The Local News Lab is a team of engineers, designers, data scientists, and journalists, working to build AI-powered, open-source products to help support local newsrooms and their businesses. We are located within the Brown Institute for Media Innovation at the Columbia Journalism School.
The Local News Lab is committed to developing and building solutions that center local newsrooms. Our team has over two decades of combined experience working in media, helping attune us to the ethical sensitivities in journalism. An Advisory Board of roughly a dozen representatives from both large and small newsrooms ensures we have input from a range of perspectives, as we move from experimentation to prototyping and production. Finally, we have a Partnerships Lead who is dedicated to the task of nurturing our newsroom relationships, so we can engage in ongoing dialogue and avoid an extractive, parachuting approach to our work.
As the first of a number of anticipated projects, we are developing an open-source smart paywall that deploys Machine Learning (ML) to go beyond one-size-fits-all solutions to audience engagement. We explicitly extend the idea of a paywall to include a variety of actions that a publisher can take to engage their readers and encourage subscription or donation. Currently, our team is exploring three avenues of data science work — paywalling likely subscribers, selecting premium content, and personalizing reader recommendations. We are looking for a data scientist to begin work on novel approaches, in addition to deepening this existing research. The end goal of this work is to deploy equitable and explainable models that help local newsrooms sustain their businesses.
Our team strives not only to produce equitable technology, but to work in as equitable a manner as possible. We make a concerted effort to distribute the administrative and emotional labor that often falls too heavily on those who are not cis men. We try our best to be inclusive of all genders and to make space for our POC and femme teammates during meetings and in decision-making processes. We don’t always get it right, but we hope to create an environment where everyone is comfortable sharing their feedback when we don’t.
Develop pipelines that extract insights from news text and clickstream data in order to:
- Enable newsroom strategists to make more informed decisions
- Facilitate automatic decision-making, including, but not limited to, paywalling likely subscribers, selecting premium content, or personalizing reader recommendations
Primarily responsible for the statistical components of the system and will serve as an authority on the topic among the development team members. Under the supervision of the Project Lead, the Data Scientist, together with the full development team, will hone the project’s formal specifications, and set technical milestones.
This is an externally funded position, contingent on performance and the continued availability of funding.
The ideal candidate has experience in one or more of the following domains:
- NLP: entity extraction, resolution, and linking. Linguistic pattern analysis, such as LIWC, readability, sentiment. Topic modeling, such as LDA, text clustering
- Network analysis: community detection, influence analysis
- Experiment design: crowdsourcing, conducting and evaluating A/B tests
- Reader analytics: experience optimizing metrics like impressions, CPM, conversions
- Data science ethics: statistical fairness, model explainability, auditing for bias
Experience with software for newsrooms and the news industry is desirable but not required.
The Magic Grant Program. The David and Helen Gurley Brown Institute for Media Innovation was founded in 2012 and is a joint effort between Stanford’s School of Engineering and Columbia Journalism School. Each year, the Brown Institute awards close to $1M in grants and fellowships to foster new tools and modes of expression, and to create stories that escape the bounds of page and screen. We are committed to radical experimentation with the potential to define new priorities and practices for both engineering and journalism.
The “Magic Grant” program provides year-long funding awards of up to $150,000 ($300,000 for teams with members of both the Columbia and Stanford communities). In addition to funding, grantees have access to a distinguished advisory and mentoring group, and an extensive and inspiring alumni network.
Successful Magic Grant projects have taken various forms — from novel works of journalism, to new software platforms, and even innovations in hardware. The common link among all our grants is that they develop new ways to find and tell stories. They can be platforms that extend our creativity, or powerful new kinds of journalism.
Since its founding, the Brown Institute has funded over 80 projects through its Magic Grant program and a complete list can be found on our website at brwn.co/magic-grants.
Evaluation and Requirements. Magic Grant proposals are evaluated on: 1) the originality of the project described; 2) its potential for impact; 3) the strength of the team; and 4) the timeline outlined to complete the work. Each of these areas should be clearly addressed in the proposal.
- Proposal submission deadline: May 1, 2021 (11:59pm GMT-10)
- Announcement of finalists: May 14, 2021
- Presentation/Q&A by Columbia finalists (virtual): May 24, 2021
- Presentation/Q&A by Stanford finalists (virtual): May 25, 2021
- Announcement of winners: June 1, 2021
- Projects start: September 2021 (July/August by special arrangement)
Additional Information. The full Magic Grant Call for Proposals can be found at brown.columbia.edu/propose or brown.stanford.edu/propose. Questions about the process and eligibility requirements can be found under the Propose menu item of our site.