An architecture of the semantic meta mining assistant for adaptive domain-oriented data processing

Abstract

Data mining is applied in various domains for extracting knowledge from domain data. The efficiency of DM algorithms usage in practice depends on the context including data characteristics, task requirements, and available resources. Semantic meta mining is the technique of building DM workflows through algorithm/model selection using a description framework that clarifies the complex relationships between tasks, data, and algorithms at different stages in the DM process. In this article, an architecture of semantic meta mining assistant for domain-oriented data processing is proposed. A case study applied proposed architecture on time series classification tasks is discussed.