Geospatial Knowledge in Housing Advertisements: Capturing and Extracting Spatial Information from Text - Archive ouverte HAL Access content directly
Conference Papers Year : 2021

Geospatial Knowledge in Housing Advertisements: Capturing and Extracting Spatial Information from Text

(1) , (2) , (1)
1
2

Abstract

Information of the geographical and spatial type is found in numerous text documents and constitutes a very challenging target for extraction. Geoparsing applications have been developed to extract geographic terms. However, off-the-shelf Named Entity Recognition (NER) models are mainly designed for Toponym recognition and are very sensitive to language specificity. In this paper, we propose a workflow to first extract geographic and spatial entities based on a BiLSTM-CRF architecture with a concatenation of several text representations. We also propose a Relation Extraction module, particularly aimed at spatial relationships extraction, to build a structured Geospatial knowledge base. We demonstrate our pipeline by applying it to the case of French housing advertisements, which generally provide information about a property's location and neighbourhood. Our results show that the workflow tackles French language and the variability and irregularity of housing advertisements, generalizes Geoparsing to all geographic and spatial terms, and successfully retrieves most of the relationships between entities from the text.
Fichier principal
Vignette du fichier
kcap-paper.pdf (510.05 Ko) Télécharger le fichier
Origin : Files produced by the author(s)

Dates and versions

hal-03518717 , version 1 (10-01-2022)

Identifiers

Cite

Lucie Cadorel, Alicia Blanchi, Andrea G. B. Tettamanzi. Geospatial Knowledge in Housing Advertisements: Capturing and Extracting Spatial Information from Text. K-CAP 2021 - International Conference on Knowledge Capture, Dec 2021, Virtual Event USA, United States. pp.41-48, ⟨10.1145/3460210.3493547⟩. ⟨hal-03518717⟩
66 View
131 Download

Altmetric

Share

Gmail Facebook Twitter LinkedIn More