PI is an Alert Winner of Tech Challenge for Atrocity Prevention sponsored by Humanity United and USAID. Upon successful development and deployment, PI will automate the collection of relevant humanitarian and human rights information from hard to access areas and verify it using crowd-sourcing and “dumb” mobile phones.
What problems or needs does PI address?
PI addresses many of the shortcomings of current documentation initiatives using crowd-sourcing: lack of relevant and quality information, no or limited assessment of the reliability of the sources and the credibility of the collected information, reliance on the Internet, lack of feedback loops and limited empowerment of those reporting information. To solve these problems, PI makes use of low cost GSM technology (e.g., SMS and voice) to establish a conversation with victims and witnesses of an incident to collect and guarantee relevant and quality information.
Why is PI‘s solution innovative?
From a technological viewpoint, most existing crowd-sourcing platforms operate along the same principles: gather (often incomplete) information from a multiplicity of sources, tag them either manually or automatically and present them in the form of a map, descriptive statistics or as a list of reports. None, however, seeks to engage in a dialogue with the victims and witnesses to collect better quality information and provide useful information in return.
PI proposes to shift from current ad-hoc organization centric collection and verification processes towards permanent user centric information collection and verification processes. With PI, users become active initiators and collaborators. Instead of monologues, we believe in conversations where victims and witnesses become initiators. They are recognized as the real experts of their situation, and are guided in providing relevant and complete information in return for actionable information, while helping relief organizations in their decision making processes and resource allocation.
PI will define the necessary functional requirements in collaboration with humanitarian and human rights actors to automate (i) the reporting and collection of information, (ii) the provision of actionable information to those reporting incidents, and (iii) the attribution of reliability scores to a source as well as credibility scores to a piece of information to help organizations make a better use of their resources.
How does PI collect relevant information?
The reporting and collection process is automated providing for a more efficient use of resources. Typically, victims and witnesses will dial a short code and she or he will be invited to answer a series of basic questions: where and when did the incident happen, who did what, how did it happen and how do you know? An information workflow provides for accurate answers. For example, when a user reports the name of a city incorrectly, the system can propose correction options. Syntactic and semantic analysis automatically assess the relevance of a report and upon reception of quality information the system provides the source with actionable information and send early warning signals to partner organizations. For example, when a victim reports a rape on a specific date and at a certain location, the system can automatically send the victim the contacts of nearby organizations providing retroviral drugs and put them directly in contact if the victim consents.
How does PI verify information?
Another problem that PI tackles is the automatic attribution of reliability scores to a source as well as credibility scores to a piece of information. Say a report is submitted by an unknown source and the information cannot be triangulated with other reports sharing similar characteristics, such as date and place. Both the reliability of the source and the credibility of the information will be unknown. In such instances or when the reliability of a source or the credibility of the information is low, the system automatically sends documentation requests depending on the permissiveness of a particular situation to a trusted network in the vicinity of the reported incident, and/or using geo-fencing technologies, to mobile users entering a specific cell-tower perimeter near the reported incident. Upon reception of new reports a similar process resumes and reliability and credibility scores are automatically updated.
Who can benefit from using PI?
The main beneficiaries can be the victims and witnesses who will have their voices heard and receive actionable information in return for quality information as well as partnering organizations who will become better informed and equipped to decide where to allocate resources and coordinate their efforts.
What are PI‘s ethics?
Humanitarian protection standards and principles will be at the core of our work to guide our answers to security and safety challenges. Hence at PI we abide by the Sphere Humanitarian Charter, the Protection Principles and Minimum Standards in Humanitarian Response and work in line with the Millennium Development Goals, Target 8.F in particular, which is to make available benefits of new technologies, especially information and communications.
This work is licensed under Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International.