An Optimized Question Classification Framework Using Dual-Channel Capsule Generative Adversarial Network and Atomic Orbital Search Algorithm

The advancement in education has emphasized the need to evaluate quality of examination questions and cognitive levels students. Many educational institutions now acknowledge Bloom’s taxonomy-based students’ evaluating subject-related learning. Therefore, this paper, a novel optimized Examination Question Classification framework, referred as QC-DcCapsGANAOSA, is proposed by combining Dual-channel Capsule generative Adversarial Network (DcCapsGAN) with Atomic Orbital Search Algorithm (AOSA) for preprocessing real-time online dataset university questions, thus identify key features from raw data using Term Frequency Inverse Document (TF-IDF) finally classifying questions. used fine-tune parameters’ weights DcCapsGAN, then uses these categorize Knowledge Level, Comprehension Application Analysis Synthesis Evaluation Level. Experimental results demonstrate superiority method (QC-DuCapsGAN-AOSA) when compared state-of-the-art methods such QC-LSTM-CNN QC-BiGRU-CNN an accuracy improvement 23.65% 29.04%, respectively.