This paper discusses the impact of the "13th Five-Year Plan for Educational Informatization" on the development of online learning spaces, emphasizing their integrated, data-driven, intelligent, and personalized characteristics. At the same time, it identifies the challenges of insufficient learning motivation and lack of deep engagement among college English online learners and proposes a solution by applying teaching interventions based on online learning analytics to college English instruction. By collecting and analyzing learner behavior data, effective teaching interventions are implemented to stimulate motivation, enhance self-regulation, and ultimately improve adaptive learning capabilities and online learning outcomes.

1. Online Learning Analytics

This section introduces the origins, theories, model construction, and research on technologies and tools in learning analytics. Learning analytics involves data collection, analysis, evaluation, prediction, and intervention, with prediction and intervention being key components. It not only gathers learner behavior data but also collects big data from learning environments to construct a comprehensive learner model. The accuracy of the learner model directly affects the provision of personalized learning resources and pathways. Learning analytics is a cyclical process where teaching interventions and adaptive interventions influence learner data, forming a feedback loop. The technology aims to predict learning outcomes by integrating data to create knowledge, cognitive, affective, and behavioral models, enabling timely interventions to enhance learning effectiveness.

2. Teaching Intervention Framework Based on Learning Analytics

This section presents a teaching intervention framework based on learning analytics, designed to address issues such as low engagement and procrastination among online learners. The framework collects learner data, analyzes and predicts outcomes, and provides feedback to both learners and instructors, facilitating self-regulation and human intervention. Specifically targeting college English blended courses, the "Learning Pass" platform is utilized to enable diversified knowledge delivery and unrestricted learning across time and space. The prediction focuses on non-English major college English learners, with a teaching intervention mechanism to promptly identify and address problems. Teaching interventions are categorized into traditional and online types, with learning analytics constructing learner models and processing collected data to enhance intervention effectiveness.

3. College Students' Autonomous English Learning Engagement

This section reviews domestic and international research on learning engagement, covering four aspects: conceptualization, structure, correlations, and measurement. Conceptual research interprets learning engagement from multiple perspectives, while structural research divides online learning engagement into behavioral, cognitive, and affective dimensions, introducing concepts such as agentic engagement and metacognitive engagement. Correlation studies indicate that learning engagement is positively associated with academic performance and higher-order skill development, serving as a key factor in addressing issues like learning burnout. In terms of measurement, domestic scholars have developed a distance learning engagement scale based on classic foreign instruments. This study constructs the "College Students' English Online Learning Engagement Scale," comprising behavioral, affective, cognitive, and agentic engagement dimensions, to assess autonomous learning engagement and influencing factors among college English learners.

4. Empirical Study on Teaching Interventions

This section introduces an empirical study on the impact of teaching interventions on autonomous online learning engagement. The participants were first-year students from the School of Management enrolled in a blended College English (II) course, divided into an experimental class and a control class. The experimental class received online teaching interventions based on learning analytics, while the control class used traditional intervention methods. The research steps included collecting student behavior data and conducting questionnaires, primarily employing data mining and descriptive statistical analysis. Findings revealed that after online interventions, the experimental class showed increased participation, particularly in login frequency, preview activities, and activity engagement, though interaction, persistence, and focus exhibited minimal change. Questionnaire results indicated that behavioral engagement was a key factor in online learning, while agentic engagement significantly influenced learning outcomes, with affective and cognitive engagement having lesser effects. The experimental class demonstrated notable improvements in behavioral and agentic engagement compared to the control class, but changes in affective and cognitive engagement were not statistically significant. This suggests that online learning engagement has transcended traditional definitions, exhibiting an ecological development trend. Teaching interventions based on learning analytics help bridge the gap between learners and systems, enhancing engagement. Future intervention mechanisms should place greater emphasis on learners' affective engagement for further improvement.

5. Conclusion and Future Prospects

In recent years, college English teaching reforms have made online learning a mainstream approach. However, online instruction often overlooks individual differences. Learning analytics technology enables online interventions, constructs personalized models, optimizes learning strategies, and enhances adaptive learning capabilities. Both online and traditional learning models hold potential in affective and cognitive engagement. Improving teaching intervention mechanisms requires leveraging smart education platforms and educational cloud resource libraries, advancing the development of educational management information systems and technological updates to boost learning engagement. This imposes higher demands on educational informatization and presents challenges for educators and administrators. Language educators must strengthen cross-disciplinary collaboration, improve big data competencies, and research the effectiveness of teaching interventions in practice.