Comparative analysis of verbal alignment in human-human and human-agent interactions