At Bar-Ilan University in Ramat-Gan, Israel, the Research Authority serves as the academic unit responsible for the administration of securing and managing external and internal research grants. The Research Authority locates funding sources, develops proposals, approves proposals for submission, negotiates contracts, and ensures financial and scientific compliance of grants awarded. The staff of the Research Authority must be able to notify faculty members from all disciplines in the sciences and the humanities in a timely and targeted fashion of funding opportunities and other research-related matters.
Statement of the Problem
During the past five years, the Research Authority designed and implemented an electronic message system to communicate with the Bar-Ilan faculty. In 1995 we introduced a mainframe-based system (Listserv) drawn from the university payroll database. This system enabled us to communicate with the entire 1,200-member faculty. While the majority of faculty members welcomed this technological advance, some researchers objected to receiving unwanted or irrelevant messages, saying it resulted in too much noise. In response, we refined the system one year later by dividing our researcher community into three faculty distribution lists: (a) Natural, Exact, and Life Sciences, (b) Social Sciences and Law, and (c) Humanities (including Judaic Studies).
This policy change helped to reduce the amount of information overload. In concrete terms, this change meant that the physicists no longer received funding opportunity notices about research programs in philosophy. However, after conducting a faculty survey and an internal Research Authority review, the Research Authority realized that we needed to further improve the message communication system. We found that some faculty members filtered or deleted unwanted Research Authority mail, and a few others asked to be removed from the list(s) altogether.
Proposed Resolution of the Problem
To ensure more precise targeting of emails--to make sure that researchers received postings that were most relevant to their field of interest--we decided to design a messaging system based on pre-defined discipline-oriented keywords.
The objective was to create a system that would address the dilemma of noise reduction on the one hand, while on the other to ensure that the researcher would not miss golden opportunities of applying for funding. The goal was to attract more researchers to read the opportunities with the hope that more researchers would apply for funding from sources outside the university.
To accomplish this goal, we developed the following specific objectives:
1. The first was to develop the size and scope of the controlled vocabulary. Within this parameter, we identified the following three issues: (a) The vocabulary must be manageable by the Research Authority administrative staff (not information specialists), yet accurately reflect the faculty members' research areas. For example, we needed to distinguish between "Disease Entities and Medical Problems" and "Neuroscience and Mental Disorders" (b)The vocabulary must not be too general or we risk loss of relevance. (c) The vocabulary must not be too specific or we risk misclassifying a message.
2. The second objective was to ensure that messages reached their targeted recipients.
3. The third objective was to equip the Research Authority administrators who create the e-mail messages with the requisite minimal knowledge to manage the system effectively. Such knowledge includes learning the following skills: searching, retrieving, judging pertinence and applicability, abstracting, classifying, and writing creatively.
With these considerations accounted for, in September 2000 the Research Authority launched the KeyMail system. As a test case we chose the faculties of Life and Exact Sciences as our initial target audience. …