Pattern Based Feature Construction in Semantic Data Mining

Abstract

The authors propose a new method for mining sets of patterns for classification, where patterns are represented as SPARQL queries over RDFS. The method contributes to so-called semantic data mining, a data mining approach where domain ontologies are used as background knowledge, and where the new challenge is to mine knowledge encoded in domain ontologies, rather than only purely empirical data. The authors have developed a tool that implements this approach. Using this the authors have conducted an experimental evaluation including comparison of our method to state-of-the-art approaches to classification of semantic data and an experimental study within emerging subfield of meta-learning called semantic meta-mining. The most important research contributions of the paper to the state-of-art are as follows. For pattern mining research or relational learning in general, the paper contributes a new algorithm for discovery of new type of patterns. For Semantic Web research, it theoretically and empirically illustrates how semantic, structured data can be used in traditional machine learning methods through a pattern-based approach for constructing semantic features.

DISESOR - decision support system for mining industry

“…This system is complete as it delivers the tools that can be applied to data storage, processing and preparation, and also to definition of the models based on expert knowledge (expert system) and the models based on the results of both historical and on-line data analysis. Due to the application and proper customisation of existing tools (RapidMiner, R) and development of the proprietary solutions (e.g., ETL2, rule induction and optimisation [19,20], rough set operators [21] and semantic analysis of data mining processes [22] that are not available in RapidMiner) a user receives a broad set of tools that can be applied to different tasks. Finally, the case study that was presented shows that the system can be practically utilised in a coal mine industry.…”

The Data Mining OPtimization Ontology

“…The capability of DMOP based meta-mined models to predict the relative performance of DM workflows was confirmed in [30]. This study used 1581 RapidMiner workflows solving a predictive modeling task on 11 UCI 7 datasets with various characteristics, whose meta-data was stored in the DMEX-DB containing over 85 million of RDF triples 8 .…”

“…Building a classifier that predicts whether a workflow is in the class of the best performing workflows or in the class of the rest of the workflows was addressed by [2] and [30].…”

Rule Quality Measures Settings in a Sequential Covering Rule Induction Algorithm - an Empirical Approach

“…In the other paper [35] a Dominance Based Rough Set Approach generated rules, and their length, determines the relevance of attributes, what becomes a criterion for a feature selection. There was also an attempt to characterise rule induction as a process specified by domain ontology [36], describing all data mining algorithms.…”