AI Supports Information Discovery and Analysis in an SPE Research Portal

Ongoing growth in the volume of raw data generated by digitized oil and gas operations has been widely documented (Spath, 2014). What may be less apparent is that the industry is also authoring dramatically more unstructured content—interpretations, learnings, case studies, etc.—on an annual basis. This is not surprising, as anecdotal evidence suggests that most decisions are taken on the basis of unstructured data (Quaadgras & Beath, 2011) (Garland, 2017) (Palkowsky, 2005) (Haines, Shaughnessy, & Briggs, 2006) (Hollingsworth & Schey, 2017). At the Society of Petroleum Engineers, the number of new papers published annually has followed an exponential growth curve that doubles approximately every 10-11 years, beginning as far back as the early 1950's.