Knowledge Engineering and
Decision Support Research Group
1 GECAD presentation
The “Grupo de Investigação em Engenharia do Conhecimento e Apoio à Decisão” GECAD (Knowledge Engineering and Decision Support Group) is a research unit
settled in the Institute of Engineering - Polytechnic of Porto (ISEP/IPP) having as
mission the promotion and development of scientific research in the Knowledge and
Decision Sciences domains, having Information Technologies as support. The group
has 35 investigators, being 9 PhD, 12 PhD students, 10 MSc students and 4 MSc.
1.1 Research Lines
There are four research lines:
Planning and Scheduling
•4 Finished Projects: SFF, SAPPIOP, SAD-3PE, SAADESI
•Finished PhD+MSc: 3 PhD, 5 MSc
Knowledge Based Systems
•3 Finished Projects : SOCRATES, SATOREN, DIAARE
• Finished PhD+MSc: 1 PhD, 4 MSc
Intelligent Agents
•2 Finished Projects : WG-IMS, AgentLink
•Finished PhD+MSc: 1 PhD, 1 MSc
Knowledge Dicovering
•2 Finished Projects : ECOAR, GIS-SISAS
• Finished PhD+MSc: 1 PhD, 1 MSc
1.2 Ongoing Projects
There are seven projects going on:
GOMEC – Optimized Management in Electical Market
IMAGE – Integrated Tools for Computer Supported Manufacturing Systems
IDDA – Intelligent Distributed System for Power Distribution Automation
SCALINTEL – Decision Support System with Scalable Inteligence for
Scheduling
SANSKI - Semi-automatic Negotiation Service for Knowledge Interoperability
ONTOMAPPER - Ontology Automatic Mapping
DAMICE - Power Systems Consumer Profiles and Contracting Strategies
supported by Data Mining
2 Background
Knowledge and ontology-based systems are important research fields in the GECAD
research group for several years. Many different projects were conducted and
developed by GECAD. Ontology-based projects are specially focused in two research
fields:
 ontology mapping in the context of Semantic Web;
 ontology assembling according to orthogonal dimensions.
2.1 Ontology Mapping
“Ontology mapping” aims to provide knowledge interoperability between
independent entities, based on the definition of semantic relations between ontologies.
Semantic relations are specified at ontological level and further applied at data level
in the transformation of data.
MAFRA Toolkit is the main outcome of the GECAD research in the field of ontology
mapping. MAFRA Toolkit includes extensive tools for the acquisition, representation,
customization and execution of semantic mappings.
MAFRA Toolkit is a service-oriented approach, playing several core roles in the
MAFRA Toolkit. In special, services are responsible for the discovery and
specification of semantic relations, and validation and execution of data
transformations. The transformation process might be executed in different modes
(batch/on-line, push/pull, source/target oriented) depending on the application
requirements.
The automatic specification of semantic relations is still under research and
development, but the current approach focus in the exploitation of external knowledge
sources (e.g. dictionaries, WordNet, domain specific glossaries) droved by
transformation services requirements. The negotiation phase of the mapping process,
concerning the inter-entities ontology mapping consensus building, is also under
development, exploiting the application of agent-based technology.
2.2 Ontology assembly
Because it is difficult to engineer a complex ontology with time, we developed a
method that allows for factorizing the complexity of the engineering process,
Factorizing ONTology Engineering complexity (FONTE). This method divides the
engineering task into building a time-less domain ontology and a temporal theory
independently from each other. FONTE method provides an operator otimes that
assembles the two independently developed ontologies into the targeted ontology.
Besides we only had studied the assembly of time into a given ontology so far, we
have reason to conjecture, that (i) FONTE method may also be applied to integrate
other important concepts like space, trust, or user access rights (concepts that pervade
a given ontology in intricate ways and necessitate management of engineering
complexity); and, (ii) building on Wiederhold's idea of an ontology algebra,
the proposed operator otimes might perhaps be further utilized for having ontologies
interoperate in manifold ways.
3 Ongoing research
One of the most challenging characteristics concerning the automation of ontologybased interoperability derives from the subjective nature of ontology mapping
process. In fact, the ontology mapping process largely depends on the domain
expertise, which raises uncertainty in any automated processes. Minimization of
uncertainty is then one of the goals in the ontology mapping process.
Factorization of ontologies with formal orthogonal ontologies provides application
ontologies with reliable, formal ontologies that can be exploited in diminishing the
uncertainty of the mapping process.
The combination of both the ontology assembly and ontology mapping research fields
becomes therefore a very important direction in our current work.
4 Possible role in PLANSERVE
Our possible role in the project includes the research in the field of ontology
engineering, either ontology assembly, mapping or combination of several
approaches. GECAD team is especially aware of ontologies developed in scope and
for the Semantic Web, their representation and manipulation technology, limitations
and potentialities.
Interoperability and negotiation between web-services is also a research field where
GECAD team can provide support due to the background in both agent-based
technology and e-business processes.
Besides, planning for manufacturing is also a core research field within GECAD.