US 11,386,885 B2Granted

Method and system for detecting intent as an ordered sequence from a user query

AssigneeWipro Limited

InventorsArindam Chatterjee, Pratik Kumar Pandey

Life of this patent

Estimating the term…

Loading patent…

Abstract

Disclosed herein is method and system for detecting intent as an ordered sequence from user query. The system identifies word embedding feature for each word using word embedding model and identifies Part of Speech tag feature. Thereafter, system determines dependency label feature based on dependency role and POS tag feature. Further, system provides feature vector comprising POS tag feature of target word, POS tag feature of previous two words of target word, word embedding feature of target word, word embedding feature of head word for each target word and dependency label feature of target word to deep neural network for detecting intent as ordered sequence. The ordered sequence includes desired action in user query, focal point pertaining to which desired action must be performed and one or more descriptors associated with focal point. In this manner, in present disclosure, overall intent from user query is captured for accurately providing response.

Claims15

Dependent claims

A method for detecting intent as an ordered sequence from a user query, the method comprising:

identifying, by an intent determination system, a word embedding feature for each word in a user query using a word embedding model;
identifying, by the intent determination system, a Part of Speech (POS) tag feature for each word in the user query;
determining, by the intent determination system, a dependency label feature for each word in the user query based on dependency role of each word in the context of the user query and POS tag feature of each word in the user query;
providing, by the intent determination system, a feature vector for the user query comprising the POS tag feature of each target word, POS tag feature of previous two words of each target word in the user query, word embedding feature of each target word in the user query, word embedding feature of a head word for each of the target word in the user query and the dependency label feature of the target word to a deep neural network, wherein the head word for each of the target word is determined using the dependency label feature of each word in the user query;
and
detecting, by the intent determination system, the intent as an ordered sequence, based on the feature vector using a deep neural network, comprising a desired action in the user query, a focal point pertaining to which the desired action has to be performed and one or more descriptors associated with the focal point in the user query.

3 more elements

An intent detection system for detecting intent as an ordered sequence from a user query, the intent detection system comprising:

a processor;
and
a memory communicatively coupled to the processor, wherein the memory stores the processor-executable instructions, which, on execution, causes the processor to:
identify a word embedding feature for each word in a user query using a word embedding model;
identifying, by the intent determination system, a Part of Speech (POS) tag feature for each word in the user query;
determine a dependency label feature for each word in the user query based on dependency role of each word in the context of the user query and POS tag feature of each word in the user query;
provide a feature vector for the user query comprising the POS tag feature of each target word, POS tag feature of previous two words of each target word in the user query, word embedding feature of each target word in the user query, word embedding feature of a head word for each of the target word in the user query, and the dependency label feature of the target word to a deep neural network, wherein the head word for each of the target word is determined using the dependency label feature of each word in the user query;
and
detect the intent as an ordered sequence, based on the feature vector using a deep neural network, comprising a desired action in the user query, a focal point pertaining to which the desired action has to be performed and one or more descriptors associated with the focal point in the user query.

6 more elements

A non-transitory computer readable medium including instructions stored thereon that when processed by at least one processor causes the processor to:

identify a word embedding feature for each word in a user query using a word embedding model;
identify a Part of Speech (POS) tag feature for each word in the user query;
determine a dependency label feature for each word in the user query based on dependency role of each word in the context of the user query and POS tag feature of each word in the user query;
provide a feature vector for the user query comprising the POS tag feature of each target word, POS tag feature of previous two words of each target word in the user query, word embedding feature of each target word in the user query, word embedding feature of a head word for each of the target word in the user query and the dependency label feature of the target word to a deep neural network, wherein the head word for each of the target word is determined using the dependency label feature of each word in the user query;
and
detect the intent as an ordered sequence, based on the feature vector using a deep neural network, comprising a desired action in the user query, a focal point pertaining to which the desired action has to be performed and one or more descriptors associated with the focal point in the user query.

3 more elements

Dependencies are inferred from the claim text.