Address Validation in Transportation and Logistics: A Machine Learning Based Entity Matching Approach

Yassine Guermazi (1, 2, 3) , Sana Sellami (1, 2, 3) , Omar Boucelma (1, 2, 3)

1 DIAMS - Data Integration, Analysis, and Management as Services
2 LIS - Laboratoire d'Informatique et des Systèmes (LIS) (Marseille, Toulon)
3 AMU - Aix Marseille Université

In the Transportation and Logistics (TL) industry, address validation is crucial. Indeed, due to the huge number of parcel shipments that are moving worldwide everyday, incorrect addresses generates several shipment returns, leading to useless financial and ecological costs. In this paper, we propose an entity-matching approach and system for validating TL entities. The approach is based on Word Embedding and Supervised Learning techniques. Experiments carried out on a real dataset demonstrate the effectiveness of the approach.

Mots clés: Transport and logistics; Word embedding; Supervised learning; Entity matching; Transport and logistics

Domaines: Apprentissage [cs.LG]; Autre

Yassine Guermazi, Sana Sellami, Omar Boucelma. Address Validation in Transportation and Logistics: A Machine Learning Based Entity Matching Approach. Communications in Computer and Information Science, 2020, ECML PKDD 2020 Workshops, 1323, pp.320 - 334. ⟨10.1007/978-3-030-65965-3_21⟩. ⟨hal-03181312⟩