Dr. Elmalech Avshalom

Crowdsourcing
My interest in crowdsourcing is primarily focused on finding ways to motivate people to do quality work for their employer and ways to automatically monitor the quality of their work. My work in this field has two different dimensions. First, I am working to identify the kinds of tasks in which human labor is more beneficial than machine labor. Second, I am working to create efficient algorithms to motivate people and mechanisms to compensate people.

Recommendation systems
I am interested in finding innovative methods for collaborative agents to better assist people facing complex decision situations. In particular, I am fascinated by the cognitive biases of people and the way these can be used for pushing them to better decisions, e.g., through intelligently restructuring the decision situation presented to them or selectively disclosing the information available to them. Recently, I proposed a new paradigm for automated advice provisioning agents, one in which the agent should not necessarily provide what it believes to be the utility-maximizing advice, but rather, it should suggest sub-optimal advice that is more likely to be adopted. In other words, by considering and modeling the likelihood that the advised subject will adopt the advice given and/or the way she will be affected by the advice given, the effectiveness of the advice given can be substantially improved. My work presented a paradigm shift in what advice these agents should provide people.

Agents and people
I am interested in intelligent user interfaces, both for adversarial and cooperative settings. Much of my work in this area deals with developing methods for selective information disclosure to be used by autonomous agents when interacting with people in order to influence them. Agents and platforms can use these methods to improve the benefits of people or agents using their services, to attract more customers, and to benefit from their users' overconsumption of their services.

Agent-based simulations
My interest in agent-based simulations mainly focuses on finding ways to reliably represent and model people and bounded rational agents acting on behalf of people. One turn-key technology I am especially interested in is PDAs. A PDA is a peer-designed agent—a computer agent developed by regular programmers, not decision-making experts. In recent research, I have managed to demonstrate the effectiveness of this technology in generating a variety of realistic strategies and behaviors, as well as how it can be used to alleviate the simulation and behavior analysis of systems populated by human individuals with diverse strategies.