Yemen: Death from above is an ongoing data visualisation which documents the devastation wreaked on Yemen by the more than 19,000 air raids over the course of 4 years.
The project aims to provide a global audience with a contextual understanding of the human cost of war through the use of maps, annotated timelines, and a unique audio data visualisation.
Built entirely following a reproducible data journalism workflow, this story aims to become an indispensable resource for future data-driven projects on Yemen.
What makes this project innovative?
Produced by data journalists from the Middle East, this project contains three distinct features which makes it innovative. 1. Transparency and reproducibility: The underlying data for this project was compiled by the Yemen Data Project and is publicly available. This data set, which has been widely used by news organisations, contains a detailed cross-referenced account of every raid across the country since 2015. In keeping with this project’s transparency and reproducibility, a core editorial feature of this data-driven story is its transparent and reproducible data analysis process and source code (written in R) available here: https://medium.com/@ajlabs/how-and-why-we-made-our-first-audio-data-visualisation-ed7c136327ae This open process sets a new standard particularly in data-driven reporting from the Middle-East and allows the story to be updated systematically as new data becomes available. 2. Audio as the vehicle for the story: In addition to an interactive data visualisation (using D3.js), this project uses audio, with increasing levels of intensity, to represent four years of non-stop raids across the country. By incorporating audio, readers are able to listen to the story unfold. This unique approach to data visualisation combines the strengths of visual and audio content consumption and adds a new dimension to this very emotive story. 3. Content strategy and contextualisation: The entire story was built to ensure that readers, regardless of consumption platform, were able to consume the story with relevant context and understanding. As part of this content strategy, we produced bespoke versions of this story for our website, Facebook, Twitter and Instagram stories. This approach ensures that the story remains relevant and impactful.
What was the impact of your project? How did you measure it?
A large part of how we define this project’s success stems from its content distribution strategy. Three key metrics were identified during the inception of the project to ensure that we were designing a user experience for maximum impact. 1. Platform independent: To maximize our reach and native consumption on all our social media platforms, we produced bespoke versions of this story for our website, Facebook, Twitter and Instagram stories. Success was not defined only by clickbacks to the main interactive but on how readers engaged with the story on the platforms themselves. In the case of Instagram we crafted several questions and follow-ups to the story linking back to the rest of Al Jazeera’s Yemen coverage including a VR experience and several social video explainers. 2. Long tail traffic: The true success of this project lies in its ongoing relevance and long-tail impact. The project, which was published just over 2 weeks ago has maintained consistent web traffic and social media engagement. By following a reproducible data analysis approach this should mean that the story will remain up-to-date and most importantly, relevant for a global audience for a long time to come. 3. Time on page: Another key indicator of success is the project’s time on page and traffic type. Nearly half of all visitors to the story were new users each spending on average 3:40 on the page. On Instagram stories, the conversion from the first slide to the last slide was among the highest numbers we’ve seen on our platform.
Source and methodology
The complete data sourcing, cleaning, processing and analysis methodology is all available on Github. https://medium.com/@ajlabs/how-and-why-we-made-our-first-audio-data-visualisation-ed7c136327ae A summary of these processes is provided below: Data source: Yemen Data Project, verified by our Yemen subject specialist. Data analysis: 1. Import the data and data wrangling packages 2. Filter the data that we’re interested in 3. Reformat columns as dates and integers 4. Find answers to all the questions we asked 5. Consolidate the output data required for the presentation
Each of the technologies used in this story was selected for their ability to produce a mobile-first, modular and reproducible output. These technologies include: React.js - Front-end responsive user interface framework. D3.js - Data visualisation elements. Adobe Audition - Create unique audio tracks. FFMpeg - Terminal based data-driven audio/video generator. R (Tidyverse) - Reproducible data analysis. Google Sheets - Collaborative data analysis.
Produced by: Faisal Edroos, Alia Chughtai, Mohammed Haddad, Talha Dar and Kali Robinson for @AJLabs