Fuzziness and Ontology in Personalization of Selection Processes in the Semantic Web

A concept of the Semantic Web provides a frame-work for development of systems with higher levels of usability, operability and independence. One of the most intriguing ideas of this new paradigm is related to the vision of web systems requiring minimal involvement of humans in activities related to a pre-selection and eventually selection of most promising alternatives among all obtained options. In order to perform such a task, systems have to be able to mimic users´ behaviour. To be more specific, such systems should know not only users´ preferences but also users´ levels of acceptance of alternatives that satisfy multiple criteria to a different degree. All this represents a challenge especially in the case when any information provided by any individuals is more likely to be vague than crisp. The approach proposed here aims at creating capability to mimic human acceptance patterns in order to make web systems more independent and human-like at the same time. To address these needs, the concepts of fuzziness and fuzzy reasoning are introduced as important aspects of systems supporting users in their selection processes. Architecture of such systems with ontologies representing fuzzy information and abilities to express vague acceptance levels of users is proposed and described. Feasibility of the presented approach is illustrated by a prototype of Semantic Web application - Hotel Reservation System.