While jellyfish are a natural feature of European Seas, massive swarms of gelatinous organisms have become a frequent sight in coastal waters leading to important socio-economic impacts in recent decades. From an ecosystem perspective, jellyfish outbreaks in the Mediterranean basin are sending warning signals of a potential phase shift from a fish to a gelatinous sea. The lack of data on those organisms makes the comprehension of the underlying mechanisms difficult. Indeed, those organisms have been considered as trophic dead ends for long, and they are very difficult to collect due to their unpredictable appearance and their fragility.
As jellyfish strandings and blooms are very noticeable and frequently reported events, using social networks, mass media and scientific publications, we will obtain a multilingual and heterogeneous corpus of documents indexed with meta-data. A data classification will aim to select relevant documents from a supervised machine learning approach initiated by experts validations. Valuable support will inform on the date, location and the jellyfish species observed.
From this complementary approach based on text and data mining, we aim to construct the most extensive database on jellyfish in the Mediterranean and Black Seas. This open access database will allow to map space-time patterns of stranded jellyfish and to assess their diversity as well as the contribution of non indigenous species to such events. Finally, we aim at testing whether these events and patterns of exotic species are related with tropicalization and meridionalisation phenomena of the Mediterranean Sea.