UVCA: a Unified Vocal Commands Approach for Mobile Applications

Abstract

Humans prefer to interact with their smart devices through speech rather than using pointing tools and keyboards. Recent mobile devices provide the capability to automatically convert voice to text but they still suffer from many issues when it comes to efficiency goal. In this paper, we present a new approach for vocal commands processing based on Information Retrieval concepts (i.e. text pre-processing, cosine similarity) as well as machine learning algorithms and which enables mobile smart devices to better and faster access to applications' functionalities.