Abstract:
This research was based on quantitative study and its aims were to evaluate the users’
satisfaction of the available services in Sabaragamuwa university library. We used
standard Process for Data Mining (CRISP-DM) process which is an industry-proven way
to build predictive analytics models across the enterprise. The process contains the six
phases as follows: business understanding, data understanding, and preparation of data,
modeling, evaluation and deployment. A self-administered questionnaire was used to
collect data in the data understanding phase. The 5 point Likert Scale questionnaire was
distributed among stratified randomly selected 360 undergraduates who represented
10% of the total university population from the Sabaragamuwa University of Sri Lanka.
The descriptive statistics, Pearson correlation and multinomial logistic regression were
applied to analyze the obtained data in modeling phase. According to the results of
descriptive statistics, selected variables have a satisfaction level and mean of their ranking
values around 3 and 4. Furthermore, results suggested that selected variables are
distributed slightly normally. Pearson Correlation results revealed that library collection
(0.662) and library services (0.658) are highly correlated with physical facilities under
the 0.05 level of significance. Furthermore, library services highly correlated with library
collection (0.651) and library website services (0.681).
