Common Sense Reasoning in Diagnostic Systems - Efficient Decision Support Systems: Practice and Challenges from Current to Future - page 100

BLOCK OF TRAINING,
ADAPTATION AND
MODIFICATION

PROBLEM
SITUATION
ANALYZER

DECISION
MAKING
BLOCK

DATA
BASE

MODEL
BASE

KNOWLEDGE
BASE

SIMULATION
BLOCK

KNOWLEGE
ASQUISITION AND
ACCUMULATION BLOCK

PREDICTION
BLOCK

Fig. 1. The generalized structure of a real-time IDSS

Formally, a RT IDSS can be defined by the set
SS = <M, R(M), F(M), F(SS)>,
where
M = {M 1 ,…, M n } is the set of formal or logic-linguistic models, implementing defined
intelligent functions;
R(M) is the function for selection of the necessary model in a current situation;
F(M) = {F(M 1 ),..., F(M n )} is the set of modification functions of models M 1 ,..., M n ;
F(SS) is the function for the SS modification, i.e. its basic components M, R(M), F(M) .
The main problems, solved by RT DSS, are:
 diagnostics and monitoring - revealing of problem situations;

 decision search - searching an optimal or admissible sequence of actions allowing to

achieve the desired goal or solve the problem situations;

 forecasting - assessing the recommended actions related to the goal achievement (to

solve a problem situation).

2. Reasoning by analogy
Nowdays there are a great number of various models, schemes, and methods that describe
mechanisms of reasoning by analogy [Haraguchi et al., 1986; Long et al., 1994; Varshavskii et
al., 2005; Eremeev et al., 2005].
In (Haraguchi et al., 1986), the authors have proposed two types of analogies, an analogy for
solving problems and an analogy for forecasting: