Bot Dialog Manager Context Handling via Phrase Vectors
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Interactive voice response systems and chatbots face challenges in providing contextually relevant responses due to the ambiguity of natural language user inputs, where phrases can have different meanings based on the dialog context, leading to difficulties in identifying the correct task and maintaining context across different communication types.
Innovation Solution
A method that converts user inputs into phrase vectors, identifies tokens using a token identification engine, selects unsupervised and supervised tokens, and determines a voted token based on recent tokens, allowing for contextually relevant responses across voice and email communications.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If natural language user inputs are processed directly, then the system can respond to user queries, but the ambiguity of phrases leads to incorrect task identification and loss of context
Solution Approach 1:
The patent segments the user input processing into multiple components: phrase vector generation, token identification, unsupervised token selection, and supervised token selection. Each component handles a specific aspect of context extraction, collectively resolving the ambiguity of natural language inputs while maintaining task identification accuracy.
Solution Approach 2:
The patent introduces phrase vectors as an intermediary representation between natural language input and task identification. The phrase vector captures the semantic meaning and context of the input, serving as a mediator that preserves context information while enabling accurate task identification despite phrase ambiguity.
2Measurement precision
If multiple token selection methods are used (unsupervised and supervised), then context accuracy is improved, but the system complexity increases
Solution Approach 1:
The patent merges unsupervised token selection and supervised token selection into a unified framework. Both methods operate in parallel on the same phrase vector and token set, with their results combined to identify the final next token. This merging approach improves context identification accuracy while managing system complexity through integrated architecture.
Solution Approach 2:
The patent creates a multi-functional token identification system that performs both unsupervised and supervised token selection through a single integrated process. The same phrase vector and token identification engine serve both selection methods, allowing the system to achieve high measurement precision without proportionally increasing device complexity.
Data Source
AI summary
A method manages bot dialogue. A user input is converted to a phrase vector. A set of identified tokens are identified by a token identification engine from the phrase vector. An unsupervised token is selected from the set of identified tokens. A supervised token is selected from the set of identified tokens. A voted token selected from the unsupervised token and the supervised token. A next token is identified based on a set of recent tokens that includes the voted token. The next token is presented as one of a voice communication and an email communication.