PERE LAPERNE, Jacques

Voir/Ouvrir

(A)KDD for Structuring Destructured Documents ICAI'18 (1).pdf (1.004Mo)

Métadonnées

Afficher la notice complète

Licence d’utilisation du document

PERE LAPERNE, Jacques
ESTIA INSTITUTE OF TECHNOLOGY

Langue

Communication dans un congrès

Ce document a été publié dans

Proceedings of the 2018 International Conference on Artificial Intelligence ICAI'18, 2018 International Conference on Artificial Intelligence ICAI'18, 2018-07-30, Las Vegas, NV. 89.

Résumé en anglais

The worldwide volume of digital data doubles every 9 months. Over 75% of these data are unstructured. This paper concerns structuring the graphic information contained in vector files, including PDF (Portable Document Format) files, which represent a very significant share of all these vector files. We say these data are destructured because they are today produced by software. When there are stored or exchanged, (for example in PDF Files) only graphic information is kept and the data structure use to create the document is lost To structure these data, we use Knowledge Discovery in Databases (KDD). The following two issues arise: • Can the KDD method be adapted to “Structuring Destructured Documents”?• If so, what points need adapting or highlighting in the method to solve the issue of structuring essentially graphic documents?The answer to the questions is “YES” with the human in the loop. This is why we talk about the Anthropocentric Knowledge Discovery in Database method, abbreviated to (A)KDD.< Réduire

Mots clés en anglais

CHI

PDF

KDD

Graphic reconstruction

Pattern recognition

Data Minning

URI

https://oskar-bordeaux.fr/handle/20.500.12278/200057

Unités de recherche

ESTIA - Recherche