LOPES, E. C.; http://lattes.cnpq.br/7197564797868150; LOPES, Expedito Carlos.
Resumo:
In the routine of many professionals is a common situation being confronted with a
problem that triggers a decision process for its resolution. However, not every problem is
considered common. When the problem presents difficulties associated with the absence of available solutions, or requires significant number of resources and effort to its solution, or to diagnose its causes, it is considered a complex problem. Case Based Reasoning (CBR) uses the history of similar cases supporting the resolution of complex cases, based on solutions extracted from a Case-Base, providing mechanisms to retain new individual
problem-solutions. But, CBR can not support the solution of complex problems when no
similar case is found in the Case-Base. On the other hand, Evidence-Based Practice (EBP) represents a decision-making process centered on justifications of relevant information contained scientific research proof found, generally, in reliable sites. But, EBP procedures do not provide mechanisms to retain strategic knowledge from individual solutions, which could facilitate the learning of decision makers, preserving used evidences. However, the integration of CBR with EBP includes important mechanisms to support solutions to complex problems, but we did not find domain-independent models that incorporate these paradigms. Furthermore, Context is a type of knowledge that supports identifying what is or is not relevant in a given situation. However, the use of contextual information in EBP is still an open issue. This thesis proposes a conceptual framework consisting of: architectural elements, and a conceptual scheme to represent the integration of evidence, context and cases domain-independent that use research evidences, oriented to assist system designers. In addition, we specify a decision support methodology to guide experts in solving complex problems using CBR and EBP, considering the contexts of decision making. To investigate the feasibility of the proposal, we developed the design of two applications in different application domains (medical and juridical). The validation was realized with a case study conducted in juridical domain, resulting in a prototype implemented.