Design and Implementation of a Data Mining System for Malware Detection
Journal of Integrated Design & Process Science
| Hi-index | 0.00 |
Resolving semantic heterogeneity across distinct data sources remains a highly relevant problem in the GIS domain requiring innovative solutions. Our approach, called GSim, semantically aligns tables from respective GIS databases by first choosing attributes for comparison. We then examine their instances and calculate a similarity value between them called entropy-based distribution (EBD) by combining two separate methods. Our primary method discerns the geographic types from instances of compared attributes. If successful, EBD is calculated using only this method. GSim further facilitates geographic type matching by using latlong values to further disambiguate between multiple types of a given instance and applying attribute weighting to quantify the uniqueness of mapped attributes. If geographic type matching is not possible, we then apply a generic schema matching method which employs normalized Google distance. In addition, we seek to address additional challenges encountered when employing clustering-based geospatial schema matching. We focus on three distinct challenges. First, many schema matching algorithms, including GSim, rely only on one instance property. Second, a consistent score for an attribute match is not produced. Third, hierarchical relationships between the data are not considered. In order to meet these challenges, we develop a successor to GSim called GeoSim. GeoSim derives clusters from attribute instances based on their geographic and semantic properties and produces a high-quality clustering by optimizing an objective function. It also captures hierarchical relationships between the GTs representing the instances in compared tables and attributes. Finally, GSim possesses the ability to execute greedy 1:N matching that reveals relationships between several attributes. 1:N matching is defined as an optimization problem and is justified with a proof of correctness. With impressive results generated, we show the effectiveness of GSim and GeoSim over traditional mining-based similarity approaches across multi-jurisdictional datasets.