-
学术期刊
-
学术著作
-
精品图书
-
投稿须知
International Education and Social Science Exploration (IESSE) June 2026, Vol.2, No.3
Construction and Practice of an Intelligence-Driven Big-Data-Based Precision Instructional Intervention System
Fang He, Shuanghao Fan, Yiwei Wei, Zhengzhong Cha, Qingyang Zheng
Rocket Force University of Engineering, Xi’an 710025, Shaanxi, China
Abstract: The digital-intelligence transformation of higher education is shifting instructional practice from experience-based judgement, uniform provision, and remedial action after failure toward multidimensional sensing, intelligent diagnosis, dynamic adaptation, and continuous support. Conventional instructional intervention relies heavily on test scores, classroom observation, and teachers’ personal experience, and therefore suffers from partial learner-state identification, delayed problem detection, weak causal diagnosis, homogeneous intervention, and insufficient outcome tracking. Synthesising three complementary lines of work, this paper defines big-data-based precision instructional intervention as a student-centred, closed-loop process that continuously integrates learner, knowledge, behavioural, task, contextual, and assessment data and uses learning analytics, knowledge tracing, causal inference, predictive modelling, generative artificial intelligence, and human–AI collaboration to sense learning states, identify risks, diagnose causes, match interventions, and evaluate effects. The paper proposes a “one core, four intelligences, five loops, and six dimensions” framework. Student development constitutes the core; data-sensing intelligence, cognitive-diagnostic intelligence, predictive-simulation intelligence, and generative-decision intelligence form the intelligent chain; state sensing, risk identification, causal diagnosis, intervention matching, and effect evaluation constitute the five-loop cycle; and six evidence dimensions cover learner, knowledge, behaviour, task, context, and assessment. Practice mechanisms are designed for pre-class anticipatory intervention, in-class real-time regulation, post-class targeted reinforcement, process intervention in projects and practica, and preventive support for developmental risk. Implementation paths include data standards, multi-graph association, multi-agent collaboration, teacher digital-intelligence competence, quasi-experimental evaluation, and ethical governance. The study argues that precision does not depend on collecting more data, but on improving educational interpretability, intervention fit, and continuous learning from effects so that technology enables earlier, more accurate, and more humane educational responses.
Keywords: big data, precision instructional intervention, artificial intelligence, learning analytics, knowledge tracing, digital-intelligence education
