Abstract:

The paper presents a set of formal tools supporting design and implementation stages of an original strategy for aggregation and integration of a heterogeneous set of feature assesments built using the concept of grading scales and linguistic variables. Feature assessments are provided by a collective of respondents answering research questionnaires. The proposed tools, both of theoretical and conceptual nature, support the analysis and implementation of the first two stages of the aggregation and integration strategy responsible for collecting feature assesments and transforming them into computationally-oriented machine representations. In the approach, each individual opinion is finaly represented by a finite set of normalized numerical intervals, which as a whole set captures its meaning. The proposed interval-based machine representation of collective assesment is the input to an original adaptation and implementation of an algorithm for collective knowledge aggregation and integration defined over the universe of numerical intervals and presented elsewhere by other authors.