Conversational Context Retrieval Using Prior Dialog Topic Routing

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Conventional methods for continuing a previous dialog with a virtual assistant require manual search through numerous prior conversations, consuming computational resources and battery power, especially on devices with limited display sizes.

Innovation Solution

Utilizing a routing agent and generative models to identify and select a conversational context related to a user query, processing user input and topic dictionaries to generate a response based on previous dialogs, reducing the need for manual search and resource consumption.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Ease of operation

If manual search through chat history is used to identify previous dialogs, then user can find relevant conversations, but computational resources and battery power are consumed unnecessarily

Engineering Contradiction:

Improveease of dialog retrievalVSAvoidbattery consumption

Core Design Contradiction:

Ease of operationVSUse of energy by moving object

Solution Approach 1:

The system performs preliminary actions by generating summaries of previous dialogs and storing them in a topic dictionary before the user needs to search. This allows the user to quickly retrieve relevant conversations through semantic matching without manually scrolling through entire chat histories, thereby reducing battery consumption while maintaining ease of dialog retrieval.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The patent introduces an intermediary mechanism (topic dictionary with summaries) that mediates between the user's search query and the actual chat history. Instead of directly searching through numerous previous dialogs, the system uses the topic dictionary as an intermediary layer to filter and present relevant conversations, reducing computational resource usage.

Inventive Principle:

Principle #24Intermediary (Mediator)

2Loss of information

If user manually reviews chat history to continue previous dialog, then relevant conversation can be found, but time and user inputs are increased

Engineering Contradiction:

Improveretrieval accuracyVSAvoidtime to find dialog

Core Design Contradiction:

Loss of informationVSLoss of time

Solution Approach 1:

The system performs preliminary summarization of dialog topics and stores them in advance. When a user wants to continue a previous dialog, the system can quickly match the user's intent against these pre-generated summaries, significantly reducing the time required to find the relevant conversation while maintaining accurate retrieval through semantic matching.

Inventive Principle:

Principle #10Preliminary action

3Area of stationary object

If chat history is displayed on client device with limited display size, then fewer prior dialogs can be viewed, but user requires more inputs to navigate

Engineering Contradiction:

Improvedisplay areaVSAvoidease of navigation

Core Design Contradiction:

Area of stationary objectVSEase of operation

Solution Approach 1:

The patent introduces a topic dictionary as an intermediary that presents condensed representations of previous dialogs rather than requiring users to navigate through entire chat histories. This intermediary layer allows users to quickly identify and access relevant conversations even on devices with limited display areas, reducing the number of navigation inputs required.

Inventive Principle:

Principle #24Intermediary (Mediator)

Data Source

PatentUS20260010569A1Generating response(s) to user input(s) for new conversation(s) by selecting and prepending conversational context(s) from prior conversation(s)

Publication Date: 2026.01.08 GOOGLE LLC

AI summary

Implementations relate to utilizing machine learning model(s) in selecting a prior conversational context related to a user query in response to receiving the user query. The prior conversation context can be based on a particular prior dialog that is selected from all prior dialogs and based on the user query. The particular prior dialog can be selected based on processing at least the user query and a list of topics, respectively determined from the prior dialogs, using the machine learning model(s). For example, an output of the machine learning model(s) can indicate a particular topic from the list of topics that is related to the user query, and the particular topic can be utilized to identify the particular prior dialog from which the particular topic is determined, thereby enabling the particular prior dialog (or a representation thereof) to be utilized as context in responding to the user query.