Authors:
Victor Pablos-Ceruelo
and
Susana Muñoz Hernández
Affiliation:
Universidad Politécnica de Madrid, Spain
Keyword(s):
Databases, Fuzzy Logic, Search Engine.
Related
Ontology
Subjects/Areas/Topics:
Approximate Reasoning and Fuzzy Inference
;
Artificial Intelligence
;
Computational Intelligence
;
Fuzzy Information Processing, Fusion, Text Mining
;
Fuzzy Information Retrieval and Data Mining
;
Fuzzy Systems
;
Soft Computing
;
Soft Computing and Intelligent Agents
Abstract:
Although the relevance of fuzzy information to represent concepts of real life is evident, almost all databases contain just crisp information. The main reason for this, apart from the tradition, is that fuzzy information is most of the times subjective and storing all users points of view is unfeasible. Allowing fuzzy concepts in the queries increases the queries' expressiveness and asking for cheap products, big size, close hotels, etc is much more interesting that asking for products with a price under X, of the size Y, hotels at most X kilometers
far, etc. The way we propose for achieving this more expressive databases' queries is adding to the basic knowledge offered by a database (e.g. distance to hotel is 5 km) the link between this crisp concept and multiple fuzzy concepts that we use in real life (e.g. close hotel). We present FleSe, a framework for searching databases in a flexible way, thanks to the fuzzy concepts that we can define. In this paper we describe the easy proc
edure that let us define fuzzy concepts and link them to crisp database fields.
(More)