Current data mining and data analysis applications are designed primarily to address commercial business intelligence requirements. Business Intelligence (BI) applications include the activities of decision support systems, query and reporting, online analytical processing (OLAP), statistical analysis, forecasting, and data mining. Conventional RDBMS-based BI applications are records-based, transaction oriented, and demand massive resource overhead for table joins and index management. Because relational database management systems (RDBMS) have established themselves as the primary standard for data storage, they have inadvertently become the foundation for data mining. The objective of this SBIR is to develop a novel cell-based approach to database design. This would enable the data model to store, process, and manipulate data much like the human brain, thereby permitting more effective and flexible use of the database and the underlying data. Performance gains are possible with cell based indexing which captures, aggregates, and handles data in native cellular format cells vice traditional indexes. Additionally, cellular database techniques facilitate value pooling, a dimension that aggregates clusters of cell associations. Entity relationship of link cells provides a powerful link analysis and knowledge discovery function, and opens the possibility of further advances in innovative data analysis capabilities.
Keywords: Cell Aggregation, Data Mining, Link Analysis, Entity Relationship
