Synthetic Data Generation for Screen Time and App Usage
On the map: Synthetic data and self-improvement for LLMs → LLM synthetic data generation
Abstract (arXiv)
Smartphone usage data can provide valuable insights for understanding interaction with technology and human behavior. However, collecting large-scale, in-the-wild smartphone usage logs is challenging due to high costs, privacy concerns, under representative user samples and biases like non-response that can skew results. These challenges call for exploring alternative approaches to obtain smartphone usage datasets. In this context, large language models (LLMs) such as Open AI's ChatGPT present a novel approach for synthetic smartphone usage data generation, addressing limitations of real-world data collection. We describe a case study on how four prompt strategies influenced the quality of generated smartphone usage data. We contribute with insights on prompt design and measures of data quality, reporting a prompting strategy comparison combining two factors, prompt level of detail (describing a user persona, describing the expected results characteristics) and seed data inclusion (with versus without an initial real usage example). Our findings suggest that using LLMs to generate structured and behaviorally plausible smartphone use datasets is feasible for some use cases, especially when using detailed prompts. Challenges remain in capturing diverse nuances of human behavioral patterns in a single synthetic dataset, and evaluating tradeoffs between data fidelity and diversity, suggesting the need for use-case-specific evaluation metrics and future research with more diverse seed data and different LLM models.
Key facts. No citing papers in the map yet (data through 14 Oct 2026). Placed in LLM synthetic data generation, one of 13 directions in Synthetic data and self-improvement for LLMs. Builds on 13 of its 29 references that are in the map. First posted on arXiv 17 Sep 2025.
Builds on
13 of 29 references.
Closest ancestors · 1
Same direction, newest first
Other references · 12
- 2020Language Models are Few-Shot Learnerscited by 32,419
- 2023GPT-4 Technical Reportcited by 18,396
- 2024GPT-4o System Cardcited by 4,588
- 2021Prompt Programming for Large Language Models: Beyond the Few-Shot Paradigmcited by 606
- 2025LLM Social Simulations Are a Promising Research Methodcited by 99
- 2022Prototypical Verbalizer for Prompt-based Few-shot Tuningcited by 53
- 2024Take Caution in Using LLMs as Human Surrogates: Scylla Ex Machinacited by 29
- 2023Mixture of Soft Prompts for Controllable Data Generationcited by 23
- 2023MindShift: Leveraging Large Language Models for Mental-States-Based Problematic Smartphone Use Interventioncited by 16
- 2024Time2Stop: Adaptive and Explainable Human-AI Loop for Smartphone Overuse Interventioncited by 5
- 2025Scrolling in the Deep: Analysing Contextual Influences on Intervention Effectiveness during Infinite Scrolling on Social Mediacited by 5
- 2024What Social Media Use Do People Regret? An Analysis of 34K Smartphone Screenshots with Multimodal LLMcited by 1
Built on by
Not cited yet.
Similar, not cited
Close in wording, no citation either way
- 2026DynamicMem: A Long-Horizon Memory Benchmark in Real-World Settings
- 2024A Population-to-individual Tuning Framework for Adapting Pretrained LM to On-device User Intent Prediction
- 2025Guided Persona-based AI Surveys: Can we replicate personal mobility preferences at scale using LLMs?
- 2026ProfileFoundry: A Synthetic Person-Object Substrate for Privacy, Memory, and Tool-Use Evaluation in LLM Agent
- 2026PersonaTrace: Synthesizing Realistic Digital Footprints with LLM Agents
- 2024Be More Real: Travel Diary Generation Using LLM Agents and Individual Profiles
- 2024Best Practices and Lessons Learned on Synthetic Data
- 2025Show or Tell? Modeling the evolution of request-making in Human-LLM conversations