Geographic Literacy in Text Analytics: Developing and Applying OpenSextant

Date: 

Thursday, January 30, 2020, 12:00pm to 1:30pm

Location: 

CGIS South Room S020

Presentation by Marc Ubaldino

Abstract: Developing text analytics in any language for any genre of text is challenging; adding geographic interpretation of text or the mentioned entities therein, further complicates the situation. OpenSextant is one of many open source text analytics efforts that focuses on
geographic analytics of unstructured data and has been applied to domains of biosurveillance,
social media demographics, digesting reporting from the field and world news analysis. While the geotagging and geocoding technology has been developed and proven in limited scenarios there remains significant evaluation challenges and research topics remain, including working with phonetics in geographic names and specializing geotagging for more languages. This talk will discuss how to make textual applications more geographically literate and how to evaluate the quality of such language processing. The key underlying approaches and architecture of OpenSextant components will be introduced at a high level. Outside of this software development discussion we will discuss how such open source technology can be used in educational and training contexts for students of all ages where interaction with geographic resources is beneficial for at many levels.

Speaker Bio: Marc Ubaldino has had a 25 year career as a systems engineer and researcher at The MITRE Corporation (http://www.mitre.org), a non-profit firm that operates federally-funded research centers. His focus areas have been in integrating and applying text analytics and geospatial solutions along with this deeper interest in extracting concrete, usable and credible intelligence from any type of data. These interests are embodied in the work he and others open sourced at http://OpenSextant.org, where he is the author of Xponents and co-authored other capabilities there. These open source libraries and activities actively support his mission area of applied text analytics and geospatial analytics. Marc has a Bachelor’s Degree from Boston University in Biomedical Engineering and a Master’s Degree in Systems Engineering from Johns Hopkins University. He makes his home with his wife, two children and pets in Concord, MA.

Lunch will be served.

geocoding-workflow.png