This article describes a simple unsupervised system for automatic extraction and classification of named entities in French novels. The solution presented combines a set of different standalone classifiers within a meta-recognition system. The system is tested on 35 classic French novels, representing 5 million words and 3,700 names of people and places. The results demonstrate that although none of the standalone methods clearly outperforms the others, their combined classification offers a robust solution in this context.

This paper presents a methodology to analyze linguistic changes in a given textual corpus allowing to overcome two common problems related to corpus linguistics studies. One of these issues is the monotonic increase of the corpus size with time, and the other one is the presence of noise in the textual data. In addition, our method allows to better target the linguistic evolution of the corpus, instead of other aspects like noise fluctuation or topics evolution. A corpus formed by two newspapers "La Gazette de Lausanne" and "Le Journal de Genève" is used, providing 4 million articles from 200 years of archives. We first perform some classical measurements on this corpus in order to provide indicators and visualizations of linguistic evolution. We then define the concept of a lexical kernel and word resilience, to face the two challenges of noises and corpus size fluctuations. This paper ends with a discussion based on the comparison of results from linguistic change analysis and concludes with possible future works continuing in that direction.

No abstract