Knowledge mining in massive collections of geoscience documents with digital technologies: our journey in the {TELLUS} community

@inproceedings{bouziat_knowledge_RM2023,
 abstract = {Since late 2022, the extreme mediatization of ChatGPT and other comparable tools has highlighted the tremendous progresses recently made by large language models (LLM), and broadly by natural language processing (NLP) technologies. Still, data science and Machine Learning applications to text and documents are not something new, and various methods have been explored for several decades before the current boom. Besides, universal models based on gigantic generalist datasets are probably not the optimal answer to all knowledge-mining challenges faced routinely by geoscientists in the subsurface industries. Quite often, a combination of less sophisticated and more specialized techniques fueled with subject matter expertise can provide solutions satisfying enough to unlock enormous efficiency gains in everyday operations. In this talk, we illustrate this observation with a sample of projects carried out in the framework of the TELLUS consortium, led by IFP Energies Nouvelles (IFPEN) and dedicated to the digital transformation of geoscience activities.},
 author = {Bouziat, Antoine and Nguyen, Minh-Tuan and Divies, Renaud and Khvoenkova, Nina and Siccardi, Olivier and Dehghan, Kaveh and Rumbach, Gabrielle},
 booktitle = {2023 {RING} meeting},
 language = {en},
 pages = {10},
 publisher = {ASGA},
 title = {Knowledge mining in massive collections of geoscience documents with digital technologies: our journey in the {TELLUS} community},
 year = {2023}
}