An Algorithm for Resolving Natural Language Ambiguity in Human-Robot Interaction.

Abstract

The use of social robots is fast becoming an essential part of human life. This is because social robots are becoming more ubiquitous and they are capable of performing complex tasks. For instance, social robots are now widely used to perform collaborative tasks with human beings in healthcare, education, entertainment, military as well as search and rescue missions where safety is a major concern. The interaction between human beings and social robots requires effective communication for the successful completion of a specific task. Nevertheless, effective, seamless and more flexible communication between human beings and social robots can be easily achieved through natural languages. This is because natural languages do not require complex graphical interfaces and environments, the knowledge of programming as well as extensive training. However, some words in natural languages are highly ambiguous, polysemous and their interpretations depend on the context in which they are used. Unfortunately, it is quite difficult for a social robot to capture the meaning of a word in the context in which it is used. This usually results in communication errors and thus a misinterpretation of the human intention which ultimately leads to the users’ dissatisfaction. This reduces the acceptance rate of social robots in the society. Hence, this paper proposes an algorithm that explicitly specifies the semantics of words in natural languages between humans and social robots. This is with a view to resolving ambiguities and facilitating seamless communication and natural interaction between human beings and social robots.