Nzuri

Group Members

• Rebecca Odim
• Sarah Rumbley
• Maja Wichrowska
• TA: Juho Kim, juhokim@mit.edu

Problem Statement

With increasingly powerful computers, the problem of visualizing large, complex data sets is central to empowering intelligence analysts to make informed decisions. With respect to the context of UI as applied to human-computer interaction (HCI), very few applications demonstrate the mission-centric criticality and power of good design than intelligence software. There are currently several tools available to analysts, one of the most prominent being i2's Analyst's Notebook. Although Analyst's Notebook is somewhat lagging in its user interface, i2's recent acquisition by IBM means that it will have the most powerful computational back-end of all data analysis tools. For our project, we will explore ways to design an improved GUI as a web application on top of Analyst's Notebook. Building off of existing data visualization frameworks, we will address usability problems in the way users interact with visualizations. To guide our efforts, we will focus on a specific case study of data sets concerning WiFi networks in Africa.

The problem of building communication networks in Africa is complicated by shifting demographic and climate patterns. At the end of 2011, an estimated 65% of the African population had cell phones, but only 13% had Internet access. Despite extreme poverty, smartphone usage is growing at a surprising rate in Africa. In fact, many people depend on smartphones to survive, i.e. using them to quickly locate food markets rather than travel for days in the wrong direction. Many African tribes are nomadic, often crossing country borders in search of water and other resources. However, because cellphone networks differ by country, it is common for one person to carry multiple phones to avoid roaming charges. One inexpensive solution to this problem is VoIP, which requires only an Internet connection and would thus eliminate the need for multiple phones. However, as previously mentioned, WiFi coverage in Africa is sparse. The Inmarsat satellites scheduled to launch in 2013-14 will provide coverage over Africa, but the Ka-band frequencies they will use are disrupted by rain. Communication failures often lead to tribal conflict, which will be a problem especially because tribes tend to follow rainfall.

Boingo, a global market leader in WiFi, is seeking to expand into Africa to allow more of the population to use VoIP as a primary means of communication. To improve their coverage, Boingo needs to find the most optimal locations to build WiFi towers by analyzing demographic and weather data. We will design a user interface for data visualization with Boingo's problem in mind.

GR1 - User and Task Analysis

GR2 - Designs

GR3 - Paper Prototyping

GR4 - Computer Prototyping

GR5 - Demo

GR6 - User Testing