- Location: San Francisco, USA, or Oxford, UK, preferred; remote possible for the right candidate
- Availability: Immediate; 12-month 1.0 FTE contract, with possibility of renewal subject to funding
- Reports to: Director of Research
- Deadline: Please apply before 27 March 2020 for full consideration.
Overview of the role
Meedan is seeking a Research Fellow to join its research team for a 12-month fellowship. The successful fellow will work primarily on the CryptoChat project led by Dr Scott Hale at Meedan and Dr Kiran Garimella at MIT’s Institute for Data, Systems, and Society. The Fellow will also contribute to other research projects at Meedan under guidance from Director of Research, Dr Hale. The Fellow’s research will include analysis of large, real-world social media datasets created and annotated by Meedan partners. This opportunity is available to suitable candidates located anywhere in the world to work remotely with in-person opportunities in Oxford, UK, or San Francisco, USA.
Meedan is a 501(c)(3) non-profit that builds digital tools for global journalism. We are a team of designers, technologists, academics, and journalists who focus on open source investigation of digital media and crowdsourced translation of social media. With commercial, media and university partners, we support research, curriculum development, and new forms of digital storytelling. We have supported major fact-checking initiatives around elections in Mexico (Verificado), France (CrossCheck), the US (Electionland), and India (Checkpoint) as well as general day-to-day fact-checking efforts in the Philippines (Rappler, Vera Files), Africa (AfricaCheck, PesaCheck) and for world changing human rights investigations with partners like Bellingcat, Syrian Archives, Watching Western Sahara, and the Berkeley Human Rights Center.
Our product suite includes Check, an open-source media annotation web application integrated with major social media platforms that allows journalists, human-rights defenders, and civil society organizations to annotate, verify, fact-check, and contextualize web content in a collaborative manner mixing human and machine intelligence.
- Develop and undertake state-of-the-art research in social data science, computational social science, or computer science.
- Lead quantitative analysis of large social-media datasets, including messaging app data already collected for the project from tip lines operated by media partners and webscraping.
- Survey existing literature, tools, and techniques in regard to messaging platforms and online communication.
- Writing, publication, and dissemination of results in articles, blog posts, and project reports aimed at both academic and non-academic audiences.
- General day-to-day administration of the project and coordination with all project partners.
- Research design in collaboration with the lead investigators.
- Maintaining the project’s online presence.
- Support legal partners working to develop and evaluate new data governance approaches.
- Participate actively in the intellectual life of Meedan sharing findings and developing research questions with Meedan’s program, product, design, and software development teams.
- Such other project-related duties as may be required by Dr Scott Hale.
- Hold a relevant Ph.D in computational social science, computer science, communication or another programme with social science theory and computational/data science methods. Those without a Ph.D. but with exceptional academic research experience will also be considered.
- Demonstrated experience in empirical data analysis using a suitable scripting language (e.g., Python)
- Ability to effectively disseminate results to a range of stakeholders
- Ability to work flexibly, prioritizing tasks to meet given deadlines and maintaining a high standard of accuracy and attention to detail
- Interpersonal skills relevant to working with colleagues as part of a global, cross-cultural team at Meedan and communicating across disciplines and industry-academia boundaries
- The ability to write reports and publications in fluent English
Optional, but a plus
- Machine learning, natural language processing, crowdsourcing, image/video processing (e.g., image hashing), or computer vision experience
- Strong experience in Computer-Supported Collaborative Work, Computer Mediated Communication, and other disciplines focused on analysing online communication.
- Demonstrated commitment to Open Science and public or policy engagement
About the CryptoChat project
The CryptoChat project is a 12-month quantitative research project analysing communication dynamics in large datasets from encrypted messaging platforms. It leverages existing datasets in a comparative analysis to understand how encrypted messaging differs from the open web and other social media platforms. The intended contributions of this research are multifold.
- Provide insights into the nature of encrypted messaging platforms through descriptive statistics about their use (e.g., number of users, number of unique messages, number of images versus textual messages, temporal trends).
- Provide exploratory analysis of the data through unsupervised methods such as clustering images based on similarity metrics and topic modeling text to generate hypotheses for further research.
- Investigate instances of dangerous and harmful content within the datasets with a specific focus on hate speech and incitement to violence.
- Provide methodological insights into how to best study mis/disinformation on encrypted platforms, with considerations for both ethics and efficacy. Specifically, we compare datasets created through misinformation tiplines with datasets created by crawling the web for join links to large, pseudo-public groups.
- Compare content in our closed messaging datasets to open web and traditional media data to understand the extent to which content overlaps and the temporal ordering of content appearance. Does content appear first on closed messaging platforms and later on the open web or vice versa?
- Build the technical capacity to generate descriptive statistics and exploratory analysis quickly. We will build interfaces for live statistics with the goal of allowing policymakers, journalists, and the public to understand the content submitted in near real-time. We will further build a robust pipeline for exploratory data analysis (e.g., content clustering) so that reports with these results can be published within one month of any tip line deployment.
- Build and test a legal and ethical framework that provides clear guidelines for data collection and data usage in a global context.
Meedan is a fast-paced, small, software development startup that relies heavily on the unique skills of each team member. Our team members are able to work flexibly and respond to challenges across a range of project areas. We are headquartered in San Francisco’s SOMA district with team members distributed through multiple cities and time zones. We encourage individuals from historically underrepresented communities in technology and journalism to apply, including women, people of color, people with disabilities and those who identify as LGBTQ.
Salary and benefits commensurate with qualifications and experience.
Please send CV and statement of interest to email@example.com