We’ve been working with the Innovation and Research Group at Al Jazeera to explore how technology might fuel innovation in journalism and news media. The Canvas Media in Context Hackathon is the first event to come out of our collaboration.
The theme — media in context — acknowledges that there is much more to telling a story than simply conveying facts. How can new technology provide access to the kind of historical, cultural and environmental information that is also needed to understand the events happening around us?
Diversity from around the world
One of the most remarkable features of this hackathon was the expertise and incredible diversity of the participants. 1400 people applied from all over the world — a stark reminder of how important this subject is to people everywhere. Narrowing that number down was no easy task but in the end 86 people from 37 countries were invited to attend. Of particular importance to us: 40% of them were women (next year we aim to achieve a higher number).
The quality of the participants was matched by the partners who chose to support the event: MIT Media Lab, Frog Design, KnightLab, Technically, TechWadi, and others all sent representatives to act as mentors and judges.
19 Teams, 12 challenges
A number of curated challenges were presented for the participants to work on, each covering different phases of the media cycle (production, distribution, consumption) and various contexts ranging from personal to situational, cultural and editorial. At the end of those 3 days, 19 technology projects were created in response to the 12 challenges that were posed to the participants.
Here are a couple of featured challenges that were worked on at the hackathon:
Mapping an Understanding: Everything happens somewhere and the news is no exception. With billions of place names in the world, understanding the geography of international news is a task made for visualization and mapping. How can we couple advanced geospatial technologies and media content to provide a geographic component to storytelling?
Giving Voice to the Voiceless With Data: One of Al Jazeera’s guiding editorial principles in its news coverage is to “give voice to the voiceless.” Using a media organization’s API or content database, analyze its content from a data science perspective. Where is the media organization meeting and exceeding its editorial goals and mission, and where is it falling short? On what groups, topics, areas, or events should it focus to ensure a balanced perspective in its coverage? How could it distribute content and coverage differently?
As might be expected, the quality of the final projects reflected the expertise and experience of the groups themselves. It was easy to forget that these people had never even talked to each other until they arrived in Doha, let alone work together. Here are a few examples:
Journalist Yana Kunichoff covered recent events in Ferguson. She noticed the while the media coverage was extensive, it didn’t adequately the emotions around the event or its impact on people there. Yana was part of the team that developed Street Stories, a media consumption tool that complies various different media into a single immersive experience. Although it was built entirely from scratch at the hackathon, Street Stories has already been used by the South China Morning Post [Link].
The Lasertag team developed a tool for news producers that better contextualizes archives. The team devised a simple algorithm for relating stories using overlap in noun phrases. For the proof of concept, the Lasertag team used 3,900 stories from the Al Jazeera API. Finding noun phrase overlap required 15.2 million database operations and resulted in more than 1 million relationships. These noun phrases allow for automated tagging and connections of important concepts between different news stories in archives, making them more useful and accessible.
Many people only read one or two articles about any given news story. But different articles can portray the same event in slightly or even very different ways. The reader’s point of view can be influenced by the author’s inherent conscious or subconscious biases, including the connotations of chosen words and the inclusion/exclusion of certain facts. Perspectives was created to discover discover different takes on news events. Perspectives is a Chrome extension that allows news readers to seamlessly discover alternate accounts of any article they read online. Through the extension, the reader is shown facts taken from other articles. These facts are specifically chosen to contain words, phrases, and information that are different from the article that was originally being read.
Each project was interesting and innovative in its own way. Check out the full list here. You can also have a say in the final prize that is awarded by voting for the one that excites you the most!
The Canvas Innovation community has been set up as a long-term effort that aims to involve a group of individuals at the global level with an interest in contributing to the mission of pushing the boundaries of technology and media. Even though the effort has been initiated by Al Jazeera Network, this is a platform for a wide community to engage and benefit from, including other media organizations, academic institutions and technology partners.
A number of the projects that were created at the hackathon are currently continuing to be developed. Opportunities are being set up for participants to receive support and coordination in pushing their projects forward, either as teams or in collaboration with other organizations. The hackathon community is keeping closely in touch helping each other and starting to involve a wider number of members and mentors to explore new opportunities in the intersection of media and technology in tangible ways.
For additional information, Andrew Thompson at Azavea has published a great overview post on the initiative.