The final report to the Mellon Foundation was completed in January 2003.
The search interface went live to users in June 2002. Statistics of use have remained around 2000 (individual searches) per month.
A revised interface, using information gathered from a second round of user testing (both in-house and remote), went live in October 2002. This interface included Boolean searching, sorting selection from the search page, new sorting methods, and the ability to access all the records retrieved. We plan on working on improvements in the future.
The number of records we harvest and make available through the search interface has nearly quadrupled in the past 6 months, from 274,062 in June 2002 to 1,070,724 records in January 2003.
We purchased the domain name "www.oaister.org" to increase access, as this domain name is easier to remember than "oaister.umdl.umich.edu."
The UNIX version of UIUC's harvester was shipped to us at the end of February. Installation, testing and feedback followed. Version 2 of the harvester was complete and shipped to us the third week of April.
We are currently using the harvester to gather metadata of registered OAI data providers. At present, we have 39 of the archives harvested and transformed to our DLXS bibliographic class format.
A short 7-question survey was created and publicized. We received 591 responses. Results will assist us in providing an appropriate interface and functionality for our service. View the summary report of the survey.
Three pages of a draft interface were tested with 9 users. We will use the results to further refine our interface. Our plan is to make the interface functional (searching the harvested archives) by the end of June.
Mike Burek and Kat Hagedorn will both be attending the Digital Library Federation Spring Forum 2002 in Chicago, IL, May 10-12. Kat will be presenting the latest project progress report (PPT format), and additionally will host a birds-of-a-feather discussion on issues surrounding OAI data and service providers.
Progress Report: February 2002
We have familiarized ourselves with the OAI protocol, its specifications, its uses, who's working in the field, where the work is being done and how our service could differentiate itself.
Our primary focus for this phase of the project was to contact individuals who could assist us in providing OAI or non-OAI compliant collections or lead us towards those who have such collections.
Most individuals have expressed interest in making their collections OAI compliant. Most are either involved in developing OAI clients or have this on their timelines for doing within the next 6 months.
We are in constant communication with UIUC who is building the harvester that we will use to gather the OAI compliant collections. The Java version of the harvester should be available for our use at the end of this month.
Simplified testing of harvesting and transforming has begun. Our work involves developing transformation scripts to be used to transform a collection's simple DC XML into a format we can index.
We are also working closely with UIUC to develop user testing methods and opportunities for educating data providers on the benefits of OAI compliance.
A survey has been developed, designed to understand the level of interest in such a service. Results of this survey should be forthcoming in March.