| Abstract |
In today's data-driven world, organizations which make use of the transactional data they have accumulated over time to come up with a more realistic picture of their operations can make more informed decisions towards attaining their goals and interests. However, due to the huge volumes of these accumulated transactional data, they cannot just be easily and readily used for reporting and analysis purposes. With this, a data warehouse is needed to store these accumulated data obtained from different sources within an organization where other decision-support applications can be built on to guide management decisions. In this study, a financial data warehouse was developed with a multidimensional construct that splits time, finance unit, account, and time dimensions, which is updated periodically with the accumulated transactional data sourced from a financial database of a university, and accessible via a Representational State Transfer application programming interface (REST API). To demonstrate the API's functionalities, we have created a data visualization tool which we integrated into our university web portal and subjected it to usability testing by its target end-users. It was shown that most of the respondents find the tool useful. Also, a query performance test was conducted comparing the execution of certain queries on the source transactional database and on the data warehouse. Result showed that the query time was greatly reduced by an average of over 50% when these queries were executed on the latter. |
Mindanao State
University - Iligan Institute of Technology