Template-based intent classification for chatbots

A computer system comprising:

• one or more processors;
• and
• a memory coupled to the one or more processors, the memory storing instructions that,
• when executed by the one or more processors, cause the one or more processors to perform processing comprising:
• receiving an input utterance from a user;
• determining, based on a set of templates associated with two or more chatbots in a chatbot system, that the input utterance is representative of a first task that a first chatbot is configured to perform, wherein the determining comprises:
• comparing the input utterance to a first template in the set of templates, wherein:
• the first template is associated with the first chatbot, and the first template is generated by:
• identifying one or more example utterances representative of the first task that the first chatbot is configured to perform, generating, for each utterance of the one or more example utterances representative of the first task, an initial template as a regular expression that includes a string pattern, repeatedly performing at least one of sentence level or word level generalization on the generated initial templates until no further reduction of the initial templates can be performed to generate the first template, and confirming that the first template does not conflict with one or more templates associated with other chatbots in the chatbot system,
• the first template is a regular expression that includes a string pattern,
• the string pattern corresponds to a generalized form of one or more example utterances that have been provided for the first chatbot, and
• the one or more example utterances are representative of the first task that the first chatbot is configured to perform;
• calculating a degree of match between the input utterance and the first template;
• comparing the input utterance to a second template in the set of templates, wherein the second template is associated with a second chatbot in the chatbot system, and the second template is representative of a second task that the second chatbot in the chatbot system is configured to perform;
• calculating a degree of match between the input utterance and the second template;
• selecting the first chatbot over the second chatbot for processing the input utterance in response to determining that the degree of match between the input utterance and the first template is greater than the degree of match between the input utterance and the second template;
• and
• responsive to determining that the input utterance is representative of the task that the first chatbot is configured to perform, communicating at least part of the input utterance as input to the first chatbot.

A non-transitory computer-readable memory storing instructions that, when executed by one or more processors within a computer system, cause the one or more processors to perform processing comprising:

• receiving an input utterance from a user;
• determining, based on a set of templates associated with two or more chatbots in a chatbot system, that the input utterance is representative of a first task that a first chatbot is configured to perform, wherein the determining comprises:
• comparing the input utterance to a first template in the set of templates, wherein:
• the first template is associated with the first chatbot, and the first template is generated by:
• identifying one or more example utterances representative of the first task that the first chatbot is configured to perform, generating, for each utterance of the one or more example utterances representative of the first task, an initial template as a regular expression that includes a string pattern, repeatedly performing at least one of sentence level or word level generalization on the generated initial templates until no further reduction of the initial templates can be performed to generate the first template, and confirming that the first template does not conflict with one or more templates associated with other chatbots in the chatbot system,
• the first template is a regular expression that includes a string pattern,
• the string pattern corresponds to a generalized form of one or more example utterances that have been provided for the first chatbot, and
• the one or more example utterances are representative of the first task that the first chatbot is configured to perform;
• calculating a degree of match between the input utterance and the first template;
• comparing the input utterance to a second template in the set of templates, wherein the second template is associated with a second chatbot in the chatbot system, and the second template is representative of a second task that the second chatbot in the chatbot system is configured to perform;
• calculating a degree of match between the input utterance and the second template;
• and
• selecting the first chatbot over the second chatbot for processing the input utterance in response to determining that the degree of match between the input utterance and the first template is greater than the degree of match between the input utterance and the second template;
• and
• responsive to determining that the input utterance is representative of the task that the first chatbot is configured to perform, communicating at least part of the input utterance as input to the first chatbot.