Database journalism

Computer-assisted journalism

it's not about how many developers you have, but what you do with them. – Adrian Holovaty

Data journalism outcomes can range from visualization to long form articles. It is about the process how to turn numbers in to a story, whether the story is comprised of words or graphics is irrelevant. By knowing the structure of your team and balancing resources, you can start to think how best to organize your group.

• https://www.youtube.com/watch?v=DxvVZe2fnvI
• Government spending by department, 2011-12: get the data, The Guardian

From Doing Journalism with Data: Frist Steps, Skills and Tolls: Lession: Login required

Key-roles, you will probably need to do at least 2 of these:

1. Research (4:58)
   • Most important role
   • On basis of a journalistic hunch, or by digging around in data
   • Time consuming
2. Writing (5:58)
   • Numbers without context are just numbers, words give context
   • Example: Yearly guide to public spending by each government department (Links to an external site.) (The Guardian)
3. Development and coding
   • Coders (programmers), can help with the research and visualization
   • Coordinate who can write and code
   • Example: Connected China (Links to an external site.) (Reuters)
4. Designing and visualizing
   • Designers can make visualizations happen
   • Example: 99% vs 1% (Links to an external site.) (The Guardian)

1. Lone rangers
   • one person does everything
   • possible by all the tools of the DDJ ecosystem:
     • OpenRefine,
     • Datawrapper,
     • Tableau,
     • Google Fusion Tables,
     • CartoDB
2. Two-person teams
   • Gay rights in the US, state by state, the Guardian
3. Small scale teams
   • Is able to produce innovative projects quickly
   • Part of the newsdesk
     • Example: 'Flooding and Flood Zones' map Hurricane Sandy (WNYC)
4. Larger teams
   • Has a deliberate strategy to create a new kind of online journalism
   • They help with finding ways to tell the story better
   • Part of the newsdesk
     • Example: 2012 Olympic Experience, New York Times

Readings in the Guardian

• Anyone can do it. Data journalism is the new punk
• Data visualisation DIY: our top tools

• Data-journalists are the new punks: Simon Rogers at TEDxPantheonSorbonne

• Data are or data is?

The Inverted Pyramid Structure at Purdue University

For data journalism, the 5 W's have never been so important:

1. Who? (1:30)
   • Where did the data come from?
   • The most important W
   • Transparency about the source is critical
   • Do not blindly trust accuracy of supplied data
2. What? (3:05)
   • What are you trying to say?
   • What points are you trying to get across?
   • Your stories are aimed at the general public
   • Example: Art Market for Dummies (Askmedia.fr)
   • Your job is to bridge the gap between the data and the user
3. When? (4:28)
   • How old is your data?
   • Near real-time data
     • Chicago homicides database
     • Example story based on this database “Englewood sees 3 homicides in 4 days, Redeye Chicago
   • Social media data
     • Twitter account Boston Marathon Bomber
       • read Dzhokhar and Tamerlan Tsarnaev
     • Example story ”We know when Dzhokhar Tsarnaev sleeps“, Quartz
   • Sensor journalism
     • Cicada Tracker, WNYC
       • Full article, ”People Power Prevails!“ (Source)

• Crowdsourced reporting
  • UK riots: every verified incident - interactive map, The Guardian
  • Data journalism reading the riots: what we know. And what we don't, The Guardian
  • Occupy movement: Help us map which protests are still active, The Guardian
  • Superstorm Sandy: every verified event mapped and detailed, The Guardian
  • Broadband Britain mapped: your web speeds, The Guardian
  • Each of these projects was based on a simple journalistic tool with a few steps:
    1. A Google form embedded on a page to collate reader responses;
    2. That data edited by journalists to compensate for repeated or offensive data;
    3. Data imported into Google Fusion Tables;
    4. The data feeding a map which updates every time it’s reloaded.

1. Where? (07:53)

• A key part of data journalism is the ability to ‘mash up’ different datasets to create a new story.
• The gun ownership and gun homicides murder map of the world, The Guardian

1. Why? (8:48)

• Hardest question
• Data journalism is less good at correlating that data to produce a cause and effect analysis.

오픈데이터저널리즘
빅데이터저널리즘
데이터중심(기반)저널리즘

Computer assisted = fact gathering
여기에 더 나아가 인터액티브한 결과물을 도출하는 과정
따라서
데이터저널리즘 =

• 데이터의 수집, 정리, 분석, 시각화, 스토리화 등의 과정을 통해 저널리즘을 실천하는 행위들

데이터 기반한 인포그래픽, 스토리텔링 → 데이터마이닝, 데이터 전문가의 역할이 중요해 짐

필요성

• 정보불평등의 해결책
• 저널리즘의 개관성과 신뢰도 향상
  • http://hot.coroke.net 충격, 경악, 황당 등 자극적이고 무의미하며 선정적인 단어로 구성된 헤드라인 사용
• 깊이 있는 정보 (사실과 사실 밑에 있는 것들 (embedded))

역할

• 저널리즘 객관성 강화
• 데이터의 공공적인 성격을 부각시켜 공개화 추구 (open 데이터) → 정부, 기업의 투명성
  • 국가안보, 국익, 프라이버시, 국민의 알권리 등과 결합하여 충돌
  • 개방에 걸림돌이 되고 있음 (Lehtonen, 2011)

• wikileaks.com
• science communication: from page to screen
• 데이터 수집과 참여에 개인이 참여 → 개인의 정치, 사회적인 활동 강화
• 스토리텔링을 보여주기 위한 기술의 강화 (증강현실 활용 등)
• https://wikileaks.org/
• http://news.coroke.net/