The Identification of a Promising Research Topic in Applying Generative Artificial Intelligence in Petroleum Engineering

This study aims to identify a promising research topic related to the use of generative artificial intelligence in the petroleum industry. It involves the collection of publications on generative artificial intelligence in the Scopus abstract database related to engineering and computer science, the systematization of the publications using the GSDMM algorithm, and the search for publications on the OnePetro platform that are close to the research objectives identified by Scopus. The analysis focused on 12 424 Scopus bibliometric records. The texts of the title and abstract fields were used to cluster the records. As a result, 21 clusters were obtained. For each cluster, stacked histograms of the difference in the occurrence of terms for this cluster and the other clusters were constructed using the program Scimago Graphica. A promising research topic could be adversarial attacks that compromise generative models by manipulating input data. This topic is underrepresented in petroleum literature, but has significant research potential because much has been written about it in publications from other subject areas. The findings of this study provide the petroleum industry professionals with the opportunity to broaden their search for publications on generative models and deepen their expertise in this research area.