AUTOMATIC SPEAKER ROLE LABELING IN AMI MEETINGS: RECOGNITION OF FORMAL AND SOCIAL ROLES

This work aims at investigating the automatic recognition of speaker role in meeting conversations from the AMI corpus. Two types of roles are considered: formal roles, fixed over the meeting duration and recognized at recording level, and social roles related to the way participants interact between themselves, recognized at speaker turn level. Various structural, lexical and prosodic features as well as Dia-log Act tags are exhaustively investigated and combined for this pur-pose. Results reveal an accuracy of 74 % in recognizing the speakers formal roles and an accuracy of 66 % (percentage of time) in cor-rectly labeling the social roles. Feature analysis reveals that lex-ical features provide the higher performances in formal/functional role recognition while prosodic features provide the higher perfor-mances in social role recognition. Furthermore results reveal that social role recognition in case of rare roles in the corpus can be im-proved through the use of lexical and Dialog Act information com-bined over short time windows