Intent Identification Using Historical Context Vectors
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current user intent identification methods in human-computer interaction systems are inaccurate, leading to inconsistent responses and a negative user experience, as they rely solely on the analysis of the current statement without considering historical context.
Innovation Solution
A method that determines user intent by analyzing a to-be-analyzed statement in conjunction with historical statements from the same HCI session, using a historical vector calculated from the statement vectors of both, which includes semantic similarity and positional relationships to improve intent accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the machine uses only the text included in the current question as a reference for identifying user intent, then the processing speed is fast and the system complexity is low, but the identified user intent is inaccurate leading to inconsistent answers
Solution Approach 1:
The system performs preliminary action by pre-processing historical statements and generating historical vectors before they are needed for intent identification. The historical vector generation unit continuously updates historical vectors based on previous interactions, so when a new question arrives, the context is already prepared and available for immediate use in improving intent accuracy without adding significant processing delay.
Solution Approach 2:
The patent introduces a new dimension to the intent identification process by incorporating historical context vectors alongside current question analysis. Instead of relying solely on the current question text, the system creates a multi-dimensional input that includes both current statement features and historical interaction patterns, thereby enhancing identification accuracy through additional contextual dimensions.
2Measurement precision
If the machine analyzes only the current statement without historical context, then the processing time is short, but the user intent identification accuracy is low resulting in negative user experience
Solution Approach 1:
The system performs preliminary action by pre-processing historical statements and generating historical vectors before they are needed for intent identification. The historical vector generation unit continuously updates historical vectors based on previous interactions, so when a new question arrives, the context is already prepared and available for immediate use in improving intent accuracy without adding significant processing delay.
Solution Approach 2:
The system maintains continuity of useful action by continuously updating and maintaining historical vectors as the conversation progresses. Rather than processing all historical context from scratch for each new question, the system continuously refines the historical vector representation, ensuring that useful contextual information is always available and up-to-date, thereby reducing processing time while maintaining accuracy.
3Measurement precision
If the machine incorporates historical statements into user intent identification, then the user intent accuracy is improved, but the computational complexity and processing load increase
Solution Approach 1:
The system applies the extraction principle by selectively extracting only the most relevant features and information from historical statements to create compact historical vectors. Instead of analyzing entire historical conversation transcripts, the system extracts key contextual elements and represents them in a condensed vector form, significantly reducing computational energy requirements while preserving the essential information needed for accurate intent identification.
Solution Approach 2:
The patent applies parameter changes by transforming historical statements into a different representation format (vectors) that are more computationally efficient to process. By changing the parameters of how historical context is stored and processed - from raw text to numerical vectors - the system reduces computational energy requirements while maintaining or improving intent identification accuracy.
Data Source
AI summary
The disclosure relates to identifying a user intent of a statement. When a user intent of a to-be-analyzed statement needs to be determined, at least one historical statement of the to-be-analyzed statement can be determined from an human-computer interaction session. A historical vector determined with reference to a statement vector of the to-be-analyzed statement and a statement vector of the at least one historical statement may comprise related content in the historical statement, and the content may be associated with the content of the to-be-analyzed statement. If the user intent of the to-be-analyzed statement is determined with reference to the historical vector, then the content related to the to-be-analyzed statement comprised in the historical vector may be used as an additional determination reference, so that the accuracy of determining the user intent of the to-be-analyzed statement can be improved.