Managing Vagueness with Fuzzy in Hierarchical Big Data1

Out of the web of linked open data, comes a sense of networked “Big Data.” This large scale interconnected data is hierarchical, and often messy and full of subjective bias particularly when mass collaboration is concerned (e.g. wikipedia). In this paper we apply fuzzy set theory, specifically the X-μ approach which is shown to be more efficient than a standard fuzzy approach, to attributes within linked data. We look at hierarchical structures, using an example from the films subset of the DBpedia data repository. The hierarchical nature of film categories lends itself well to our application, and we apply fuzzy models to handle the vagueness in attributes such as film length, film budget, and box office takings.