Dialogue system and domain determination method

1. A dialogue system, comprising:

an input unit configured to acquire utterance contents of a user in a dialogue;

a mode determining unit configured to determine, based on the utterance contents acquired by the input unit, whether a mode of the dialogue is task-oriented or non-task-oriented;

a plurality of intention understanding units each corresponding to a specific domain; and

a domain determining unit configured to

transmit, when the mode of the dialogue is task-oriented, the utterance contents to all of the plurality of intention understanding units, and determine a domain of the dialogue based on a result of intention understanding of the utterance contents performed using each of the plurality of intention understanding units, and

determine, when the mode of the dialogue is non-task-oriented, the domain of the dialogue prior to transmitting the utterance contents to any of the intention understanding units.

2. The dialogue system according to claim 1 , wherein

each of the plurality of intention understanding units outputs a certainty factor of the utterance contents being consistent with each of a plurality of intentions determined in advance.

3. The dialogue system according to claim 2 , wherein

each of the plurality of intention understanding units vectorizes the utterance contents, and determines, based on a distance between a position of the utterance contents and a center of gravity position of an intention determined in advance in a vector space, a certainty factor corresponding to the intention.

4. The dialogue system according to claim 2 , wherein

each of the plurality of intention understanding units outputs a certainty factor for a prescribed number of intentions starting from an intention with a highest certainty factor, and

the prescribed number is common among the plurality of intention understanding units.

5. The dialogue system according to claim 2 , wherein

each of the plurality of intention understanding units calculates the certainty factor based on same calculation criteria.

6. The dialogue system according to claim 2 , wherein

the domain determining unit determines, as the domain of the dialogue, a domain corresponding to an intention understanding unit having output a highest certainty factor.

7. The dialogue system according to claim 1 , wherein

the domain determining unit estimates a domain from a word string included in the utterance contents when the mode of the dialogue is non-task-oriented.

8. The dialogue system according to claim 1 , wherein

when the mode of the dialogue is switched between the task-oriented mode and the non-task-oriented mode, the domain determining unit determines a domain after the switch by taking into consideration a domain before the switch.

9. The dialogue system according to claim 1 , wherein

the input unit includes: a speech input unit configured to acquire uttered speech of the user; and a speech recognizing unit configured to apply speech recognition to the uttered speech to acquire the utterance contents.

10. The dialogue system according to claim 1 , wherein

the input unit acquires a text of utterance contents from the user.

11. A domain determination method carried out by a dialogue system, the domain determination method comprising:

an input step of acquiring utterance contents of a user in a dialogue;

a mode determining step of determining, based on the utterance contents acquired in the input step, whether a mode of the dialogue is task-oriented or non-task-oriented;

an intention understanding step of performing intention understanding by a plurality of intention understanding units each corresponding to a specific domain; and

a domain determining step of determining a domain of the dialogue, wherein

when the dialogue is task-oriented, the utterance contents are transmitted to all of the plurality of intention understanding units, and the domain of the dialogue is determined based on a result of intention understanding performed using the intention understanding units each corresponding to the specific domain, and

when the mode of the dialogue is non-task-oriented, the domain of the dialogue is determined prior to transmitting the utterance contents to any of the intention understanding units.