Frequent Itemset Mining and Multi-Layer Network-Based Analysis of RDF Databases

Triplestores or resource description framework (RDF) stores are purpose-built databases used to organise, store and share data with context. Knowledge extraction from a large amount of interconnected requires effective tools methods address the complexity underlying structure semantic information. We propose method that generates an interpretable multilayered network RDF database. The utilises frequent itemset mining (FIM) subjects, predicates objects data, automatically extracts informative subsets database for analysis. results form layers in analysable multidimensional network. methodology enables consistent, transparent, multi-aspect-oriented knowledge linked dataset. To demonstrate usability effectiveness methodology, we analyse how science sustainability climate change structured using Microsoft Academic Graph. In case study, FIM forms networks disciplines reveal significant interdisciplinary communities change. constructed multilayer then analysis scientific areas. proposed process, search measure rank their multidisciplinary effects. identifies discipline similarities, pinpointing similarity between atmospheric meteorology as well geomorphology oceanography. confirm provides sampled which can be simultaneously analysed