Keith E Shafer
|描述：||pages 39-40 : illustrations|
|责任：||project manager, Keith Shafer.|
Every document collection has an underlying corpus structure, which seldom has a readily available, concise expression. Without an explicit corpus structure expression, it is difficult to build or use a database of arbitrary documents. Three basic steps are needed: (1) identify the corpus structure, (2) design the database, and (3) design the interface. This report describes these steps and presents tools developed to perform them.