Invited to give the Distinguished Lecture on Big Data next month for the Science and Technology Exploration and Production (STEPS) program run by Halliburton. The program aims to foster geoscience excellence through the facilitation of thematic research and offers the opportunity for academics to engage with Landmark (Halliburton) and the wider exploration and production community.
The lecture title is Big Data – Small Patterns: Applying Geoscience Sentiment Analysis to Unstructured Text.
Will be sharing recent results and findings of the Geoscience aware sentiment AnalyZER (GAZER) algorithm I developed in Python which has been applied to Geological elements in public domain texts. It is designed to surface interesting associative patterns relating to concepts such as ‘source rock’, ‘reservoir’, ‘trap’ and ‘seal’ that might be unknown to exploration geoscientists as they are buried in volumes of documents too large to ever be read and too subtle to be detected by traditional search engines.
The hypothesis is that if a geoscientist can be surprised by these patterns, and there is legitimate evidence for that ‘surprise’, it is likely to lead to a learning event and potentially a new play/model; changing what people know – or think they know.