AP® Computer Science Principles review sheet from Aim for Five (aimforfive.com/csp/units/5/5-4)
Unit 5 · Topic 5.4
5.4 Crowdsourcing
The internet lets huge numbers of people contribute to one project, from a single fact to a few dollars to hours of computer time. This topic covers crowdsourcing, citizen science and how open data helps people find and solve problems together.
Key terms
- crowdsourcing
- citizen science
- collaboration
- public data
Crowdsourcing
Crowdsourcing means getting ideas, information, work or money from a large number of people through the internet, instead of from a small group like one company's employees.
Examples are everywhere. Wikipedia articles are written and corrected by volunteers. Navigation apps collect reports of crashes and traffic jams from drivers. Review sites gather ratings from customers. Crowdfunding sites like Kickstarter let thousands of people each give a small amount to launch a product, a film or a community project.
Crowdfunding is one of the new ways computing connects people with causes and businesses they want to support. A charity can raise money for a disaster within hours, and an inventor can test whether people want a product before building it.
Citizen science
Citizen science is research done fully or partly by many people, most of whom aren't professional scientists, who collect or analyze data using their own phones and computers.
Bird watchers submit sightings to the eBird project, giving researchers a picture of bird populations over whole continents. On the Zooniverse platform, volunteers classify galaxy images, transcribe old handwritten records and count animals in camera-trap photos. In 2011, players of the online puzzle game Foldit built a model of a protein from a monkey virus related to HIV that let scientists finally work out its structure, a problem that had stumped researchers for more than a decade.
Some projects don't need people's judgment at all, just their spare computing power. Volunteers install a program that runs scientific calculations when their computer is idle, which is distributed computing (topic 4.3) carried out by the public.
Why it works, and its limits
Many people can cover far more ground than a few experts. Thousands of volunteers can look at millions of images or report from places no research team could reach.
Open, public data helps every step: identifying a problem (seeing that a disease is spreading), developing solutions (outside programmers building tools with a city's open transit data) and sharing the results with everyone who needs them.
There are limits. Volunteers make mistakes, and some may submit false information on purpose, so projects check contributions, for example by having several people classify the same image and keeping the majority answer. Contributors are also unpaid, and the people who join may not represent everyone, which can tilt the data (topic 5.3).
Worked examples
Try each one yourself first, then open the solution.
- Example 1
Explaining how a crowd helps
A city wants to find every pothole on its roads. Instead of sending inspectors street by street, it launches an app that lets residents take a photo of a pothole, which is tagged with its location. Explain how this is crowdsourcing, one benefit, and one drawback.
Show the solutionHide the solution
- Step 1: Crowdsourcing: the city gets information from a large number of residents through the internet, rather than from its own small team.
- Step 2: Benefit: thousands of residents drive the roads every day, so potholes are reported faster and from more places than inspectors could cover.
- Step 3: Drawback: neighborhoods where fewer people have smartphones or use the app will get fewer reports, so their potholes may be fixed later. That's the digital divide shaping the data.
- Step 4: A good answer also notes a way to manage it, such as checking reports with photos and still sending inspectors to areas with few reports.
Answer: It gathers reports from many residents online; it finds potholes faster and more widely, but areas with less smartphone use may be underreported.
Common mistakes
- Calling any website with many users crowdsourcing. The crowd has to be contributing the information, work or money.
- Thinking citizen scientists must be trained scientists. Most aren't; that's the point.
- Ignoring quality control. Crowdsourced data can contain errors or false entries, so projects need ways to check it.
On the exam
- Expect questions asking which scenario is an example of crowdsourcing or citizen science, or how widespread access to data helps solve a problem.
- Questions may ask for a benefit of crowdsourcing (scale, speed, many viewpoints) or a concern (accuracy, unrepresentative contributors).
Connected topics
Videos
Check yourself
4 questions on 5.4 Crowdsourcing. Pick an answer to see if you got it, and why.
Which of the following is the best example of crowdsourcing?
Thousands of volunteers use a phone app to photograph and log the birds they see. Scientists use the data to track how bird populations change. Which term best describes this project?
An inventor raises $40,000 for a new kind of bike lock by collecting small donations from 2,000 people on a website. What does this example best show?
A city posts its data on potholes, streetlight outages and repair times online for anyone to use. Which is the most likely benefit?
0 of 4 answered