In order to access digital objects (DO), we need to have categorised or labelled the DO in some way, for example in a traditional library books will have been indexed or classified by subject (e.g. using Dewey Decimal or Library of Congress schemes), similarly art works are labelled with relevant categories (e.g., year of creation, artist’s country of residence, or some other grouping, such as portrait, still life, or abstract). When we search for web pages, relevant pages may be returned to some extent based on the labels the author has given them, but most likely the results will be based on some kind of automatic categorisation of the web pages performed by the search engine.

In the case of the last example, we are very used to digital documents being automatically processed in order to identify those relevant to search terms, since there are so many of them it would be prohibitive to label them by hand (or to trust their authors to label them accurately by hand). This emphasises the fact that labelling DOs, and doing so consistently, is a difficult and time consuming task.

This need is the motivation behind the two tools which we present in this module, PROPheT and PeriCoDe.