AI-BASED SOLO TAXONOMY OBSERVATION METHOD AND PROCESS

Abstract

This research investigates the integration of artificial intelligence technologies with the Structure of Observed Learning Outcomes (SOLO) taxonomy to develop automated assessment and learning observation methodologies. The SOLO taxonomy, conceptualized by Biggs and Collis, provides a systematic framework for classifying learning outcomes based on the complexity of understanding demonstrated by learners. This study explores how AI algorithms, particularly machine learning and natural language processing techniques, can be leveraged to automatically classify and evaluate student responses according to SOLO taxonomy levels. Through extensive analysis of secondary literature and primary experimental data, this research identifies significant gaps in current implementations, particularly regarding the feature selection and dimensionality reduction techniques when processing educational datasets. The paper proposes a novel approach to feature selection that addresses these gaps, demonstrating improved classification accuracy and reduced computational requirements. Results indicate that applying optimized feature selection algorithms to educational omics datasets can significantly enhance the performance of AI-based SOLO taxonomy classification systems. This research contributes valuable insights for educational technology researchers and practitioners seeking to implement automated assessment systems aligned with established pedagogical frameworks.