Abstract
This paper investigates proactivity, a characteristic phenomenon of collaborative human-humaninteraction, where a participant in the dialogue offers the addressee some useful and not explicitlyrequested information. More precisely, a proactive behaviour is: (i) self-prompted and not simplyreactive, that is, the speaker does not act merely in response to the requests the other participanthas made; (ii) somehow effective for the achievement of the dialogue goal, since the speaker has along-term, goal-directed behaviour that predicts future states and needs. Proactivity has been poorlyinvestigated from a theoretical point of view, and there is a general need of empirical data for bothquantitative and qualitative research. The paper provides an extensive analysis of proactivity in sev-eral human-human task-oriented dialogic corpora, selected with different characteristics, includingchat exchanges and telephone calls, collection modalities such as natural setting and Wizard of Oz,and two languages, Italian and English. The main result is theD-Pro Corpus, a new resource manually annotated at the utterance level with proactivity and dialogue acts, which allows to investigateproactivity in the context of task-oriented dialogues. There are several findings from our empiricalinvestigation of proactivity: (i) we find that about 20% of turns in our corpus are proactive turns,showing that this is a very diffused and relevant phenomenon; (ii) we confirm the non-reactivenature of proactivity, highlighting the presence of a pattern where a turn in the dialogue triggersa reaction in a following turn and a proactive utterance is then added to the turn; (iii) we showthat only a limited number of dialogue acts are actually involved in expressing proactivity, andwe discuss the theoretical implications of this finding; (iv) we empirically confirm that proactivityhas a crucial role in recovering from goal-failure situations, contributing to the effectiveness of thewhole dialogue; (v) we support the intuition of a non-uniform distribution of proactive utterancesthroughout the dialogue. Our empirical findings and the D-Pro Corpus provide relevant insights fordeeper theoretical investigations, as well as crucial resources for improving proactivity in currenttask-oriented dialogue systems.