| 摘要 |
A computer readable storage medium includes executable instructions to retrieve a dataset from a data source. The dataset includes a first set of categories. A data structure with nodes connected by edges is built to represent the dataset. A first set of merit values is calculated for the first set of categories. Each merit value is a measure of how efficient a category is in filtering data. The first set of categories is ordered based on a criterion. The first set of categories is then returned.
|
