creates clusters of items based on statistical analysis of content (with or without metadata) NOT just keywords, kind of like find other documents like this one
could be interesting addition to Chandler providing a means to find items in repositories that are related that do not have specific attributes in common or that are know a priori like keywords.
This is particularly useful for large and/or unfamiliar data sets, eg. Blogs, example searched RSS feeds for query to retrieve ranked relevance items that might not contain ANY or the search terms in the query
Set expectations we are trying to determine how NITLE institutions' requirements differ from Canoga and Westwood
In particular focus on interoperability/migration needs vis a vis Exchange/Outlook (evidentially a number of institutions are using Exchange servers to Eudora clients, unclear if they're talking IMAP or MAPI)
Try to field ASAP. We may have difficulty getting responses because of running into an end of semester/gone for the summer problem
send to Carla for comments
Carla will email to NITLE members (CIOs to delegate) with cover from her, responses will be emailed to Pieter to tabulate and draw conclusions