The British Art Information Project at Tate, has as its key aims, the digitisation and indexing of 50,000 works from Tate's British Collection. The indexers describe the content of the works of art by adding keywords according to a subject index which has been created specifically according to the needs of Tate's collection. The project would, I imagine fall into your suggested category of creation of new records. * The types of works we are dealing with are all fine art objects, including, paintings, sculpture, works on paper (including the Turner Bequest - 30,000 sketches and watercolours), installations and prints. * Each work already has a catalogue record drawn from the collections database which gives 'tombstone' information such as artist, title, accession number, date etc. * Only about 10-20% of the works have pre-existing interpretative data eg caption texts, catalogue texts etc * Indexing staff can be classed as curatorial in that they have knowledge and experience in the interpretation of works of art * Workflow targets are based around two and a half persons working time over an average 4 day week (allowing for holidays and sickness etc) for completion of this task within an eighteen-month timeframe. That averages out at around 125 works per day, although in the early stages of the project progress was much slower, allowing for refining the index and getting used to the software etc. Therefore the current daily workflow target is 150-200 works per day. This varies immensely according to the nature of the work being described and the available interpretative information about it. Hope this is useful. Rachel Bhandari BAIP Content Co-ordinator Tate