Optimizing Vocational English Writing Teaching with Computer-Assisted Feedback: A Big Data Platform Approach
Year: 2025
Type: conference-paper
Study designs: Randomized Controlled Trial, Clinical Trial
Abstract: This study addresses the challenges of delayed feedback (>72 hours) and lack of personalization in vocational English writing instruction by developing a big data-powered computer-assisted teaching platform. Through a four-layer closed-loop architecture (data collection-intelligent processing-decision engine-instructional application), it integrates three core innovations: an AI feedback engine enhanced with industry corpora (92.6% accuracy) that dynamically extracts cross-industry terms from 1.2 million job descriptions (JD); a 9-dimensional competency digital profiling system (grammar/terminology/logic) enabling precise capability modeling through error data inversion; and a reinforcement learning-driven dynamic teaching strategy algorithm (α=0.7 error adaptation, β=0.3 preference weight) that boosts learning efficiency by 41.8%. A 16-week randomized controlled trial (N=244) showed learners reduced grammatical errors by 37.2%, increased industry term usage by 2.8 times; teachers improved grading efficiency by 68% (from 12.5 minutes per essay to 4.0 minutes); and self-efficacy significantly increased (+19.4%, p<0.05). The platform ensures data security and algorithm transparency through federated learning and SHAP explainability technology, establishing a new paradigm of “data-driven, real-time feedback, and dynamic optimization” for vocational English writing instruction.
Author: Chunyan Hao
Institution: Qingdao Technical College
Language: English
Subfield: Education
Field: Social Sciences
Domain: Social Sciences
Open Access status: gold
OpenAlex ID: W7155406293
