The rapid proliferation of mobile applications (apps) has led to a significant shift in how individuals interact with technology and consume digital content. Understanding user app usage behaviour is crucial for app developers, marketers, and researchers to optimize user experience and engagement. However, it is difficult to obtain large-scale app usage data due to privacy concerns. This proposed student research project aims to leverage Large Language Model (LLM) generative agents to simulate and analyze app usage behaviour patterns. Similar to a recent work (Generative Agents: Interactive Simulacra of Human Behaviour), powered by LLMs such as GPT and LLAMA, we can simulate a few interactive agents. This project will then focus on the app usage behaviours of the simulated agents. We are also interested in using open-sourced LLMs such as LLAMA2 so that we can fine-tune the model with real-world app usage behaviour data if needed. This project requires experience with LLM-related tools such as HuggingFace. 


Computer Science and Engineering

Research Area

Deep learning | Spatio-temporal data | Large language models

The research will be conducted at the School of CSE together with the supervision team.

  • The outcomes by the end of the project will include a technical report, a code package, and a video demo (to explain the projects, technologies, and achievements). We are planning to extend or write outcomes as a technical paper targeting a publication in a top venue.
Professor and Cisco Chair in Digital Transport
View Profile
View Profile
View Profile