Dataflood Filter. Sorting and verifing information during a crisis situation: Saving the World.

From TechCampGlobal
Jump to: navigation, search

Project Information

Project Title: Dataflood Filter. Sorting and verifing information during a crisis situation: Saving the World.

Group Number: 14

NGO Project Lead:

Project Participants:

  • Poomjit Sirawongprasert
  • Klaikong Vaidhyakarn
  • Visetsinh Huangvilay
  • Arpaporn Winijkulchai
  • Yuwadee Prinnarith
  • Chatchai Kadum
  • Marie Sicat
  • Nathaniel Manning
  • Katie Dowd


Summary of the Problem

This is a description of the problem situation. What kind of people mission better if this problem is solved?

How do we sort information quickly and accurately during a crisis situation?


Proposed Solution

This should provide an outline of the basic need that would solve the above problem.

Create an online platform that combines computer search power (NLP) to filter information, with the power of human collaboration to sort and make decisions. The platform would combine crowdsourced information with verified databases (government and NGOs) and inputs from sensors. Journalists and crisis responders would act as the curators and editors of the incoming information. The platform will also send early-warning alerts when crowd and sensor inputs report a warning. An SMS warning system will be integrated with large media and crisis management hotlines (1111) to warn the public.

For instance, in the case of floods in Thailand, sensors can be placed in klongs that will report water depth daily. The platform will also pull information from the crowd using Twitter, SMS, email, Facebook, radio, and more (multi-channel platform). In addition the government can offer to upload their databases for help-center locations and official reports. The platform will convert all the data to the same format standards. Data from previous floods can also be uploaded into the platform, allowing for a benchmark when measuring for alert systems.

Once the information is on the platform, volunteers, journalists, and crisis responders can use NLP and Geolocation search to filter the inputs. Then the teams can work together to organize the information, just like working on Google Docs, from both internet and smart phone access. That information can then be consumed online and send out alerts via SMS. The online public portal can show the information on a map, in a feed, on a timeline, in a photogallery, and with analytics.

Ideally the platform would have an intelligent reputation system that could help to evaluate the validity of each incoming piece of information from the crowd. It would be similar to Wikipedia’s system, giving preference to people who have a history of truth or people who are official journalists or crisis responders.


Technology to Solve the Problem

Areas of technology that you believe have the potential to solve this problem? (eg mobile devices, social networks, geospatial technology, etc.). As the proposed solution could be used?

Combine crowdsourcing, sensors, and volunteer efforts into a digital platform. Use mobile and SMS for alerts. Use NLP for searching data, and mapping for presenting the information. Use well-placed sensors for early-warning detection. Integrate all these inputs into the same format. Use Google Translator and organized diaspora groups to translate when needed.

Existing solutions or relevant links

What other information could help technologists working on this idea?

Swiftly.org Ushahidi.com Yahoo Pipes ESRI Pachube.com Twilio.com Google Translator Google Person Finder

Follow-on Actions

Once a potential solution has been developed, what is your plan of action to implement it? Should arise as a series of steps and goals, preferably together with actions assigned to a specific group of people.

  1. Develop technology, deploy sensors, and integrate existing data sets with technology developers from civil society and with buy-in from government crisis management departments.
  2. Work with government, civil society, and crisis responders to set alert benchmarks and monitor inputs.
  3. Practice run-through of crisis response and platform use.
  4. Wait for a crisis. Everyone eats delicious Thai food.
  5. Once alert system warns about an incoming crisis (for instance a flood), crisis responders, journalists, stand-by volunteers, and government go into action.
  • a.Journalists and crisis responders break into ‘ground teams’ and ‘curator teams.’ Ground teams send in information. Curator teams organize information and reach out to the organizations that can respond to help requests.
  • b.Government uploads their data (help centers, help hotlines, etc.)
  • c.People on the ground map their location in real time, and the curators organize it.

TechCamp Bangkok Folder