# US 12,002,454 B2 · Method and apparatus for intent recognition and intent prediction based upon user interaction and behavior > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 12,002,454 B2 - Document ID: US12002454B2 - Status: Granted - Kind: B2 - Application number: 17/126,703 - Filed: 2020-12-18 - Published: 2021-06-24 - Granted: 2024-06-04 - Earliest claimed: 2019-12-20 - Assignees: Swampfox Technologies, Inc. - Applicants: Swampfox Technologies, Inc. - Inventors: Sergey A. Razin; Robert S. Cooper; Rick Ulmer; Tom Hanson - Web page: https://patentgrep.com/patent/US12002454B2 ## CPC - G10L 15/18: using natural language modelling - G06F 40/30: Semantic analysis - G10L 15/063: {Training} - G10L 15/22: Procedures used during a speech recognition process, e.g. man-machine dialogue - G10L 15/30: Distributed recognition, e.g. in client-server systems, for mobile phones or network applications - G10L 15/02: Feature extraction for speech recognition; Selection of recognition unit - G10L 15/1822: {Parsing for meaning understanding} - G10L 15/183: using context dependencies, e.g. language models - G10L 2015/223: {Execution procedure of a spoken command} - G10L 2015/225: {Feedback of the input speech} - G10L: SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING - G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N) ## Abstract Embodiments of the innovation relate to, in a contact center apparatus, a method for recognizing user intent associated with user interaction with the contact center apparatus. The method includes receiving user interaction data; performing a feature extraction process on the user interaction data to generate feature data; performing an intent extraction operation on the feature data to extract topics included with the feature data; executing a classification engine on the topics extracted from the feature data to classify a user intent associated with the user interaction data; and directing the user to a corresponding working agent based upon the classified user intent ## Claims ### Claim 1 In a contact center apparatus, a method for recognizing user intent associated with user interaction with the contact center apparatus, comprising: receiving user interaction data; performing a feature extraction process on the user interaction data to generate feature data; performing an intent extraction operation on the feature data to extract topics included with the feature data; executing a classification engine on the topics extracted from the feature data to classify a user intent associated with the user interaction data; performing an intent prediction operation on the user interaction data to generate an intent prediction, comprising: transforming the user interaction data into a prediction vector having the format ; providing the prediction vector and a prediction model to a framework; and receiving the intent prediction and a confidence score from the framework based upon the prediction vector and the prediction model; generating a next best action based upon the intent prediction and a confidence score, the next best action configured as an automated action performed by the contact center apparatus to act on the intent prediction, the next best action including an advertisement directed toward the user; and providing the next best action to the user. ### Claim 2 (depends on 1) The method of claim 1, wherein extracting features from the user interaction data comprises: utilizing a data cleansing engine to reduce the size of the data set included in the user interaction data and to generate cleansed user interaction data; and applying a feature importance engine to the cleansed user interaction data to identify statistically important words associated with the user interaction data. ### Claim 3 (depends on 1) The method of claim 1, wherein performing the intent extraction operation on the feature data comprises applying Gensim's Latent Dirichlet Allocation (LDA) model to the feature data to extract topics included with the feature data. ### Claim 4 (depends on 1) The method of claim 1, wherein executing the classification engine further comprises confirming an accuracy of the classification of the topics extracted from the feature data and associated with the user interaction data. ### Claim 5 (depends on 1) The method of claim 1, further comprising, in response to receiving a user response to the next best action, providing response data to the framework to reinforce the prediction model. ### Claim 6 In a contact center apparatus, a method for predicting user intent associated with user interaction with the contact center apparatus, comprising: receiving user interaction data; transforming the user interaction data into a prediction vector having the format ; providing the prediction vector and a prediction model to a framework; receiving an intent prediction and a confidence score from the framework based upon the prediction vector and the prediction model; generating a next best action based upon the intent prediction and a confidence score, the next best action configured as an automated action performed by the contact center apparatus to act on the intent prediction, the next best action including an advertisement directed toward the user; and providing the next best action to the user. ### Claim 7 (depends on 6) The method of claim 6, further comprising, in response to receiving a user response to the next best action, providing response data to the framework to reinforce the prediction model. ### Claim 8 A contact center apparatus, comprising: a controller having a processor and a memory, the controller configured to: receive user interaction data; perform a feature extraction process on the user interaction data to generate feature data; perform an intent extraction operation on the feature data to extract topics included with the feature data; wherein when performing the intent prediction operation on the user interaction data the controller is configured to: transform the user interaction data into a prediction vector having the format ; provide the prediction vector and a prediction model to a framework; and receive the intent prediction and a confidence score from the framework based upon the prediction vector and the prediction model; execute a classification engine on the topics extracted from the feature data to classify a user intent associated with the user interaction data; perform an intent prediction operation on the user interaction data to generate an intent prediction; generate a next best action based upon the intent prediction and a confidence score, the next best action configured as an automated action performed by the contact center apparatus to act on the intent prediction, the next best action including an advertisement directed toward the user; and provide the next best action to the user. ### Claim 9 (depends on 8) The contact center apparatus of claim 8, wherein when extracting features from the user interaction data, the controller is configured to: utilize a data cleansing engine to reduce the size of the data set included in the user interaction data and to generate cleansed user interaction data; and apply a feature importance engine to the cleansed user interaction data to identify statistically important words associated with the user interaction data. ### Claim 10 (depends on 8) The contact center apparatus of claim 8, wherein when performing the intent extraction operation on the feature data, the controller is configured to apply Gensim's Latent Dirichlet Allocation (LDA) model to the feature data to extract topics included with the feature data. ### Claim 11 (depends on 8) The contact center apparatus of claim 8, wherein when executing the classification engine the controller is further configured to confirm an accuracy of the classification of the topics extracted from the feature data and associated with the user interaction data. ### Claim 12 (depends on 8) The contact center apparatus of claim 8, wherein the controller is further configured to, in response to receiving a user response to the next best action, provide response data to the framework to reinforce the prediction model. ### Claim 13 A contact center apparatus, comprising: a controller having a processor and a memory, the controller configured to: receive user interaction data; transform the user interaction data into a prediction vector having the format ; provide the prediction vector and a prediction model to a framework; receive an intent prediction and a confidence score from the framework based upon the prediction vector and the prediction model; generate a next best action based upon the intent prediction and a confidence score, the next best action configured as an automated action performed by the contact center apparatus to act on the intent prediction, the next best action including an advertisement directed toward the user; and provide the next best action to the user. ### Claim 14 (depends on 13) The contact center apparatus of claim 13, wherein the controller is further configured to, in response to receiving a user response to the next best action, provide response data to the framework to reinforce the prediction model.