A Geodyssey – Geoscience Text Analytics and Enterprise Search Research
A voluntary non-commercial research blog since 2015
GeoGPT Data Security Risks for Geoscientists. The Deep-time Digital Earth (DDE) online proprietary GeoGPT Application operated by Zhejiang Lab has data security risks. In the interests of safeguarding the geological community, geologists should be fully aware so they can make informed decisions.Extracts below of a letter published in the ‘Geoscientist’ today of the Geological Society... Continue Reading →
Geo-LLM: Using Llama 4 to create 3D geological models from geological reports. Innovative work using the Llama 4 API on geological reports, converting into domain specific language for input files to visualise in GemPy. All the code is Open-sourced in GitHub.Congratulations to William Davis and Gain Boonvanich for winning “Best Llama API Usage” at the... Continue Reading →
On this day exactly 10 years ago, I created a free community outreach blog for geoscience data and AI. Thank you to everyone for all the encouragement and support you have given over the years.The blog was initially focused on techniques leveraging content in documents - unstructured data, search (IR) and NLP, tying this to... Continue Reading →
Experimenting on applying machine vision to Dinosaur Footprints with Mark Baggott from the Herefordshire and Worcestershire Earth Heritage Trust. These are some of the photographs of Dinosaur footprints and tracks Mark and his colleagues have taken recently from a variety of UK locations.As a hobby, it's still early days, seeing if we can create a... Continue Reading →
Some interesting research on how employees use Open AI's ChatGPT. Whether its private sector companies, non-profits or governments funding these online proprietary chatbots, and whether they are generic or built to target a domain vertical, we need to raise awareness. Once sensitive data from your company or organisation is uploaded (through an API or user... Continue Reading →
Researchers at Imperial College London, draw on geological data science to reveal regions that can host giant earthquakes, regardless of the history of seismic activity. Including in places where none have occurred in the last 100 years.“We hope that by using the geological data itself as the principal indicator of risk potential, we can tap... Continue Reading →
The subsurface: our hidden asset. How geoscience knowledge and data supports government priorities. Excellent summary from the BGS.“While often ‘out of sight, out of mind’, the subsurface is a natural habitat, a source of critical resources, and a space for the infrastructure that underpins modern economies; in the coming decades the drive for urbanisation, innovation... Continue Reading →
Assessing named entity recognition by using geoscience domain schemas: the case of mineral systems. Excellent paper published in Frontier Earth Sciences by Villacorta Chambi Sandra Paula , Lindsay Mark , Klump Jens , Gessner Klaus , Gray Erin , McFarlane Helen (2025).AbstractNamed Entity Recognition (NER) is crucial for accurately extracting and classifying specialized domain terms... Continue Reading →
Government officials, industry chairs, technical experts and academics recently participated in a round table on how subsurface data sharing can drive sustainable development. I took part representing Robert Gordon University, attended by Members of Parliament, Government, Business and Academia. Report write up in the magazine link below. https://www.geplus.co.uk/features/round-table-the-economic-environmental-and-social-benefits-of-subsurface-data-sharing-07-05-2025/
Seismic hazards 3d Earth Viewer Open Data. Earthquakes on land and in the oceans with magnitude >=6 from the US Geological survey catalogue from 1970 to recent. Uses hypocenter point radius animation based on date and time of the earthquake event.From the University of Washington and partners to help foundation K12 education. Link in the... Continue Reading →
A Geodyssey - Geoscience Text Analytics and Enterprise Search Research 