Enabling autonomous agents to discriminate between questions and requests

A system comprising:

• a non-transitory computer-readable medium storing computer-executable program instructions;
• and
• a processing device communicatively coupled to the non-transitory computer-readable medium for executing the computer-executable program instructions, wherein executing the computer-executable program instructions configures the processing device to perform operations comprising:
• accessing an utterance comprising text;
• generating a parse tree for the utterance, wherein the parse tree comprises at least one terminal node comprising a word type, wherein each terminal node represents a word or phrase of the utterance and wherein the parse tree identifies syntactic information of the utterance;
• applying a classification model to the parse tree, wherein the classification model was trained by:
• accessing a set of training data comprising a set of training data pairs, wherein each training data pair comprises text and an expected classification, and wherein the set of training data includes both (i) a first training data pair that comprises features indicating a request and (ii) a second training data pair that comprises features indicating a question;
• and
• iteratively providing one of the training data pairs to the classification model and adjusting internal parameters of the classification model based on a difference between a determined classification and the expected classification,
• wherein the applying includes performing actions comprising:
• responsive to identifying, in the parse tree, an imperative verb as a first word of the utterance by determining that (a) a tense of a verb in the parse tree is present tense or that a voice of the verb is active voice and that (b) the verb is neither modal, mental, or a form of the verb “to be,” classifying the utterance as a request;
• responsive to identifying, in the parse tree, a first person pronoun and that the verb in the parse tree is a mental verb, classifying the utterance as a request;
• and
• responsive to identifying that a verb in the parse tree is an imperative verb and that the parse tree lacks a first person pronoun, classifying the utterance as a question;
• receiving, from the classification model, a classification of a question, a request or unknown;
• and
• based on the classification, sending a message to a user device or adjust a configuration of an external device.

A non-transitory computer-readable medium storing computer-executable program instructions that when executed by a processor, perform operations comprising:

• accessing an utterance comprising text;
• generating a parse tree for the utterance, wherein the parse tree comprises at least one terminal node comprising a word type, wherein each terminal node represents a word or phrase of the utterance and wherein the parse tree identifies syntactic information of the utterance;
• applying a classification model to the parse tree, wherein the classification model was trained by:
• accessing a set of training data comprising a set of training data pairs, wherein each training data pair comprises text and an expected classification, and wherein the set of training data includes both (i) a first training data pair that comprises features indicating a request and (ii) a second training data pair that comprises features indicating a question;
• and
• iteratively providing one of the training data pairs to the classification model and adjusting internal parameters of the classification model based on a difference between a determined classification and the expected classification,
• wherein the applying includes performing actions comprising:
• responsive to identifying, in the parse tree, an imperative verb as a first word of the utterance by determining that (a) a tense of a verb in the parse tree is present tense or that a voice of the verb is active voice and that (b) the verb is neither modal, mental, or a form of the verb “to be,” classifying the utterance as a request;
• responsive to identifying, in the parse tree, a first person pronoun and that the verb in the parse tree is a mental verb, classifying the utterance as a request;
• and
• responsive to identifying that a verb in the parse tree is an imperative verb and that the parse tree lacks a first person pronoun, classifying the utterance as a question;
• receiving, from the classification model, classification of a question, a request, or unknown;
• and
• based on the classification, send a message to a user device or adjust a configuration of an external device.