Roussanka Loukanova: Situation Theory and its Applications
Tid: On 2014-12-10 kl 13.30 - 15.00
Plats: Room 15, building 5, Kräftriket, Department of mathematics, Stockholm university
Situation Theory is a powerful, highly expressive theory of finely-grained information that is partial, underspecified, and situational. Its most prominent application, known as Situation Semantics, is to computational semantics of human language. Recently, new applications of Situation Theory have been emerging and proliferating, most distinctively in areas such as models of context in computational sciences, Artificial Intelligence, Natural Language Processing, neuroscience of language and other sub-areas of computational neuroscience. Despite of diversity of its applications, mathematical foundation of Situation Theory had legged behind them. We will consider both its new applications and a new approach to its mathematics.
I will introduce the major concepts of Situation Theory, as it was initiated originally by Barwise and Perry in 80s --- as a relational model-theory of partial information with content. Then we will discuss a new approach to its major concepts:
- Primitive relations, argument roles, saturation of argument roles
- Types for situational information
- Basic information units
- Situations and events
- Situated propositions
- Complex information units
- Situation types
- Partial information
- Semantic parameters as typed objects: primitive, complex, restricted
- Parametric information
The concepts of situated, partial, and parametric information are central for Situation Theory. Parametric information is modeled by generalized, basic or complex, parametric objects. We discuss restricted parameters with constraints that include information with situated components.
While I give the introduction to Situation Theory by using examples from human language, the importance of its development is broader, to computational technologies in other emerging areas. We will touch upon them.
My target is to investigate and develop formalization of Situation Theory with Dependent Type Theory as its mathematical foundation, for modern applications to advanced technologies.