Incremental Intention Recognition
- 963 Downloads
This chapter describes a novel method for incremental intention recognition. It is performed and achieved by incrementally constructing a threelayer Bayesian Network model as more actions are observed. The Bayesian Network is composed from a knowledge base of readily specified and readily maintained Bayesian Network fragments. Their simple structures allow easy maintenance by domain experts and also facilitate automatic construction from available plan corpora. Improved performance of the method on the Linux plan corpus case is exhibited.
KeywordsBayesian Network Action Node Conditional Probability Distribution Bayesian Network Model Plan Session
Unable to display preview. Download preview PDF.