Help Shape the Future of Dataiku Join our User Research Program

Researcher Frank Romo (University of Michigan) - Mapping Police Fatal Encounters to Inform Future Policy

Team members:
Frank Romo, Master of Urban Planning Researcher
Harley Etienne, Professor

Country:
United States

Organization:
University of Michigan - Independent Researchers

Description:
The team consists of Professor Harley Etienne and myself, Frank Romo. Our research on this topic has been going on for over five years and has been presented in various formats presentations, maps, videos, and community workshops. Harley and I are independent researchers working to support community safety and change through our academic research and community’s action.

Awards Categories:

  • Organizational Transformation
  • Excellence in Research


Challenge:


The research project focuses on mapping Police Fatal Encounters.

Our team cleaned, mapped, and analyzed thousands of records from various data sets to better understand the spatial distribution of fatal encounters in the United States between the years 2015 and 2020.

The main challenge we faced was comparing records across multiple data sets and building a comprehensive dataset from various partial sources. Using Dataiku, our team was able to combine multiple datasets and create the first ever comprehensive dataset on this topic.


Solution:


Dataiku helped by working with us to establish a workflow that allowed us to not only create a comprehensive dataset but also perform spatial analysis, as well as run regressions and static tests. We had a great support from Lorena De La Parra (AI Strategist) and other team members to develop our strategy, testing methods, and to deliver a final dataset that we could use for future maps and visualizations.

461308BE-CB17-4489-8995-6A7F4A1C579E.png


Impact:


The results of this collaboration allowed Professor Harley Etienne and myself to submit abstracts to multiple academic journals. Currently, our research on Race and Policing in America is being examined and reviewed by multiple academic journals for potential publication.

In addition, the dataset we created during this process was mapped and used in various community presentations. In fact, our research, maps and analysis have been highlighted on recent podcasts at MIT Community Innovators Lab and within the geospatial industry with ESRI.

We will continue to build on this great momentum and continue to use the tools that Dataiku provides to help clean and refine our data so that it can be presented to the public and help inform future policy discussions.

Attachments
Version history
Last update:
‎07-19-2021 12:41 PM
Updated by:
Contributors