Automated Geotagging

An ongoing effort to explore approaches for identifying the geographic locations of news stories
Project Type
Collaboration, Automation, Location Recognition, Data Analysis
Start Date
August 2020
Bloom Labs
Stephen Jefferson,

Since 2015, Bloom Labs has focused on a wide range of experiments with geotagged news stories where we have encountered questions about automation regularly. These have been in conversations with newsrooms who hesitate to assign their editorial team with another manual task and prospective business partners looking for consistently available geotagged stories.

We are dedicated to supporting the work being done today to improve automated geotagging to make the technology and its capabilities more accessible to newsrooms. Over the years, we've kept in touch with professors who have spent decades trying to achieve better accuracy and scale for the approach, and continuously give support to software engineers pushing ahead on particular language processing challenges. Whether automation can alleviate geotagging adoption challenges and how it's implemented is still uncertain. This page serves as a repository of our ongoing, collaborative research for these uncertainties. We plan to make our findings publicly accessible in the near future so you can follow along and potentially find ways you'd like to work with us.

Research Goals

The goals of this ongoing research focus on gaining insight about the use, habits, limitations, and capabilities of geotagging local news stories, both manually and automatically, including:

  • To learn current challenges of manual geotagging techniques conducted by journalists;
  • To learn specific limitations on the capabilities of current automated geotagging software;
  • To learn how automated techniques can be designed to be useful for journalists for identifying locations in news stories, considering the capabilities identified.


Updates about the technology and research that makes Bloom Labs an innovative tool for journalism.

Read and Subscribe