4. Megadata analysis

Having dealt with the storage and access of megadata, we now turn our attention to their processing, in order to describe them (descriptive analysis) or model them (modeling, predictive analysis, generation of synthetic data). The nature of this analysis, also known as "analytics", depends not only on its purpose, but also on the megadata itself and the rate at which it is produced. Applications range from decision support, information retrieval, prediction, filtering, visualization and automatic classification, to automatic dashboard generation, content recommendation systems and suggested followings for social networks. When the data is textual or linguistic, there is a strong link with the field of Automatic Language Processing (sentiment and opinion analysis, automatic summarization, translation, conversational agents and voice assistants, etc.).

In what follows,...