Aplicaciones de técnicas de extracción de informacion a bibliotecas digitales = Information Extraction to feed Digital Library

Todos os direitos pertencem a Biblioteca Digital Miguel de Cervantes

Most often, Digital Libraries have the need to extract information

from poorly marked-up documents to fill databases or create new hypertext

documents with a highly structured markup. In this work, we

approach the problem of extracting bibliographic information from literary

reports in HTML format to fill a Digital Library database of

Galician publications used for Internet searchs. An information extraction

approach that takes advantage of both HTML markup and

Natural Language Processing (NLP) techniques was successfully used

for this purpose