Search for Articles:

MECS Press Menu

Yih-Chang Chen

Work place: Department of Social Work, Chang Jung Christian University, Tainan 711, Taiwan, China

E-mail: cheny@mail.cjcu.edu.tw

Website:

Research Interests: Artificial Intelligence

Biography

Yih-Chang Chen holds a Master of Science in Information Systems Security from the London School of Economics and Political Science (LSE), University of London, as well as a Doctorate in Computer Science from the University of Warwick, United Kingdom. He presently serves as an Assistant Professor in both the Department of Social Work and the Department of Information Management at Chang Jung Christian University in Taiwan. His academic expertise encompasses a range of interdisciplinary fields, including artificial intelligence (AI), software engineering, machine learning, social media applications, social work management, and long-term care.
His research is motivated by a dedication to bridging the divide between technology and its social applications. He actively participates in cross-disciplinary initiatives that merge information technology, management science, and social welfare to address complex societal and healthcare issues. In addition to his academic duties, he is currently an Audit Committee Member in the President Office at Chang Jung Christian University and leads projects under the auspices of Taiwan’s Ministry of Labor, specifically aimed at the development and implementation of Employment-Oriented Curriculum Programs.

Author Articles
Empowering Community Work: Using Semi-Supervised Learning to Identify Emerging Community Needs and Service Gaps from Massive Unstructured Text

By Yih-Chang Chen

DOI: https://doi.org/10.5815/ijisa.2026.02.01, Pub. Date: 8 Apr. 2026

Community engagement is essential to social service delivery, yet traditional community needs assessment remains time-consuming and poorly suited for timely monitoring. This study proposes a semi-supervised learning framework to identify emerging community needs and service gaps from massive, mostly unlabeled, unstructured text. We construct an explicit heterogeneous text graph where each record is a document node linked to keyword and need-category nodes; document–document edges are built using a weighted combination of semantic similarity (BERT cosine), lexical overlap (keyword Jaccard), and temporal proximity. A graph neural network with iterative self-training leverages 3% expert-labeled seed data and the remaining unlabeled corpus to classify records into a 10-category need taxonomy. On 176,602 records, the proposed model achieves F1 = 0.895 and Recall = 0.899, outperforming supervised baselines trained on the same labeled ratio by 23.8% (macro-F1). Post-hoc quarterly aggregation of predictions enables trend monitoring and prioritization of service-gap severity for decision support.

[...] Read more.
Other Articles

Subscribe to receive issue release notifications and newsletters from MECS Press journals