Flexible Interpretations: A Computational Model for Dynamic Uncertainty Assessment

In particular, we are interested here in the nature of the control structure of computer programs that can support multiple interpretation and smooth transitions between them, in real time. Each step of the processing involves the interpretation of one input item and the appropriate re-establishment of the system's confidence of the correctness of its interpretation(s). First, the input to the program may be highly unreliable either due to noise at the input channel or due to excessive irrelevant information. We have developed two computational schemes to deal with uncertainty during interpretation tasks. The most obvious advantage of using this scheme is that it may provide a broader interpretation of the situation and helps reduce biases so that relevant new information is not missed.