March 1, 2012 MEG Luncheon: DATA MINING

Extracting Map Data from Mining Documents

Geographic Search and Referencing Platforms

The mining industry generates millions of documents a year, not only from geotechnical data but from training, regulatory, personnel, and legal departments. While specialized geotechnical and GIS applications enable the storage, analysis and retrieval of the structured portion of this information in map views, the key piece of information needed to support a business critical decision is more likely to be stored on a shared drive in Word, PowerPoint, Adobe or Excel, in a SharePoint document repository, or on a web page.

Recent field-proven deployments of Natural Language Processing engines and machine metaheuristics have given the mining industry the ability to automatically scan, digitize, and extract geospatial metadata from legacy hard copy documents. This enables rapid searches from a map in a variety of geo-enabled portals and applications. Integration of structured and unstructured data is a business need originally identified by “big data” industry sectors, scaled to smaller operators, and recently applied to the mining sector. Patented and customized Geographic Data Modules are being developed to extract geospatial taxonomies from thousands of documents per second and recognize place names as they are uniquely used in the mining industry.

Author Biography
Jess Kozman is the Director of the Mining and Minerals sector for QBASE, a North America based company specializing in "big data" management and analytics. He is a geophysicist by degree, explorationist by training, IT manager by experience, project management practitioner by certification, and a data and business intelligence specialist by choice. He has spent over 35 years working with geotechnical data from onshore and offshore regions of every continent on our planet, and has consulted on the management of data for projects as varied as oil and gas extraction, geothermal energy, CO2 sequestration, environmental policy impacts of hydraulic fracturing, post earthquake tectonic studies, landslide remediation,  and methane hydrates as an energy source for Mars exploration. He was recently based in Australia where he field validated the transfer of data management best practices from oil and gas operations to the mineral exploration groups of RioTinto and BHP Billiton.