Fuzzy-based methods for processing hybrid and temporal data in Linked Open Data environment

The web becomes a huge repository of data and information. In order to deal with large quantities of data, the users need intelligent-based tools and methods to access data, process it, and make it useful for variety of purposes. Proposed as a part of Semantic Web, Resource Description Framework (RDF) is an important way of representing information. Its intrinsic feature of high connectivity creates an opportunity to process web data in a way that information and knowledge embedded in it can be extracted and absorbed. RDF has become a building block of Linked Open Data. In this talk, we show how fuzzy-based technologies can be used to process RDF data and provide the users with capabilities to cope with versatile and complex information. These technologies, when employed, are able to ensure a human-like way of interacting with the web when looking for useful facts and information. At the beginning, we focus on methods suitable for dealing with hybrid, i.e., numerical and symbolic, data in the distributed environment. We adopt a linguistic representation model to identify items that match a given reference to the highest possible degree, and satisfying - at the same time - variety of criteria. The proposed methodology deals with different types of features describing items, and utilizes comparison mechanisms most suitable for each feature type.