Variable Binding
- Variable binding is a problem for connectionist systems in general.
- It is a short-term binding of two or more concepts.
- A sample problem would be the presentation of a Red-Square, and
a Blue-Circle.
- All 4 concepts are active, but how does the system know the
colour of the square? That is, how is Red bound to Square?
- Activation alone cannot be the solution.
- The variable binding problem occurs in the above example, but
also occurs with:
- multiple presentations of the same object (e.g. 3 squares)
- any rule based behaviour
- language in particular
- A complete model of memory needs to solve the variable binding
problem.
