Abstract:
The structured description of competencies in educational resources is a growing demand,
driven by the adoption of competency-based curricula, such as the Computing
Curricula 2020 (CC2020) from the Association for Computing Machinery (ACM) and
the Institute of Electrical and Electronics Engineers (IEEE), and by the need to align
pedagogical practices with frameworks that integrate knowledge, skills, and attitudes.
Within this context, this dissertation proposes the Competency Specification Process
(CSP), a systematic methodological process for specifying and annotating competencies
within educational tasks in the field of Computer Science. Its central objective is to
promote greater clarity, reuse, and interoperability of educational data, thereby supporting
pedagogical planning and the personalization of learning. This research adopts the
Design Science Research (DSR) approach, structured around three interdependent cycles:
the relevance cycle (problem identification), the rigor cycle (theoretical foundation
and requirement definition), and the design cycle (development and evaluation). The
CSP was conceived based on the principles of CC2020 and Bloom’s Taxonomy (BT) and
was applied to tasks designed according to Problem-Based Learning (PBL) methodology,
previously used in Theory of Computation courses. For the semantic representation
of the annotated competencies, an extension to the Learning Object Metadata (LOM)
standard was proposed, named LOM-Competence (LOMc), formalized in the Resource
Description Framework (RDF). Evaluation by faculty in the field indicated that the CSP
facilitates precise and contextualized competency formulation, improves the alignment
between tasks and educational objectives, and contributes to more effective practices for
curating and recommending educational resources. The research also identifies challenges,
such as the need for manual review of annotations, suggesting the future integration
with educational ontologies to support process automation and scalability.
