Abstract
This paper presents a topical text segmentation method based on intended boundaries detection and compares it to a well known default boundaries detection method, c99. We compared the two methods by running them on two different corpora of French texts and results are evaluated by two different methods: one using a modified classic measure, the FScore, the other based on a manual evaluation one the Internet. Our results showed that algorithms that are close when automatically evaluated can be quite far when manually evaluated.