The usefulness and effectiveness of any database depends on the quality of information that can be derived from it. Queries convert data into useful information. When querying a database, users often specify what they do not want rather than what they do want. Regular fuzzy query languages adopt a “symmetric” approach where what has to be accepted is considered the complement of what has to be rejected. In reality, this is not the case: “what has to be rejected is not necessarily the complement of what has to be accepted”--a phenomenon “known as the heterogeneous bipolar nature of expressing information needs.” For instance, a statement like “I don’t want a black car” does not mean that the person’s preference for all other colors is the same. “I would like a black or blue car” does not give a precise indication of whether the user has an equal preference for black and blue cars.
In this paper, the authors discuss fuzzy querying of relational databases that takes into account heterogeneous bipolarity, and present a framework for bipolar query satisfaction modeling. The paper begins with heterogeneous bipolar database querying approaches. The authors present their bipolar criteria-handling framework and explain how to create bipolar flexible queries. They then explain the usefulness of bipolar queries in different scenarios, using several excellent examples, and present the advantages and limitations of the proposed techniques. According to the authors, their proposed framework and approaches are similar to the way “humans deal with selecting the best solution in real life situations.”
This excellent paper is a must-read for researchers and students working in the data mining, knowledge discovery, information processing, database management, and data warehousing fields. The designers and developers of data mining, knowledge discovery, and other information processing tools would also benefit greatly from the techniques described.