A Study on Document Warehousing and Text Mining for Business Intelligence Management

[[abstract]]Although data warehousing and data mining techniques are powerful and helpful in multi-dimensional analyses and knowledge discovery of cumulated historical business data for contemporary administrative decision-making, however, it is believed that there is only about 20% business intelligence can be extracted from formatted data stored in relational databases. The remaining 80% business intelligence is hidden in unstructured or semi-structured documents. Therefore, one of the next challenges will be the study of topics about document warehousing and text mining to help enterprises on obtaining the complete business intelligence. Since documents are multi-dimensional in nature, we claim that traditional indexing methods are not really suitable for document warehousing. In this paper, based on the concepts of R-tree, we propose an indexing structure called D-tree to fit the requirements of constructing document cubes. We present the formal definition, study the properties and devise related algorithms for D-trees. The work is essential for establishing an infrastructure to help combining text processing with numeric OLAP processing technologies.[[note]]NSC91-2416-H327-00