Effective Organization of a Motivated Community to Obtain Effective Algorithmic Solutions
College of Computer and Information Science, Northeastern University, Boston.
Joint Work with Ahmed Abdelmeged
and Ruiyang Xu.
We introduce a new game, called side-choosing game, that helps to organize computational problem solving communities and, more generally, constructive formal scientific communities. We study the social choice theory of side-choosing games, including appropriate axioms, a representation theorem and a meritocracy theorem. While traditional social-choice theory studies game-theoretic models of political institutions we study game-theoretic models of formal scientific communities.
Effective algorithmic solutions are in high demand
in numerous domains, such as big data, and we claim that side-choosing games
are an ideal tool to use Human Computation to find effective algorithmic solutions.
Side-choosing games offer the benefit of fair, collusion-resistant peer evaluation without a central authority, thereby lowering the effort of organizing competitions similar to the ones held on TopCoder.com. Side-choosing games are educational for players, giving them targeted feedback on their choices and their defense. We report on our experience in building computational problem solving communities in algorithm and software development courses as a first application of our theory.
Short Bio of Speaker:
is a Professor in the College of Computer and
Information Science at Northeastern University.
He has contributed to
Algorithms (P-optimal Algorithms, Clause Learning for Satisfiability)
Modular Software Design (Law of Demeter and several systems
for Adaptive Programming, a kind of Aspect-Oriented Programming).
His latest interest is in systems and their foundations for
Human Computation for complex tasks, e.g., algorithm development.