Using Substitutive Itemset Mining Framework for Finding Synonymous Properties in Linked Data
[ 1 ] Instytut Informatyki, Wydział Informatyki, Politechnika Poznańska | [ P ] employee
2015
paper
english
EN Over the last two decades frequent itemset and association rule mining has attracted huge attention from the scientific community which resulted in numerous publications, models, algorithms, and optimizations of basic frameworks. In this paper we introduce an extension of the frequent itemset framework, called substitutive itemsets. Substitutive itemsets allow to discover equivalences between items, i.e., they represent pairs of items that can be used interchangeably in many contexts. In the paper we present basic notions pertaining to substitutive itemsets, describe the implementation of the proposed method available as a RapidMiner plugin, and illustrate the use of the framework for mining substitutive object properties in the Linked Data.
422 - 430
WoS (15)