Pro-Active Systems and Influenceable Users: Simulating Pro-Activity in Task-oriented Dialogues

We investigate proactivity, the capacity of a dialogue system to provide relevant information even when not explicitly requested, in the context of task-oriented dialogues. We propose to extend the current task-oriented framework, and have investigated four aspects of proactivity: (i) the degree of proactivity provided by the system during a dialogue; (ii) the propensity of the user to be influenced by the system proactivity; (iii) the complexity of the domain ontology; (iv) the relation between user needs and application domain, in terms of expected failure situations. Under the hypothesis that proactivity helps to increase effectiveness and efficiency of dialogues, we set up a framework based on dialogue simulations, and experimented the four aspects mentioned above. Although the current implementation allows to simulate a limited amount of dialogue phenomena (e.g., system initiative only), we are able to show that proactivity might have strong effects on dialogues, reducing up to 60% of dialogue turns in an an application domain of medium complexity.

This attitude obeys to the so called principles of cooperative dialogue, which have been summarized in the popular Grices’s maxims (Grice, 1975). Among the others, proactivity is very common in instruction-giving dialogues. For instance, assuming the following dialogue between speakers A and B,

(A) What time is the next train to London?

(B) The next train is at 1015. It arrives at 1245.

It arrives at 1245 is a proactive response, because information given by B is more than was asked for by A, and because B guesses that this is the sort of information A might also need, and so offers it unsolicited.

very common in human-human dialogues, proactivity is almost absent from the current research in task-oriented dialogue systems. To the best of our knowledge, proactivity is largely under represented in most of the datasets