Very soon after the recent magnitude-7.8 earthquake in Nepal, we were contacted by multiple groups involved in directly responding with aid and rescue teams, asking if we could assist in the efforts getting underway to crowdsource the mapping of the region. One of those groups was Rescue Global, an independent reconnaissance charity that works across multiple areas of disaster risk reduction and response. Rescue Global also works with our collaborators in machine learning here at Oxford, combining human and computer inputs for disaster response in a project called Orchid. And they asked us to help them pinpoint the areas with the most urgent unfulfilled need for aid.
And so we sprang into action. The satellite company Planet Labs generously shared all its available data on Nepal with us. The resolution of Planet Labs’ imagery – about 5 metres per pixel – is perfect for rapid examination of large ground areas while showing enough detail to easily spot the signs of cities, farms and other settlements. After discussions with Rescue Global we decided to focus on the area surrounding Kathmandu, with a bias westward toward the quake epicentre, as much of this area is heavily populated but we knew many other, complementary efforts were focusing on the capital itself. We sliced about 13,000 km2 of land imagery into classifiable tiles, and created a new project using brand new Zooniverse software (coming very very soon!) that allows rapid project creation.
We also realised that if we combined our work with the results of some of the aforementioned complementary efforts, we needn’t wait for the clouds to part so that we could get post-quake images. For example, the Humanitarian OpenStreetMap Team (HOT) is doing brilliant work providing exquisitely detailed maps for use in the relief efforts. But here’s the thing: Nepal is pretty big (larger than England). And accurate, detailed maps take time. So in the days immediately following the earthquake, our area of focus – which we already knew had been severely affected – hadn’t been fully covered by HOT yet. And by comparing rapid, broad classifications of a relatively large area of focus with the detailed maps of smaller areas provided by HOT efforts, we could still make very confident predictions about where aid would most be needed even with just pre-quake images.
Because our images were in the sweet spot of area coverage and resolution, we were able to classify the entire image set in just a couple of days with the combined effort of only about 25 people, comprising students and staff members from Oxford and Rescue Global staff. For each image, we asked each person about any visible settlements and about how “classifiable” the image was (sometimes there are clouds or artefacts).
After the classifications were collected, the machine learning team applied a Bayesian classifier combination code that we first used in the Zooniverse on the Galaxy Zoo: Supernova project. After comparing these results with the latest maps from the HOT team, we saw two towns that were outside the areas currently covered by other crisis mapping, but that our classifiers had marked as high priority.
We passed this on to Rescue Global, who have added it to the other information they have about where aid is urgently needed in Nepal. The relief efforts are now in a phase of recovery, cleanup, and ensuring the survivors have the basic necessities they need to carry on, like clean water and food. Now they are coping with the damage from the second earthquake too.
Those on the ground are still busy providing day-to-day aid, so it’s early days yet to properly characterise what impact we may have had, but the initial feedback has been very good. We will be analysing this project in the days and weeks to come to understand how we can respond even more rapidly and accurately next time. That likely includes much larger-scale projects where we will be calling on our volunteers to help with classification efforts. We believe the Zooniverse, Planet Labs, and partners like Rescue Global and Orchid (and QCRI, our partner on other in-the-works humanitarian projects) can make a unique and complementary contribution to the humanitarian and crisis relief sphere. We will keep you posted on the results of our Nepal efforts and those of other, future crises.
PS: This activity was carried out under a programme of, and funded by, the European Space Agency; we would also like to acknowledge our funders for the current Zooniverse platform as a whole, principally our Google Global Impact award and the Alfred P. Sloan Foundation. And, to our team of developers who worked so hard to make this happen: you rock.
Header image adapted from OpenStreetMap, © OpenStreetMap contributors.
Today’s blog post comes from Fran Wilson, a second grade teacher at Madeira Elementary School. Fran strives to promote an interest in science in her classroom and help students discover that not all scientists work in labs wearing white lab coats and safety goggles. She seeks meaningful opportunities for her students to participate in citizen scientist work to be responsible citizens, inspire future careers in science, and to connect science concepts to the real world.
This fall I decided to implement Zooniverse’s Floating Forests in my second grade classroom. As soon as I read the description of the project, I knew it was perfect for addressing both my state science and social studies standards dealing with interactions within habitats – living things impact the environment in which they live and the environment can also impact living things. The best part was that my students would be able to engage in meaningful work to acquire these concepts. I like to incorporate project-based learning whenever I can to allow my students to assume ownership of their learning and Floating Forests was no exception. My first challenge was determining how to introduce my students to the project when I didn’t even live anywhere close to an ocean.
Introducing the Problem by Integrating Curriculum
Sea otters are very cute! I decided to use second graders’ love for animals as the entry to the study of kelp. I believe that children learn efficiently when curriculum is integrated across content areas so I made a plan. I selected the book Sea Otters by Suzi Eszterhas. The text tells how a mother sea otter cares for her growing pup. The book’s full page color photos with just the right amount of text on each page made this an ideal book. I chose to begin with a language arts lesson. I projected the book onto my Smartboard and modeled a lesson on determining the main idea and details using a page of the text. Sea Otters does not contain subtitles so I told my students that determining the main idea for a page of text was like creating the subtitle to accompany a portion of text. My students eagerly participated in guided practice of this skill while oohing and aahing at the photos of the sea otters and becoming increasingly more intrigued with the information presented in the book.
By the time we finished reading the book, the children had seen the word kelp in the text and noticed the sea otters lounging on top of the ocean in a bed of kelp. It was the perfect time for me to pose some questions: What exactly is kelp? and Why is it important to the sea otter? My naturally curious students shared their thoughts and the interest in the sea otter and kelp escalated.
How can we learn about the sea otter and kelp? That was the next question I posed after the groundwork was laid for engaging my students in collaborative research. Of course Sophia suggested that we find some more books on sea otters and Jon Miguel added that we should even find some on kelp. Tommaso proposed that we do an internet search to locate information on kelp. This planning step empowered my children with making the decisions about how to learn as well as reinforcing the steps a scientist might undergo while researching. Foreseeing my students’ plan, I had already checked out multiple books on sea otters written at various readability levels along with the few books on kelp that I was able to find.
I selected the book Sea Otters by Laura Marsh to read aloud next to my students. This enabled them to compare the information presented in two books on sea otters. My students listened closely to identify the facts from the text that highlighted the importance of kelp to the sea otters. I started a large chart titled “Sea Otters and Kelp Facts” and modeled how to take notes for our shared research. After reading aloud each of our class notes, the students decided that they had learned some ways in which sea otters depended on kelp but that they still didn’t know much about kelp. At that point we started our internet search.
The Floating Forests webpage provides some great resources, even for use with second graders! Under the education tab of the site, I found a link to a video produced by NOAA to introduce the kelp forest to my students. I was excited that one of my students suggested that they should take notes about kelp in their science journals. (I so love when they take the initiative in their learning!). I discovered several other informative, kid friendly sites with information and videos that we viewed in class and my students continued to take notes. After watching the videos Miki suddenly made the connection and proclaimed, “Hey we eat kelp at my house!” The next day she brought kelp in for everyone in the class to taste.
Websites for Learning about Kelp:
Books for Learning About Sea Otters and Kelp:
Compiling the Research:
What a list of kelp facts my students generated! After reading and researching about kelp on the internet, I compiled all of their facts onto our classroom chart. I sensed my students’ enthusiasm towards learning and researching but this was confirmed when I opened my email the next morning from a parent.
What should we do with all of these facts? That was the next question I posed. Addy had the answer to that and she shared that the facts should be placed into categories. I cut apart all of the kelp facts on the chart and we laid them out in our meeting area. The students quickly sorted the facts into categories. Some of these categories included: What is kelp, Parts of kelp, Kelp forests, Animals that live in kelp, Fish and the kelp forest, Sea otters and kelp, and Scientists and kelp. Next, some children volunteered to work in small groups to write the information into a paragraph with a main idea sentence and details. (Yay! This writing linked back to the initial reading of text for main ideas and key details.). Other children volunteered to illustrate the text with crayons and watercolors. The class research on kelp was almost finished until we started…
Discovering the Ecosystem:
Extending learning across the curriculum is really important to me so while the children were working collaboratively to research kelp through viewing websites and the few books I found, I was meeting with guided reading groups to read and discuss books on ocean life. The children began to think about the ocean as a habitat for many animals and the kelp forest as a very important habitat! After sharing the book What if There Were No Sea Otters?: A Book About the Ocean Ecosystem by Suzanne Slade with a small group of children, Ben announced, “I get it! It’s all connected like a big puzzle!” Kalley latched onto the term “keystone species” highlighted for the sea otters in the text and Sonia explained the relationship between the sea otter and the kelp in the ocean habitat.
All these relationships between the many sea animals in the kelp habitat had the children talking. We needed to solidify their thoughts in a way that we could see them. That’s when the children created a giant model of a kelp habitat. The kelp stalks grew quickly on the large blue poster paper while sea otters were being drawn in a corner of the room, prickly purple and red sea urchins were crafted, fish with fins formed, and kelp labels were created. Of course a new page was added to the children’s book on kelp. Now it was time to publish!
Finding Kelp on Floating Forests
It was finally the right time! My students knew about kelp and understood what an important habitat it was for many sea creatures. Now was the time for sharing Zooniverse’s Floating Forests project with my class. Do you think you’d like to help some scientists with a special project on kelp? I asked. My students were SO excited to become involved. They were even more excited when they realized that they would be looking for kelp on real satellite photos taken from space!
First, I prepared for the children’s “official training.” I connected my computer to my Smartboard and the children viewed the brief tutorial on the Floating Forests website. They quickly learned how to classify the satellite photos and circle the kelp. We circled hundreds of photos together and each time they spotted kelp they became very excited.
Circling kelp on the Floating Forests website continues to be a favorite classroom activity. My students enjoy working in teams of two or three on an iPad taking turns to mark the satellite photos. They often keep a tally of how many times they identified kelp on a photo. I love the discussion it prompts among teams of children circling photos. Through their work, they’ve learned that kelp is found near coastlines. They’re intrigued with the places in the world that kelp might be found.
Participating as citizen scientists with the Floating Forest Project has enabled my students to engage in meaningful work. They feel responsible contributing to important scientific research. My students know that some of the kelp forests are disappearing and they are genuinely concerned. This work has made them more interested in their world and has instilled a need to work collaboratively to care for our earth. My students’ interest in science has been fostered and perhaps some of them will even be inspired to become scientists. I feel like my students have gained so much from this learning opportunity but perhaps it’s what they think that counts most.
Student Responses to Floating Forests
We are often asked who our community are by project scientists, sociologists, and by the community itself. A recent Oxford study tried to find out, and working with them we conducted a survey of volunteers. The results were interesting and when combined with various statistics that we have at Zooniverse (web logs, analytics, etc) we can start to see a pretty good picture of who volunteers at the Zooniverse.
Much of what follows comes from a survey was conducted last Summer as part of Masters student Victoria Homsy’s thesis, though the results are broadly consistent with other surveys we have performed. We asked a small subset of the Zooniverse community to answer an online questionnaire. We contacted about 3000 people regarding the survey and around 300 responded. They were not a random sample of users, rather they were people who had logged-in to the Zooniverse at least once in the three months before we emailed them.
The remaining aspects of this post involve data gathered by our own system (classification counts, log-in rates, etc) and data from our use of Google Analytics.
So with that preamble done: let’s see who you are…
This visualisation is of Talk data from last Summer. It doesn’t cover every project (e.g. Planet Hunters is missing) but it gives you a good flavour for how our community is structured. Each node (circle) is one volunteer, sized proportionally according to how many posts they have made overall. You can see one power-mod who has commented more than 16,000 times on Talk near the centre. Volunteers are connected to others by talking in the same threads (a proxy for having conversations). They have been automatically coloured by network analysis, to reflect sub-networks within the Zooniverse as a whole. The result is that we see the different projects’ Talk sites.
There are users that rise largely out of those sub-communities and talk across many sites, but mostly people stick to one group. You can also see how relatively few power users help glue the whole together, and how there are individuals talking to large numbers of others, who in turn may not participate much otherwise – these are likely examples of experienced users answering questions from others.
One thing we can’t tell from our own metrics is a person’s gender, but we did ask in the survey. The Zooniverse community seems to be in a 60/40 split, which in some ways is not as bad as I would have thought. However, we can do better, and this provides a metric to measure ourselves against in the future.
It is also interesting to note that there is very little skew in the ages of our volunteers. There is a slight tilt away from older people, but overall the community appears to be made up of people of all ages. This reflects the experience of chatting to people on Talk.
We know that the Zooniverse is English-language dominated, and specifically UK/US dominated. This is always where we have found the best press coverage, and where we have the most links ourselves. The breakdown between US/UK/the rest is basically a three-way split. This split is seen not just in this survey but also generally in our analytics overall.
Only 2% of the users responding to our survey only came from the developing world. As you can see in a recent blog post, we do get visitors from all over the world. It may be that the survey has the effect of filtering out these people (it was conducted via an online form), or maybe that there is language barrier.
We also asked people about their employment status. We find a about half of our community is employed (either full- or part-time). Looking at the age distribution, we might expect up a fifth or sixth of people to be retired (15% is fairly close). This leaves us with about 10% unemployed, nearly twice the UK or US unemployment rate, and about 4% unable to work due to disability (about the UK averaged, by comparison). This is interesting, especially in relation to the next question, on motivation for participating.
We also asked them to tell us what they do and the result is the above word cloud (thanks, Wordle!) which shows a wonderful array of occupations including professor, admin, guard, and dogsbody. You should note a high instance of technical jobs on this list, possibly indicating that people need to have, or be near, a computer to work on Zooniverse projects in their daily life.
When asked why they take part in Zooniverse projects we find that the most-common response (91%) is a desire to contribute to progress. How very noble. Closely following that (84%) are the many people who are interested in the subject matter. It falls of rapidly then to ‘entertainment’, ‘distraction’ and ‘other’. We are forever telling people that the community is motivated mainly by science and contribution, and for whatever reason they usually don’t believe us. It’s nice to see this result reproducing an important part of the Raddick et. al. 2009 study, which first demonstrated it.
It is roughly what I would have expected to see that people tend to classify mostly in their spare time, and that most don’t have dedicated ‘Zooniverse’ time every day. It’s more interesting to see why, if they tend to stop and start, i.e. if they answered in the purple category above. Here is a word cloud showing the reason people stop participating in Zooniverse. TL;DR they have the rest of their life to get on with.
We’ll obviously have to fix this by making Zooniverse their whole life!
This is my final blog post as a part of the Zooniverse team. It has been by pleasure to work at the Zooniverse for the last five years. Much of that time has been spent trying to motivate and engage the amazing community of volunteers who come to click, chat, and work on all our projects. You’re an incredible bunch, motivated by science and a desire to be part of something important and worthwhile online. I think you’re awesome. In the last five years I have seen the Zooniverse grow into a community of more than one million online volunteers, willing to tackle big questions, and trying and understand the world around us.
Thank you for your enthusiasm and your time. I’ll see you online…
Hashtags are an important element of how the current generation of Zooniverse’s Talk discussion system* helps to power citizen science. By adding hashtags to the short comments left directly on classification objects, users can help each other (and the science teams) find certain types of objects—for instance, a #leopard on Snapshot Serengeti, #frost on Planet Four, or a #curved-band on Cyclone Center. (As on Twitter, hashtags on Talk are generated using the # symbol.)
One of the ways in which zooites can take advantage of hashtags is by using Talk’s tag group feature. A tag group (also called a “keyword collection”) is a collection that automatically populates with all of the objects that have been given a specific hashtag by a volunteer.
For instance, here is a Galaxy Zoo tag group that populates with all Galaxy Zoo objects that have been tagged #starforming. It will continue to automatically add new images that are given the #starforming tag as well.
There are two ways to tell that this is a tag-group collection, not a manually curated one. The first is that the fourth letter in the last part of the URL (CGZL000056) is an L, for “live” collection. (The other type will have an S as the fourth letter, for “static” collection.) The second is that under “description,” the conditions for the tag group will be displayed: what tags it includes and excludes.
Users can create a tag group in either of two ways: 1. Click the “create a tag group” button that will appear underneath the “tags” on the right side of any object page that has at least one hashtag (and then edit the conditions to their liking), or 2. Add “/#/collections/new/keywords/” to the end of the Talk URL; for instance, talk.planktonportal.org/#/collections/new/keywords/
At this point, there is no way to create a collection that includes, say, on Operation War Diary, #casualty or #sniper—only objects that have #casualty and #sniper. You can, however, exclude certain tags: e.g., all #casualty objects not also tagged #sniper, or #casualty and #sniper but no #horses.
Also, please note that, like all collections, these tag groups are currently capped at 500 total visible images.
It is likely that the next generation of Talk (currently being built) will feature a more refined method of curating collections from hashtags, as well as a more effective search functionality. For now, however, zooites should keep the tag group feature in mind… especially as it will be a critical feature of an upcoming project!
* As of January 2015, the Zooniverse projects using the most recent generation of Talk are: Galaxy Zoo, Planet Hunters, Operation War Diary, Milky Way Project, Snapshot Serengeti, Planet Four, Galaxy Zoo Radio, Asteroid Zoo, Disk Detective, Sunspotter, Cyclone Center, Plankton Portal, Notes from Nature, Condor Watch, Floating Forests, Penguin Watch, Worm Watch Lab, Higgs Hunters, and Chicago Wildlife Watch.
The Zooniverse is the subject of a new artwork co-commissioned by the Open Data Institute (ODI) and The Space (a website for artists and audiences around the world to create and explore digital art). We Need Us is a ‘living’ dynamic artwork, powered by your activity on the Zooniverse, driven by the thriving mass of participation across various Zooniverse sites. You can learn more about it at www.thespace.org/weneedus
We Need Us has been created by artist Julie Freeman. She takes anonymised information from your clicks, counting the number of volunteers active on various Zooniverse projects, and classifications that you all create, every minute. She stores this in a new database as sets of values, while also recording the frequency of activity over an hour, a day, and a month. These sets of values create rhythms that are translated into moving shapes, and play different sounds.
The result is a set of living artworks – one for each of 10 Zooniverse projects – and more are on the way! The live data ensures constant change to the visual and sonic composition. The sounds are processed and manipulated just like the data.
While many researchers have tried to analyse and understand the Zooniverse, We Need Us will be the first time someone has tackled the idea from the perspective of art. The Zooniverse community is an engine of discovery and a force unlike any other. We Need Us highlights its rhythms and patterns, showing how diverse and vibrant Zooniverse citizen scientists really are.
You can run the artwork in your web browser by visiting http://www.weneedus.org
Yesterday we launched our latest project: Penguin Watch. It is already proving to be one of the most popular projects we run, with over one hundred thousand classifications in the first day! The data come from 50 cameras focussed on the nesting areas of penguin colonies around the Southern Ocean. Volunteers are asked to tag adult penguins, chicks, and eggs.
Here are my favourite images uncovered by our volunteers so far: (click on an image to see what people are saying about it on Penguin Watch Talk)
See what amazing pictures you can find right now at www.penguinwatch.org