Classical query languages are insufficient when working on flexible values, because they do not include flexible conditions. As a result, some extensions of the relational model of databases have been proposed, leading to the fuzzy relational database model. This paper focuses on the problem of deduction in relational databases, under the existence of flexible queries, and/or imprecise and implicit information; this is an important point, since, in real world problems, it is usually the case that not all of the information obtained is accurate or even possible to get.
The authors continue a research line that was developed in previous papers; in particular, the results presented are based on the so-called generalized fuzzy relational database (GEFRED) model. To put the contribution into context, just recall that there are two different mechanisms to carry out deduction in a logical database: the one oriented to the values of the attributes in a tuple, and the one oriented to sets of tuples.
A modified tuple-oriented algorithm for deduction with the logical representation of a GEFRED database is introduced, a definition for generalized rules that can represent and handle imprecise information is given, and the main features of an extended algorithm for deducting with fuzzy data by the application of these generalized rules are presented.
The paper is well written and structured, although the overall impression is a little bit unsatisfactory. The nature of the paper is essentially expositive, in that a number of definitions are simply presented to introduce the proposed approach. No theoretical results are stated about the proposed approach, and, instead of providing an experimental basis for the proposed modification, only some small examples are included.