Production of origin destination matrix by extracting information from unstructured textual data

Abstract

In this paper, we present an approach for the production of origin destination matrices by extracting information from unstructured textual data of websites. This approach, which we called “Origin Destination Matrix Extractor” is based on three main modules of Information Extraction: an extraction of events module with which we tried to extract any travel events contained in a given text, a named entity recognition module for the recovery and detection of named entities that correspond to the different target information and finally a dependency syntactic analysis module to check the existence of interdependencies between extracted entities and detected travel events. The experiments carried out on a set of real data show that the proposed method gives satisfactory results with a precision of over 90%.