Learning, search, social computation and networks
AbstractFormal Concept Analysis (FCA) begins from a context, given as a binary relation between some objects and some attributes, and derives a lattice of concepts, where each concept is given as a set of objects and a set of attributes, such that the first set consists of all objects that satisfy all attributes in the second, and vice versa. Many applications, though, provide contexts with quantitative information, telling not just whether an object satisfies an attribute, but also quantifying this satisfaction. Contexts in this form arise as rating matrices in recommender systems, as occurrence matrices in text analysis, as pixel intensity matrices in digital image processing, etc. Such applications have attracted a lot of attention, and several numeric extensions of FCA have been proposed. We propose the framework of proximity sets (proxets), which subsume partially ordered sets (posets) as well as metric spaces. One feature of this approach is that it extracts from quantified contexts quantified concepts, and thus allows full use of the available information. Another feature is that the categorical approach allows analyzing any universal properties that the classical FCA and the new versions may have, and thus provides structural guidance for aligning and combining the approaches.
AbstractIn a previous FAST paper, I presented a quantitative model of the process of trust building, and showed that trust is accumulated like wealth: the rich get richer. This explained the pervasive phenomenon of adverse selection of trust certificates, as well as the fragility of trust networks in general. But a simple explanation does not always suggest a simple solution. It turns out that it is impossible to alter the fragile distribution of trust without sacrificing some of its fundamental functions. A solution for the vulnerability of trust must thus be sought elsewhere, without tampering with its distribution. This observation was the starting point of the present paper. It explores a different method for securing trust: not by redistributing it, but by mining for its sources. The method used to break privacy is thus also used to secure trust. A high level view of the mining methods that connect the two is provided in terms of similarity networks, and spectral decomposition of similarity preserving maps. This view may be of independent interest, as it uncovers a common conceptual and structural foundation of mathematical classification theory on one hand, and of the spectral methods of graph clustering and data mining on the other hand.
AbstractGame theoretic equilibria are mathematical expressions of rationality. Rational agents are used to model not only humans and their software representatives, but also organisms, populations, species and genes, interacting with each other and with the environment. Rational behaviors are achieved not only through conscious reasoning, but also through spontaneous stabilization at equilibrium points.
Formal theories of rationality are usually guided by informal intuitions, which are acquired by observing some concrete economic, biological, or network processes. Treating such processes as instances of computation, we reconstruct and refine some basic notions of equilibrium and rationality from the some basic structures of computation.
It is, of course, well known that equilibria arise as fixed points; the point is that semantics of computation of fixed points seems to be providing novel methods, algebraic and coalgebraic, for reasoning about them.