Bridging Academic Boundaries: A Design and UAT Study of the Automated Credit Transfer System (ACTS)

Abstract

This paper presents the design and User Acceptance Testing (UAT) of a Natural Language Processing (NLP)-driven Automated Credit Transfer System (ACTS) developed to automate and digitalize the academic credit transfer process at Universiti Poly-Tech Malaysia (UPTM). The system matches courses using TF-IDF and cosine similarity algorithms to assess syllabus alignment and determine credit eligibility. The design phase was afforded the highest priority to develop usability, transparency and efficiency per institutional needs. 20 members of the Faculty of Computing & Multimedia (FCOM), who acted as Resource Persons (RP), attended the UAT to test the system's usability, correctness and functionality using real-time examples. A Technology Acceptance Model (TAM)-based questionnaire and semi-structured interviews were conducted for data gathering. Results are rated very high for end-user satisfaction and perceived usefulness, implying both support for the broader institutional feasibility of implementation and the priority it deserves. Results highlight the importance of including end-users in the design and evaluation phases and call for redesigning and improving adoption and performance.