Despite centuries of scientific observations and biodiversity researches, much of the knowledge remains locked in formats (printed texts, PDFs, or unstructured narrative content). To give an immediate access to data, we need to extract the data contained within the scientific publications so they can be mobilized as soon as published. The data liberated from the publications in biodiversity-related publications enable us to monitor the contribution of an institution or a publisher, to the understanding of the natural world. The core mission of Natural history institution is to contribute to the understanding of the natural world and to disseminate its knowledge. Scientific publications are part of the cycle of research, this is particularly true in taxonomy because taxonomic papers are not only legal documents that determine the legitimacy of a name of a (new) species, they also cover every creation, description, modification, and use of a taxon. It is crucial to link the information describing a taxon and its history, contained within the publication. This presentation will question three angles to stress out the need for data extraction: 1) Why do we need the data; 2) What data do we need; 3) How should we process in the future.