Conversational System Indirect Utterance Resolution via Knowledge Graph

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Developing artificial intelligence applications for natural language-based interaction with end users is challenging, particularly in establishing a logical connection between indirect user utterances and intended goals, as human-authored dialog flows are costly and often fail to anticipate user intentions.

Innovation Solution

A method that generates a parse structure for indirect utterances and connects it to a user goal through a knowledge graph, using a matching process to identify the intended request, allowing the system to respond appropriately.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If human-authored dialog flows are used to guide conversation, then the system can handle expected user inputs, but it fails to anticipate indirect user utterances and requires significant time and money to develop

Engineering Contradiction:

Improveability to handle user inputsVSAvoiddialog flow development complexity

Core Design Contradiction:

ReliabilityVSDevice complexity

Solution Approach 1:

The patent introduces an intermediary mechanism between the user utterance and the dialog flow: a knowledge graph that stores semantic relationships between entities and concepts. This knowledge graph acts as a mediator that can infer the intent behind indirect utterances by finding semantic connections between the uttered words and predefined dialog goals, eliminating the need for exhaustive hand-crafted dialog flows for every possible user input scenario

Inventive Principle:

Principle #24Intermediary (Mediator)

Solution Approach 2:

The patent replaces the mechanical, rule-based dialog flow matching system with a semantic-based system using natural language processing and knowledge graphs. Instead of relying on exact string matching or predetermined dialog trees, the system uses semantic similarity computation and graph-based reasoning to understand user intent, substituting rigid mechanical matching with flexible semantic understanding

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

2Measurement precision

If the system waits for explicit goal statements from users, then it can accurately understand user intent, but it cannot process indirect utterances that imply goals without stating them

Engineering Contradiction:

Improveuser intent understanding accuracyVSAvoidability to handle indirect utterances

Core Design Contradiction:

Measurement precisionVSAdaptability or versatility

Solution Approach 1:

The system performs preliminary action by pre-building a knowledge graph that contains semantic relationships, entities, and concepts relevant to the domain before user interaction begins. This pre-computed knowledge structure enables the system to quickly infer user intent from indirect utterances during conversation, rather than requiring explicit goal statements, as the semantic framework is already in place to capture implied meanings

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The patent changes the parameter of intent recognition from exact string matching to semantic similarity measurement. By computing similarity between the parsed utterance and knowledge graph entities/concepts, the system can recognize intents even when the user's wording doesn't exactly match predefined patterns, thereby handling indirect utterances while maintaining accuracy

Inventive Principle:

Principle #35Parameter changes

3Reliability

If hand-crafted dialog flows are used, then coverage of expected scenarios is achieved, but development time and cost increase significantly

Engineering Contradiction:

Improvedialogue coverageVSAvoiddialog flow development time

Core Design Contradiction:

ReliabilityVSLoss of time

Solution Approach 1:

The patent uses copying by leveraging existing knowledge graphs from similar domains or pre-built semantic frameworks that can be adapted to specific applications. Instead of creating dialog flows from scratch for each application, the system copies and adapts existing knowledge structures, significantly reducing development time while maintaining comprehensive scenario coverage through the reusable semantic framework

Inventive Principle:

Principle #26Copying

Data Source

PatentUS10832006B2Responding to an indirect utterance by a conversational system

Publication Date: 2020.11.10 INTERNATIONAL BUSINESS MACHINE CORPORATION

AI summary

A method, apparatus and computer program product for responding to an indirect utterance in a dialogue between a user and a conversational system is described. An indirect utterance is received. A parse structure of the indirect utterance is generated. The indirect utterance is an utterance which does not match a user goal expressed as elements of a knowledge graph. The parse structure is connected through the knowledge graph to a user goal to issue a user request which is not stated in the indirect utterance. The parse structure is connected using a matching process which matches the parse structure with the connected user goal in the knowledge graph according to a similarity of the parse structure and a portion of the knowledge graph including the connected user goal. A system response is performed based on the connected user goal.