Estrella, Ponay of CICS discuss telehealth application, bilingual data clustering at conference in Malaysia

College of Information and Computing Sciences faculty members Assoc. Prof. Noel E. Estrella, who is the former chair of the Information Technology Department, and Asst. Prof. Charmaine S. Ponay virtually presented their respective research papers during the 7th International Conference on Software Engineering and Computer Systems (ICSECS 2021) and 4th International Conference on Computational Science and Information Management (ICoCSIM 2021) held on August 24 and 25, 2021, The conference was organized by the University Pahang Malaysia (UMP).

Estrella’s paper was titled “Sentinel: The Development of a Web and Mobile Application for the Development and Testing of an E-service Learning Interprofessional Telehealth Community Based Rehabilitation Program among Hypertensive Clients.”

Estrella and his team composed of Information Technology students, Rafael Benedict E. Bacungan, Kurt Martin C. Choi, Jansen Patrick A. Chua, and Jericho P. Dupo, developed a mobile application called Sentinel that can help healthcare providers synchronously consult with hypertensive clients through personal messaging. It also has asynchronous functions where patients can download brochures and other information to aid in managing their condition.

After a User Acceptance Test (UAT) with 17 healthcare providers testing the application, “Sentinel” met the requirements for the development of a web and mobile application.

Ponay’s paper was “Topic Modelling and Clustering of Disaster-Related Tweets using Bilingual Latent Dirichlet Allocation and Incremental Clustering Algorithm with Support Vector Machines for Need Assessment.”

Ponay and her team composed of Computer Science students, Lady Angelica Buen Guerzo, Hans Aaron Kilkenny, Raphael Noel Osorio, and Andrei Hart Villegas, studied the data clustering of social media posts on Twitter on the topic of disasters, including the classifications of tweets for prayers, cash, shelter, relief, and rescue. Due to the tendency of Filipinos to post in either Filipino or English as a majority, the system would normally have difficulty clustering data with the same disaster types, but in different languages.

To solve this bilinguality problem, the team employed a Bilingual Latent Dirichlet Allocation (BiLDA), which will help the system determine correctly whether a post is related to a disaster or not. It was found that the metrics precision, recall, F-measure, and area under the curve (AUC) of Incremental Clustering in the system, when it comes to clustering and classifying tweets according to its disaster type yielded an improvement on the overall accuracy.

This conference series served as a forum for an international community of researchers, practitioners and vendors on all aspects of Software Engineering and Computer Systems. The presenters were from countries including Malaysia, Indonesia, India, Oman, Saudi Arabia, Bangladesh and the Philippines.


Like this article?

Share on Facebook
Share on Twitter