An approach to a digital library of newspapers

This paper presents a new application for retrieving news from a large electronic bank of newspapers. Specifically, the application is intended to manage past issues of newspapers in such a way that users are able to draw up chronicles and trends about reported topics. The article first focuses on modelling the structure and temporal properties of newspapers. The resulting data model is object-oriented and uses metaclasses to cope with all the requirements of the application. The proposed system is then completed with both a Document Definition Language and a temporal Document Retrieval Language. The former allows the user to declare all the features supported by our data model. For the latter, some query-processing strategies are also outlined in this paper.