Will Fortin, Ph. D.
Lead Data Scientist, Hum
My research aim is to extract as much quantifiable information as possible from data anywhere within reach of human interaction. My current work focus is using computational methods (AI/ML/DL) to understand human language.
Previously, as an academic at Columbia / Lamont, I worked with large seismic datasets to understand the subsurface from the ocean interior, to crustal geology, to our deepest drilling projects.
Check out my research pages for more information.
- 2022 June. Society for Scholarly Publishing meeting in Chicago
- 2022 Feb. Rollout engagement project
- 2022 Jan. Official full-time start with Hum
- 2022 Jan. Won citizen's series snowshoe race!
- 2021 August. Natural Language Processing taxonomy project rollout
- 2021 March. Ph.D. student Joanna's successful defense!!
- 2021 Feb. Partial time project with Hum
- 2020 July. Press Article on Emperor Ridge project published
-2020 June. BLM science strike
-2020 Feb. Invited Speaker, OCP seminar at Lamont-Doherty Earth Observatory
-2019 Dec. Presenter x2, AGU fall meeting
-2019 Nov. Invited Speaker, ConocoPhillips department seminar at Texas A&M university
Data Science - Hum
As Lead Data Scientist at Hum, I focus on understanding scholarly publishing and association organization content.
Nobody wants junk emails and ads. At Hum, we capitalize on underutilized existing data and metadata to deliver value to organizations who value the interest and engagement of their users. As a customer data platform we specialize in helping organizations and marketers realize the full potential of their existing content and how their users use it.
More specifically, I have been focused on building artificial intelligence models in natural language processing to build taxonomies, measure user engagement, predict user behavior, and determine likely impact for recently published material.
Natural Hazards Risk Analyses
I am working with collaborators to investigate fluid overpressure off the U.S. east coast and its role in large submarine landslides. Such slides occur frequently, covering ~20% of the seafloor, and are capable of producing tsunamis along the heavily populated coast. Using advanced computational techniques like prestack waveform inversion and machine learning, we gain quantitative insight into seabed conditions and work toward a comprehensive understanding of risk associated with regional submarine landslides.
By applying novel processing techniques to legacy data, I mapped CO2 storage potential of Mesozoic rift basins offshore the US east coast. Detailed prestack waveform inversion yield insight into basin compositions, and thus the character of the rock infill and volumetric storage potential.
My research goal is to explore energetics in the oceanic interior by mapping diapycnal diffusivity of meso- and sub-mesoscale ocean structures in conjunction with temperature and salinity profiles inverted from seismic data. I am particularly interested in regions with high internal wave activity, seamounts, rough bathymetry, and large eddies. Field sites include the Caribbean, South China Sea, Adriatic, and the North Pacific.