Normalizing Spatial Information to Better Combine Criteria in Geographical Information Retrieval
SALLABERRY, Christian
Laboratoire Informatique de l'Université de Pau et des Pays de l'Adour [LIUPPA]
Laboratoire Informatique de l'Université de Pau et des Pays de l'Adour [LIUPPA]
GAIO, Mauro
Laboratoire Informatique de l'Université de Pau et des Pays de l'Adour [LIUPPA]
Linguistic signs, grammar and meaning: computational logic for natural language [SIGNES]
Laboratoire Informatique de l'Université de Pau et des Pays de l'Adour [LIUPPA]
Linguistic signs, grammar and meaning: computational logic for natural language [SIGNES]
SALLABERRY, Christian
Laboratoire Informatique de l'Université de Pau et des Pays de l'Adour [LIUPPA]
Laboratoire Informatique de l'Université de Pau et des Pays de l'Adour [LIUPPA]
GAIO, Mauro
Laboratoire Informatique de l'Université de Pau et des Pays de l'Adour [LIUPPA]
Linguistic signs, grammar and meaning: computational logic for natural language [SIGNES]
< Réduire
Laboratoire Informatique de l'Université de Pau et des Pays de l'Adour [LIUPPA]
Linguistic signs, grammar and meaning: computational logic for natural language [SIGNES]
Langue
en
Communication dans un congrès
Ce document a été publié dans
ECIR GIIW: 31st European Conference on Information Retrieval, Geographic Information on the Internet Workshop (GIIW), ECIR GIIW: 31st European Conference on Information Retrieval, Geographic Information on the Internet Workshop (GIIW), ECIR, 31st European Conference on Information Retrieval, 2009-04-06, Toulouse. 2009-04-06p. 37-49
Résumé en anglais
It is generally accepted that geographical information or G.I. (such as texts, maps and tables) is chiefly composed of 3 kinds of criteria : spatial, temporal and thematic criteria. The main focus of this article is spatial ...Lire la suite >
It is generally accepted that geographical information or G.I. (such as texts, maps and tables) is chiefly composed of 3 kinds of criteria : spatial, temporal and thematic criteria. The main focus of this article is spatial criteria. More specifically, we have developed a processing sequence that can extract the spatial information contained in non-structured cultural heritage texts. This processing sequence indexes spatial information, which enables information retrieval (I.R.) based on the same criteria. Our goal is to normalize heterogeneous spatial information. This normalization is carried out at the index level by grouping spatial information together and by using statistics to calculate weights of spatial areas and the pertinence of the results. Thus, we aim to develop a general IR strategy that is dedicated to spatial information, but which can be applied to temporal and thematic information as well. By generalizing this approach, homogeneous IR strategies will be able to combine spatial, temporal and thematic criteria for more efficient geographic IR methods.< Réduire
Origine
Importé de halUnités de recherche