Reasoning about data : The combination of multiple and non homogeneous sources of knowledgeFunded under: JRC-ADVMAT 1C
In structural engineering the combination of multiple and uncertain sources of knowledge is critical. These sources are usually any set of data, expert opinions, physical models or outcomes from instruments or techniques monitoring a particular event. Whilst the structure of the composition problem may differ from one case to another, a general pattern is preserved. The investigation and characterisation of this pattern can help to highlight the nature of the dependencies between the different sources. The paper aims at presenting a rigorous and general approach to managing the imperfection of data. The proposed analysis is performed in the framework of the theory of fuzzy uncertainty measures. The expressiveness of these measures makes them suitable for representing a wide range of situations. Two examples of the application of the above analysis to real problems are presented. The perspective of the analysis may be used in knowledge engineering.
Bibliographic Reference: Paper presented: International Course "Expert Systems in Structural Safety Assessment", Stuttgart (DE), Oct. 2-4, 1989
Availability: Available from (1) as Paper EN 35078 ORA
Record Number: 199010472 / Last updated on: 1994-12-01
Original language: en
Available languages: en