Julie S. Weber

A biometric attendance terminal and its application to health programs in India

Tracking attendance is a necessity in a variety of contexts in the developing world, encompassing health programs, schools, government offices, and a litany of other milieux. While electronic attendance tracking systems exist and perform their core function well, they are expensive, monolithic and offer little customizability. In this paper we describe a fingerprint-based biometric attendance system implemented using off-the-shelf components: a netbook computer, a commodity fingerprint reader, and a low-cost mobile phone. The system identifies visitors based only on their fingerprint, and uploads attendance logs to a central location via SMS. Its functionality goes beyond that of existing market offerings while improving modularity, extensibility, and cost of ownership. We deployed this system in two health programs - supporting tuberculosis patients in New Delhi and sex workers in Bangalore -- and logged over 550 users and 4,500 visits over the course of several months. Our experience suggests that the system is usable in real-world contexts, though incentives are needed to sustain usage over time. We reflect on the sociocultural factors surrounding adoption and describe the potential to impact health outcomes in the future.

Usability Engineering Methods for Interactive Intelligent Systems

The field of Human-Computer Interaction (HCI) offers designers and developers of interactive systems a large repertoire of methods for ensuring that their systems will be both usable and useful. This article offers a brief introduction to these methods, focusing on the ways in which they sometimes need to be adapted and extended to take into account the characteristic properties of systems that include some sort of AI. The discussion is organized around three types of activity: understanding users needs, interaction design, and evaluation.

Simulating users to support the design of activity management systems

We describe a simulation system that models the user of a calendar-management tool. The tool is intended to learn the users scheduling preferences, and we employ the simulator to evaluate learning strategies. The simulated user is instantiated with a set of preferences over local and global features of a schedule such as the level of importance of a particular meeting and the amount of preparation time available before it is to begin. The system then processes a set of simulated meeting requests, and over time and through user feedback, it learns the users preferences, affording it the ability to thereafter manage the users schedule more autonomously