Dear AMICOâs,

One of the things we can do to improve the data which RLG has for the testbed year is to give them parsed sub-fields in addition to text strings in some fields where the data dictionary provides for both. This will facilitate searching in some places. As you know, we contracted to have some parsing done automatically, and are generally pleased with the results, but need you to tell us if any of it is ready to pass along to RLG.

On the Focus Computing site, you will see a new listing for the Parsing Experiment. Please look at the report of the parsing experiment for your institution and tell us if the parses that were made by our automatic, but quite conservative, routines, work for your data. Note that some fields of your data were not parsed because the result, from the perspective of the automated routines, was ambiguous. We would still like to see this data parsed eventually, either by more sophisticated routines, or by your editors. This will be under discussion in future months. For now, we want to know which, if any, parsed fields from your data can be sent on to RLG.

!NOTE! We are continuing this month with experiments of matching your data to the AAT, ULAN and TGN vocabularies. The results will probably not be used in the testbed year, but we will keep you briefed on what we find and will be discussing the results with you.