focuses primarily on one of the objects that are arguments of the main predicate.
The way Joshi applied this idea is by replacing, in part, relational predicate logic with
monadic predicate logic. In each case, a restriction (from the O-machine to the finite
state machine, from the directed graph with loops, multiple edges, and cycles to the
acyclic, directed graph (tree), from the polyadic logic to the monadic logic) makes the
natural language processing problem easier because it builds in the constraints of the
problem.
It is interesting to wonder whether we process language with truly general
mechanisms or with constrained systems localized in various areas of the brain
customized for the task they undertake. I suspect that Joshi and I would agree that
the latter is more likely.
1. Aravind K. Joshi, “Relationship between Natural Language Processing and AI: Role of Constrained
Formal-Computational Systems,” AI Magazine 19(Fall 1998): (3)95-107.
More intriguing information
1. Om Økonomi, matematik og videnskabelighed - et bud på provokation2. The name is absent
3. Long-Term Capital Movements
4. Gianluigi Zenti, President, Academia Barilla SpA - The Changing Consumer: Demanding but Predictable
5. Graphical Data Representation in Bankruptcy Analysis
6. Two-Part Tax Controls for Forest Density and Rotation Time
7. Une Classe de Concepts
8. ANTI-COMPETITIVE FINANCIAL CONTRACTING: THE DESIGN OF FINANCIAL CLAIMS.
9. HOW WILL PRODUCTION, MARKETING, AND CONSUMPTION BE COORDINATED? FROM A FARM ORGANIZATION VIEWPOINT
10. The name is absent