Abstract:
Library acquires resources from time to time to have a balanced resource within the increment of
library users and to have the newest resources as soon as they are published especially in
academic libraries. As resources increase, it becomes difficult to users to select important
references and information of their interest. Therefore, this study aims to design Selective
Dissemination of Information (SDI) service that provides information alerting service to keep
individuals informed of new resources (books, article, etc) in their particular fields of interest.
Design science research method (DSRM) which creates and evaluates IT works proposed to
solve recognized organizational problem and the process of inspiring, designing, demonstrating,
evaluating, and communicating the artifact was followed. To this end, a prototype SDI system is
developed to recommend arrival of new books and journals using python programming language
for Jimma University Library System (JULS) users by applying an information filtering
approach. Concerned population for this study was 1610 academic staff of Jimma University; out
of which 921 considered for sample selection since the rest users have no staff profile on Jimma
University website. Hence, Profiles of eighty-six (9.34%) academic staff were registered in user
database. Among these eighty-six users, twelfth (12%) of them were used for user acceptance
testing. MySQL version 5.5 was used for recording user profile. For testing the prototype SDI
system, twenty percent of the data is used and the rest data is used for training. In this work,
different matching schemes are experimented; among them, TF*IDF weighting technique with
Vector Space Model (VSM) has registered the best performance of 78.76% precision. In
addition, the SDI system achieves 95% in user‟s acceptance testing which shows that it has high
user‟s acceptance. This further means that, it is advisable to use SDI system to enhance library
services. However, the proposed prototype system does not recommend books written in other
than English language. Therefore, future research direction is to develop an SDI system that
recommends books written in different languages, including local languages.
