Can We Predict Your Next Move Without Breaking Your Privacy?

TL;DR

This paper introduces FLLL3M, a privacy-preserving federated learning framework utilizing large language models for accurate next-location prediction without compromising user privacy.

Contribution

It presents a novel federated learning approach with an efficient mechanism to leverage LLMs for mobility modeling while maintaining privacy and reducing resource consumption.

Findings

Achieves high accuracy on multiple mobility datasets.

Reduces model parameters by up to 45.6%.

Lowers memory usage by 52.7%.

Abstract

We propose FLLL3M--Federated Learning with Large Language Models for Mobility Modeling--a privacy-preserving framework for Next-Location Prediction (NxLP). By retaining user data locally and leveraging LLMs through an efficient outer product mechanism, FLLL3M ensures high accuracy with low resource demands. It achieves SOT results on Gowalla (Acc@1: 12.55, MRR: 0.1422), WeePlace (10.71, 0.1285), Brightkite (10.42, 0.1169), and FourSquare (8.71, 0.1023), while reducing parameters by up to 45.6% and memory usage by 52.7%.