Having said that, we would like researchers to be able to learn from our experiences to better prepare for future disasters. With the exception of the open dataset above, all the messages, chat logs, and other material, including the content of this website, are Copyright Mission 4636 and cannot be reproduced without our permission. The initiative was coordinated by an open ‘IRC’ online chat, but because of the sensitive nature of some of the content we consider the chat transcript to also be private. The Mission 4636 messages are private messages sent from individuals to response efforts, so they are anonymized. The Mission 4636 text messages and a copy of the chat-log used to coordinate the initiative are available for researchers. However, enough people believed that the entire message should be omitted in this case, so we omitted all messages with sensitive data. The majority Haitians said that it was ok to remove sensitive information from the messages, but still share them with that information removed. There was only one case where we decided to be more careful than the majority vote. Where the two did not line up, we took the most conservative option of the two, keeping more data private. We also took the best practices in data protection globally, using GDPR as the basis for definitions of personally identifying and sensitive data. We ran a survey of Haitians who took part in 4636 and asked them what they felt to be appropriate data to share and what was sensitive. To decide which data to open source, we used two sources. We ask that you treat the subjects of these messages with dignity and respect. We are grateful that people can analyze the experience that we had so that others can avoid this experience in the future. Udacity ( ) is using the data for a data science course that 1000s of people will take, and will learn how to process and analyze data in disasters. We are happy to share that two programs are now using this data for good:ĪI-4-All ( is using the data to teach under-represented people about AI, especially girls and people of color. For example, the Machine Translation teams at Microsoft and Google have used the data to make sure that English to Haitian Kreyòl translation is accurate for emergency and health related information. The data is in a format that should help with Machine Learning and other Data Science methods. The company was called “CrowdFlower” at that time. The data is published by Figure Eight, which is the same company that we used to translate, map and categorize the messages in 2010. Some of the data has been evaluated as non-sensitive, and we share this openly with the world.

Mirc chat rooms medford oregon