Abstract
In this final Chapter we discuss what further challenges machines need to meet when learning how to engage in useful, meaningful, and natural dialogues with humans. In contrast to the approach adopted by many “chatbots” (see Figure 10), which typically use shallow pattern-matching techniques to retrieve a next system move, we argue that coherent dialogue actions can only be chosen when the system knows the likely meanings of its dialogue contributions, and has a representation of dialogue states which allows it to compute the utility of each of those moves.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
Copyright information
© 2011 Springer-Verlag Berlin Heidelberg
About this chapter
Cite this chapter
Rieser, V., Lemon, O. (2011). Conclusion. In: Reinforcement Learning for Adaptive Dialogue Systems. Theory and Applications of Natural Language Processing. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-642-24942-6_10
Download citation
DOI: https://doi.org/10.1007/978-3-642-24942-6_10
Published:
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-642-24941-9
Online ISBN: 978-3-642-24942-6
eBook Packages: Computer ScienceComputer Science (R0)