Mapping illocutionary speech in human interaction to human-machine interaction

Abstract

With the current development of automated technology, it is becoming more important to facilitate interaction between humans and machines. Human-computer interaction can be discussed and analysed from different points of view, for example psychology, cognitive science, or communication. This project studied and analysed two influential human-to-human communication theories, Speech Acts Theory and Theory of Communicative Action, and adapted them to create ideal guidelines for human-to-machine interaction, especially for the case of interactive machines. The guidelines called Framework version 1.0 were then used to analyse four existing interactive machines of different functions to see how the machines score against the guidelines.

Framework version 1.0 was used to design an improved ATM flow. The improved ATM flow was then validated through tests. The outcome of the tests revealed that only two points in the framework need to be revised. The first one is that the usage of common indirect speech acts is permitted while the second one advises the usage of familiar and common directives on top of unambiguous directives. An improved framework called Framework 1.1 was thus introduced. The results of the validation tests showed that the framework is effective for human-to-machine interaction.

At the end of the report, future work was proposed to further enhance the framework. The Framework version 1.1 can be tested on other interactive machines of different functions from those of an ATM.