Human factor research of voice search on mobile internet

Abstract

Voice search is a kind of promising search approach in the era of mobile internet. Currently, most of voice search services are still manual work, which are expected to be replaced by human-machine voice search based on voice recognition technology. Despite numerous disadvantages of human-human voice search, its interaction approach as natural language communication is obviously superior to human-machine voice interaction. In order to get better human-machine voice search interaction scheme, we conduct an investigation into current human-human voice search service with the purpose of understanding human habits and patterns in using voice search, elements that impact user experience, and getting enlightenment for designing automatic voice search interaction. Based on the findings, we continue to explore underlying rules and principles by building models which include two dimension model on search intention, state transition model and user experience hierarchical model. Finally, we propose a blueprint of automatic voice search interaction scheme according to investigation findings and models, hoping it could elicit better design.