Source-linked AI summary
RIACT: A Responsible AI System for Personalized Study Habit Tracking and Early Burnout Signal Detection in University Students
Ria Sidhu
TL;DR
Student burnout is prevalent but typically identified retrospectively, while students and existing tools have limited visibility into individualized study behaviour. RIACT combines structured session logging with deterministic burnout rules and constrained language-model interpretation to provide personalized, non-diagnostic guidance. The paper presents the design and a framework for future validation, but its thresholds and signals have not yet been empirically validated.
Problem
Student burnout is prevalent and usually recognized retrospectively, while existing productivity tools record activity without interpreting students’ individual study behaviour.
Method
RIACT combines structured logging of study location, timing and breaks with deterministic week-over-week burnout rules and constrained AI for pattern interpretation and recommendations.
Results
RIACT’s design combines personalized behavioural feedback with auditable warnings, non-diagnostic language and limited self-logged data collection.
Takeaways & Limitations
RIACT provides a foundation for responsible, personalized study-habit and student-wellbeing research, with empirical validation left for future work.
Takeaways & Limitations
RIACT has not been evaluated with a real user cohort, and its burnout thresholds have not been validated against established clinical measures.
Abstract
from arXiv · showhide
Student burnout is highly prevalent in higher education, with reported rates ranging from 12% to over 70% and consistently exceeding those of the working population - yet it is typically identified only retrospectively, after academic decline has already occurred. A contributing factor is that students have little structured visibility into their own study behaviour, and existing productivity tools record activity without interpreting it. This paper presents RIACT (Record, Insight, Analyze, Coach, Track), a web-based application that combines structured study session logging with a hybrid AI architecture to surface personalized insights and early burnout signals. Students log sessions by location and time; the system computes net focus time by accounting for breaks, detects burnout signals through transparent, deterministic rules operating on week-over-week behavioural comparisons, and uses a large language model - constrained to a fixed output schema - to contextualize patterns and generate personalized recommendations. The design embeds responsible AI principles throughout: warnings are governed by auditable rules rather than model judgement, all output is framed as an observation rather than a diagnosis and data collection is limited to self-logged behavioural fields. We describe the system's design rationale, situate it within the literature on student burnout and explainable AI in education and propose an evaluation framework for validating its behavioural signals against established burnout instruments.
1 Introduction
Student burnout is widespread yet usually recognized only after harm has occurred, while existing tools record study activity without interpreting individual patterns. RIACT responds with structured logging and personalized, responsible AI-supported insights.
- The problem: 12% to over 70%: reported university-student burnout prevalence exceeds that of the working population.Burnout is commonly identified retrospectively, after it has already taken hold.
- The problem: Students lack structured visibility into where, when and how effectively they study, limiting recognition of unproductive habits and warning signs.Existing trackers record activity but do not provide personalized analysis tied to individual learning outcomes.
- Why structured logging matters: Memory-based self-monitoring distorts estimates of study duration, break frequency and gradual behavioural change.Structured logging externalizes the record and enables week-over-week comparison.
- Existing tools: Commercial productivity tools support time tracking and reports but generally record activity without interpreting its relationship to student wellbeing.RIACT is motivated by the absence of burnout detection, individualized recommendations and academic-context-aware tracking.
- RIACT: RIACT is a web application that logs study sessions by location and time, computes net focus around breaks, and uses constrained AI to surface patterns, burnout signals and recommendations.Its outputs are observations rather than diagnoses, and students retain control over resulting decisions.
2 Related Work
Prior research establishes student burnout as prevalent, multidimensional and gradually developing, while also identifying risks and explainability needs for educational AI. RIACT positions its behaviour-focused, constrained architecture as a response to these gaps.
- Student burnout: Student burnout is measured across emotional exhaustion, cynicism and reduced academic efficacy in university populations.The student adaptation interprets these dimensions through study fatigue, mental distancing and diminished academic competence.
- Student burnout: 12% to over 70%: student burnout prevalence varies by population and methodology and consistently exceeds rates in the working population.A German study found stress and burnout across five academic fields, with informatics and engineering among the highest.
- Student burnout: Burnout develops gradually, and behavioural decline may be detectable before burnout becomes acute.RIACT therefore treats declining session length, rising break frequency and late-night displacement as proxies for eroding engagement.
- AI in education: Open-ended educational AI can encourage overreliance because it is disconnected from the student’s actual learning context.RIACT constrains AI to analysing logged behaviour, so its output informs study decisions rather than substituting for studying.
- Ethics and responsibility: Explainable educational AI requires transparent decisions that learners can trust and act upon.RIACT applies this principle through deterministic burnout warnings, observations rather than diagnoses, minimal data collection and professional-resource links without clinical claims.
3 System Design
RIACT combines minimal structured session data with a hybrid AI layer for personalized study insights and early burnout signals. Responsible design separates auditable warnings from model interpretation while preserving user control and non-clinical framing.
- System architecture: RIACT is a web application for understanding study habits through structured session logging and AI-generated insights.Its implementation uses Next.js, Supabase and the OpenAI API, with row-level security protecting user records.
- Session tracking and data model: Net study time subtracts recorded break duration from total session time, providing a richer focus signal than raw duration alone.Sessions capture location, timing, projected end time, elapsed minutes and break frequency.
- Session tracking and data model: RIACT collects only where, when and how continuously students study, excluding biometric data, device telemetry and study content.The minimal schema is intended to support behavioural inference without invasive data collection.
- AI features: After three logged sessions, dashboards, pattern insights, coaching and weekly plans use the student’s accumulated history and goals.Additional sessions refine the system’s picture of individual habits and recommendations.
- Session workflow: Projected end times anchor commitment, while repeated early endings can become a behavioural signal of declining engagement.Students can review and edit session records before saving, including breaks and computed net study time.
- Burnout detection: Burnout warnings come from deterministic rolling week-over-week rules, while the language model contextualizes patterns and recommendations.The rules evaluate declining session length, rising break frequency, late-night studying and declining goal completion; any triggered signal is shown with the warning.
- AI features: Pattern recognition compares locations and times of day using net study time and break frequency to produce actionable individual insights.Personalized recommendations are based on historical behaviour rather than generic advice.
- Responsible AI design: RIACT presents observations rather than diagnoses, makes no clinical claims or treatment recommendations, and keeps AI advisory rather than gatekeeping.Students retain control over data review, editing, goal management and decisions following AI output.
4 Discussion
RIACT’s discussion frames personalization and continuous behavioural monitoring as alternatives to generic advice and retrospective burnout identification, while acknowledging design tradeoffs and unresolved validation questions.
- Why Personalization Matters: Personalized recommendations use each student’s behavioural history rather than population averages, creating a feedback loop that can become more informative as logging accumulates.The approach aims to make recommendations actionable and falsifiable through continued observation of whether individual patterns hold.
- Early Burnout Detection as a Preventive Tool: RIACT monitors continuous study-behaviour shifts between formal assessments, but presents resulting burnout prompts as self-reflection rather than clinical support.The monitored shifts include declining session length, increasing break frequency and late-night cramming.
- Design Tradeoffs: The system’s design trades broader subject-level context for lower logging burden and comparable location and time-of-day data across a student’s history.Location-based tracking cannot distinguish subjects with different cognitive demands, while manual logging introduces self-report bias and friction but avoids passive-surveillance costs.
- Design Tradeoffs: Three sessions balance statistical caution with user retention, so early AI insights are framed as provisional observations rather than conclusions.A higher threshold could improve reliability but might delay users’ experience of the system’s value.
- Limitations: The principal limitation is that RIACT has not been evaluated with a real user cohort, leaving its behavioural thresholds unvalidated against established clinical measures.The proposed thresholds could be either too sensitive or too conservative until empirical calibration against self-reported outcomes is conducted.
- Limitations: Self-reported logging can degrade insight quality, and disengaged students may stop logging before they stop studying, reducing visibility when warnings would matter most.The system currently has no mechanism for detecting logging quality.
- Evaluation Considerations: Evaluation must separately test sustained use, signal validity and outcome change over sufficiently long periods and full academic cycles.Short windows risk overstating viability because novelty effects and exam-period workload fluctuations can distort behavioural patterns.
- Evaluation Considerations: A controlled comparison with a logging-only condition could isolate the AI insight layer from the reflective effects of logging itself.The proposed next step is a full-term pilot measuring study consistency, burnout awareness or goal completion against a control group.
5 Conclusion
RIACT addresses persistent student burnout and ineffective study habits by translating individual behavioural data into personalized, actionable guidance. The paper presents the system as a responsible foundation for future research, while empirical validation remains future work.
- RIACT combines structured session logging with AI-driven pattern recognition, burnout detection, and coaching grounded in each student’s history.The approach is intended to replace generalized advice with personalized guidance based on students’ own behavioural data.
- The system’s responsible AI design supports potential deployment with a vulnerable student population by framing feedback as data-driven guidance.The paper describes RIACT as openly accessible and as a foundation for future research into AI-assisted student wellbeing.