In the master's thesis, we analyzed abstracts of articles on computational thinking. For this purpose, we used artificial intelligence methods, specifically machine learning, text mining, and semantic technologies, to gain insights into computational thinking concepts. We examined research articles on computational thinking from three well-known bibliographic databases: Google Scholar, Scopus, and Semantic Scholar.
In the theoretical part of the thesis, we described various definitions of computational thinking, as well as the skills and competencies it encompasses. We connected computational thinking skills with the goals outlined in the Slovenian curriculum for the elective and mandatory elective subject of computer science in primary school, as well as with the objectives set by the ACM K-12 curriculum. We described a computational analysis of text documents using the tool OntoGen, which allows for analysis of textual documents, document classification, document clustering, visualization.
In the empirical part of the thesis, we described the analysis of abstracts from selected bibliographic databases and the experimental results showing how the articles were grouped based on various factors: their target audience (computer science teachers, general educators, researchers-academics), the age group or educational level they address (primary school, secondary school, academic), the skills they cover (coding, algorithms, abstraction, decomposition, pattern recognition), the approaches they use (project-based learning, problem-based learning, inquiry-based learning, etc.), the environments they discuss (unplugged activities, block-based programming, text-based programming, physical computing), and the learning context (connection to the curriculum of a specific subject, cross-curricular integration, extracurricular activities). By analyzing the articles and grouping them, as well as identifying their key differences and similarities, we aimed to gain a broader understanding of the target audiences of these articles. The analysis revealed that Google Scholar is aimed at researchers, teachers, students, and all those searching for best practices in the research filed, while Scopus primarily serves researchers, academics, institutions, professionals, and policymakers working on academic research in the field of computational thinking. Furthermore, Semantic Scholar is intended mainly for professionals and new participants from various scientific and technical disciplines interested in integrating computational thinking into their professional practice. Scopus emphasizes academic research and the Computing Education Research (CER) community, offering a comprehensive overview. Google Scholar, on the other hand, focuses on practical aspects, making it a valuable resource for teachers implementing computational thinking in the classroom. Researchers can choose the source that best suits their needs: Scopus for academic research, Semantic Scholar for a general overview, and Google Scholar for practical classroom ideas.
|
