# Exploring the Feasibility of Automated Data Standardization using Large Language Models for Seamless Positioning

**Authors:** Max J. L. Lee, Ju Lin, Li-Ta Hsu  
**arXiv:** [2408.12080](https://arxiv.org/abs/2408.12080)  
**Date:** 2024-08-22  
**Summary page:** https://www.summarizepaper.com/en/arxiv-id/2408.12080v1/

## Key points

- Study explores use of Large Language Models (LLMs) for data standardization and positioning accuracy in IoT environments
- Framework integrates sensor data from smartphones, IoT devices, and UWB systems using Extended Kalman Filter (EKF)
- Intelligent Data Standardization Module (IDSM) powered by LLM converts varied sensor data into standardized format
- Transformation Rule Generation Module (TRGM) automates generation of transformation rules and scripts for continuous data standardization
- Real-time evaluations demonstrate adaptability and scalability, improving operational efficiency and navigation accuracy
- Research showcases potential of advanced LLMs in addressing challenges with sensor data integration for precise IoT navigation solutions
- Feasibility study accepted at IPIN 2024 and set to be published in IEEE Xplore, highlighting significance in advancing technological innovations within positioning systems.

## Summary

Our study explores the use of Large Language Models (LLMs) to automate data standardization and improve positioning accuracy in IoT environments. By integrating diverse sensor data from smartphones, IoT devices, and specialized systems like Ultra-Wideband (UWB), our proposed framework ensures data compatibility and enhances navigation experiences through the utilization of the Extended Kalman Filter (EKF). The Intelligent Data Standardization Module (IDSM), powered by a finely-tuned LLM, converts varied sensor data into a standardized format. Additionally, the Transformation Rule Generation Module (TRGM) automates the generation of transformation rules and scripts for continuous data standardization processes. Through real-time evaluations in dynamic environments, our study demonstrates adaptability and scalability, resulting in improved operational efficiency and heightened accuracy in seamless navigation experiences. By showcasing the potential of advanced LLMs in mitigating challenges associated with sensor data integration complexities, our research paves the way for more scalable and precise solutions for IoT navigation. Accepted at IPIN 2024 and set to be published in IEEE Xplore, this feasibility study underscores its significance in advancing technological innovations within positioning systems.