| 摘要 |
A technique is disclosed for analyzing data entities of interest, including text, image, waveform, audio, video and other types of data. The data entities are accessed and identified in accordance with a domain definition and attributes contained in the entities. The attributes may be user-defined, as may the domain definition. The entities are classified in a one-to-many classification according to the domain definition. User-viewable representations are generated that present analysis of the entities, such as identifying similarities and distinctions between them based upon their attributes and the domain definition.
|
