Title: Map reduce-based scalable Lempel-Ziv and application in route prediction

Authors: Vishnu Shankar Tiwari; Sudha Chaturvedi; Arti Arya

Addresses: Visvesvarya Technological University, Belagavi, Karnataka, 590018, India ' Lingaya's Vidyapeeth, Faridabad, Haryana, 121003, India ' PES University, Hosur Road, Bengaluru, Karnataka, 560100, India

Abstract: Prediction of route based on historical trip observation of users is widely employed in location-based services. This work concentrates on building a route prediction system using Lempel-Ziv technique applied to a historical corpus of user travel data. Huge continuous logs of historical GPS traces representing the user's location in past are decomposed into smaller logical units known as trips. User trips are converted into sequences of road network edges using a process known as map matching. Lempel-Ziv is applied on road network edges to build the prediction model that captures the user's travel pattern in the past. A two-phased model is proposed using a map reduce framework without losing accuracy and efficiency. Model is then used to predict the user's end-to-end route given a partial route travelled by the user at any point in time. The objective of the proposed work is to build a Route Prediction system in which model building and prediction both are horizontally scalable.

Keywords: Lempel-Ziv; route prediction; horizontally scalable; map reduce; distributed computing.

DOI: 10.1504/IJBDI.2024.138934

International Journal of Big Data Intelligence, 2024 Vol.8 No.2, pp.89 - 99

Received: 04 Aug 2020
Accepted: 15 Nov 2021
Published online: 04 Jun 2024 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article