КМ из Бразилии (набрел по EAGE в Лондоне 2007)
Laura Silveira Mastella concluded the master program in Computer science by the Federal University of the Rio Grande Do Sul (UFRGS) in 2005. She has started the Doctorate in October 2006, working in the Institut Franзais du Pйtrole and registered at the Йcole des Mines de Paris, France.
She works in Computer Science, emphasis in Artificial Intelligence, with application to the Petroleum Exploration. In her Lattes resume, the most frequent terms in the scientific production are: Knowledge Acquisition, Knowledge Engineering, Intelligent Databases, Domain Ontologies, Knowledge Representation, Petrographic Description, Expertise, Inference e Artificial Intelligence.
http://www.inf.ufrgs.br/~mastella/english/index.html ========
Luнs Alvaro de Lima Silva
E-mail: l (dot) silva (at) cs (dot) ucl (dot) ac (dot) uk
Research student at
Department of Computer Science
University College London - UCL
Address:
University College London
Department of Computer Science
Malet Place
London - UK
Mr Silva holds a B.A. and a MSc. degree in Computer Science from Instituto de Informatica at Universidade Federal do Rio Grande do Sul – UFRGS, Brazil.
Currently, he is a research student at Department of Computer Science, University College London – UCL. He is being supervised by Prof. Bernard Buxton (Image Processing and Image Interpretation), with additional help from Prof. John A. Campbell (Artificial Intelligence). His research activities are supported by Brazilian Research Agency CAPES.
The PhD topic we are pursuing concerns the enhancement of the expressiveness of case representations for case-based reasoning, with particular reference to expert argumentation about image and other perceptual data.
In this research, we are describing, implementing and evaluating an approach to: (i) the capture and (ii) the effective computational use of argumentation knowledge regarding visual features, their inter-relationships, semantics, and related conceptual facts arising from knowledge of the application domain. We are investigating the challenging relationship between argumentation and visual interpretation by exploiting the way experts apply their knowledge and experience when they find it most natural to base their interpretations on previously analysed cases.
Therefore, the aim is to develop a case-based reasoning (CBR) approach which is an enhancement of traditional CBR to take argumentation characteristics into account. Such argumentation aspects are being collected, formalized and used to create augmented case representations. Our approach to CBR and argumentation integration innovates by taking advantage of numerical taxonomy techniques in investigating the similarity between factual and argumentation properties from these enhanced case structures.
The development and validation of this work is being supported on the study of two different applications where expert behaviour is primarily about reasoning on cases:
As contributions to Computer Science, the thesis will offer both the programs, as examples of captured knowledge, and the techniques that support them (e.g., reasoning templates for collecting and grounding argumentation knowledge, argumentation patterns for formalizing argumentation characteristics, direct association of decision-making steps in CBR with the arguments that have led to them). We believe these knowledge-based techniques are not necessarily confined to being relevant only to knowledge that refers to visual material.
http://www.cs.ucl.ac.uk/staff/L.Silva/