According to Jurafsky and Martin (1999: 178), the segmentation is the process of taking undifferentiated sequence of symbols and segmenting it into meaningful linguistic units. As units can be defined the sentences as well as the word or the topic. It distinguishes for example between the sentence and word segmentation. The task of the sentence segmentation is to find the sentence boundaries in the text. Similarly the task of word segmentation is to split the text into word boundaries.

Depending on the language, the task of the segmentation can be more or less complicated.