information Society Technologies" of the 6th Framework Programme, objective 2.3.1.7. "Semantic-based knowledge systems" in particular and also made a significant contribution to the goals of 2.3.1.11"e-Health".

ASPIC is an FP6 Information Society Technologies STREP project

• that started 1/1/2004 with a duration of 36 months

• involving 11 partners from 8 countries (10 from EU, 1 from USA)

• With a mix of academic, research and commercial skills (6 Universities, 2 Research Organizations, 3 Commercial Companies)

• With a budget of 520 Person Months ( MEuro )

Work in ASPIC is organized in 7 Workpackages:

    WP1 - Case study systematic review and database - 44MM
    WP2 - Formal analysis and semantics - 75MM
    WP3 - Experimental testing and validation - 107MM
    WP4 - Generic argumentation software components - 134MM
    WP5 - Large scale demonstrators - 80MM
    WP6 - Exploitation and dissemination - 47MM
    WP7 - Project management - 33MM

Following ISO 9001:2000 standard guidelines, project management procedures and tools define the Project's Organizational Chart

• Key Players Roles and interactions

• Management and conflict/problem resolution Procedures

• Communication means and channels within the consortium

• Clear Metrics to measure quality in project work and deliverables

• Known Risk Factors along with appropriate measures

• Auditing/Reporting Procedures

• Documentation standards, guidelines and procedures

A clear hierarchy exists in the project.
Explicit Roles have been specified.
Decisions are made at all levels and nodes.
Conflicts and Problems are escalated where needed.
Communication is both formal and informal.

WP1. Case study systematic review and database

This work-package :

• carried out a comprehensive survey of the domains and applications in which argumentation techniques have been used

• recorded the results in a publicly accessible database.

The data provided the basis for a state of the art review of current argumentation systems and technology with a view to developing a consensus framework that defines the core argumentation concepts. This provided a basis for: developing a unified formal theory setting out requirements for standard and reusable software components that was the main technological innovation of the ASPIC platform setting out initial requirements of argumentation systems, including user requirements.

 

WP2. Formal Analysis and Semantics

WP2 developed a formal framework for argumentation systems to be used in single-agent and multi-agent applications, drawing on results and requirements defined in the informal State of the Art and consensus requirements from WP1. We developed:

• a general theory for the argumentation systems that fall within this general requirements, covering as many of the requirements as possible (including at least inference and decision-making, learning and inter-agent communication)

• associated semantics that provided a secure foundation for the rest of the project. In particular provided a basis for developing an operational semantics for the software components which are an important deliverable of ASPIC (WP4) and which must be provably sound, reliable and scalable if they are to be satisfactory for building the large-scale demonstrators in WP 5.

 

WP3. Experimental testing and validation
 

This work-package complements WP2 which is driven by formal and theoretical issues with experimental and rapid prototyping work on problems acquired from problem "owners", focusing on:

• Small scale scenarios designed to exemplify specific requirements and argumentation capabilities

• Medium scale scenarios derived from real applications in the domains of eHealth, eCommerce and legal reasoning.

• Deliverables include working models of argumentation systems which provide a basis for discussion between theorists, technologists and application developers.

 

WP4. Generic Argumentation Components

The primary objective of WP4 is to develop a core set of software components from which a range of single-and multi-agent argumentation applications can be developed.
 

These components identified and specified using the theoretical work of WP2 and experimental work of WP3. Functional requirements also informed by the needs of the application builders, so this work-package plays an important bridging role between the theoretical and practical aspects of ASPIC.

Best software engineering practices used to produce high-quality, scalable components in a standard language, probably Java.

WP5. Large Scale Demonstrator

The objective of this work package is to demonstrate the successful use of the ASPIC software component infrastructure to support various roles of argumentation, in inference decision-making, dialogue between distributed agents and learning functions.
 

Work drew on results of the consensus argumentation model to establish the scope of the final demonstrator scenario, and then use the ASPIC components to build two demonstrators:

• a version which is designed to integrate with NAVUS "conventional" agent platform GENESYS

• a "fresh start" demonstrator which is intended to provide more technical freedom in using ASPIC tools, to address technical challenges, if any, which cannot be easily addressed from within the GENESYS tool.

 

WP6. Exploitation and dissemination
 

• The main objective of this work package is to manage project results for exploitation and dissemination.

• Academic partners seek to disseminate research results to relevant research communities (argumentation, agents, knowledge systems etc.)

• Commercial partners and applications-oriented researchers built credible application demonstrators and begin the process of creating argumentation based products and services.
   

WP7. Project Management
 

• Ensuring proper management of work and coordination of partners compliant with the workplan

• Facilitating required collaboration through regular meetings and internal workshops

• Ensuring timely completion of deliverables and meeting of milestones

• Supporting scientific quality and appropriate dissemination of results

• Ensuring proper adherence to budget profiles.

Translating theory into practice (1): The technology pipeline
 

• ASPIC includes leading research groups in argumentation theory and practice. It offers an opportunity to pull through basic research in this new area into practical technologies and applications.

• We have established a "pipeline" to take results forward, from state of the art review and theoretical analysis into prototyping and experimental work, and from there to soundly engineered software components for the ASPIC platform.

• Each part of the pipeline is being carefully managed to ensure effective communication and collaboration between communities and disciplines, and proper translation of research into practice.
   

* NB proposal for a standard interchange format is being developed in consultation with the community, through the ASPIC network.

Translating theory into practice (2): The "moose" model

• One of the key challenges within ASPIC is ensuring that the insights and formal results from WP1 and WP2 are properly reflected in the engineered products developed in WP4.

• A rapid prototyping methodology has been developed to mediate between these work packages.

• The method brings together theory driven and problem prototyping and capture the results as succinct and executable logic programs.

• These programs provide both clear specifications for engineering and reference engines for use in testing behaviour of engineered components.

 

Translating theory into practice (3): Scenarios and prototypes

• Validating theoretical concepts in practical problem scenarios

• Looking at a wide range of technical issues which are not primarily formal: the ARTS of argumentation
      ” Assumptions of integrated argumentation systems.
      ” Roles that argumentation services can support.
      ” Technological options for designing component architecture.
      ” Specification of software.

• Investigating practical design, engineering issues "in the small" without making premature risky investments in software engineering

 

ASPIC Platform, "fresh start"

• Argumentation components reengineered from scratch, using xUML and Java.

• The design, implement, test cycle lead to scalable and scalable components that embody a sound interpretation of WP2/WP3 results.

• This tested and validated on large scale demonstrator(s) developed in WP5.

• The components can be delivered for incorporation into future versions of TalliSYS or other host applications, or as a basis for an entirely new platform.
   

CONTACTS:

Prof. John Fox - University of Oxford - Department of Engineering Science
Scientific Coordinator: john.fox@eng.ox.ac.uk

 

Mrs. Gianna Tsakou - Singular Logic S.A.
Administrative Coordinator: ytsakou@singularlogic.eu