Conversational Agent Deployment via Intent-Entity Matching

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Users face challenges in finding and accessing relevant information from multiple electronic sources due to the need for diverse skills, including knowledge of question types, information sources, and programming expertise, making it inefficient for developers to create execution plans for conversational agents.

Innovation Solution

A method for deploying a computerized conversational agent that simplifies the configuration and deployment process by receiving configuration information, determining actual entities from possible entities based on user requests, and providing fulfillment, using a graphical user interface and similarity scoring algorithms to match user inputs with available data sources.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If developers manually create execution plans with multiple skills (question types, information sources, access methods, formatting, programming), then the conversational agent can answer questions accurately, but the deployment process becomes complex and time-consuming

Engineering Contradiction:

Improveanswer accuracyVSAvoiddeployment complexity

Core Design Contradiction:

ReliabilityVSDevice complexity

Solution Approach 1:

The patent introduces an intermediary system that automatically generates execution plans by matching user questions with pre-defined question templates and information sources. This intermediary layer translates natural language questions into structured queries without requiring manual programming, thereby maintaining answer accuracy while eliminating deployment complexity.

Inventive Principle:

Principle #24Intermediary (Mediator)

Solution Approach 2:

The system enables self-service by allowing the conversational agent to automatically determine which information sources to query and how to format responses based on the detected question type. The agent autonomously selects appropriate execution plans from a library of templates, eliminating the need for developer intervention in the deployment process.

Inventive Principle:

Principle #25Self-service

2Adaptability or versatility

If developers spend time creating custom execution plans for each question type, then the conversational agent can be highly customized, but developer productivity decreases

Engineering Contradiction:

Improveconversational agent customizationVSAvoiddeveloper productivity

Core Design Contradiction:

Adaptability or versatilityVSProductivity

Solution Approach 1:

The patent implements preliminary action by pre-defining execution plans for common question types during system setup. These templates include pre-configured information sources, query structures, and response formats. When deployed, the system automatically matches incoming questions to these pre-prepared templates, providing high customization without requiring developers to create custom plans for each scenario.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system achieves universality by creating a single executable file that contains a library of question templates and information source configurations. This universal deployment package can handle multiple question types and domains without requiring separate customization for each, thereby maintaining adaptability while dramatically improving developer productivity.

Inventive Principle:

Principle #6Universality (Multi-functionality)

3Loss of information

If the system provides detailed information from multiple electronic sources, then answer completeness improves, but information retrieval becomes cumbersome

Engineering Contradiction:

Improveinformation completenessVSAvoidinformation access ease

Core Design Contradiction:

Loss of informationVSEase of operation

Solution Approach 1:

The patent merges information from multiple electronic sources by automatically querying relevant databases and APIs based on the detected question type. The system consolidates results from different sources into a unified response structure, providing complete information while maintaining ease of access through a single conversational interface.

Inventive Principle:

Principle #5Merging (Combining)

Solution Approach 2:

The system segments the information retrieval process into distinct stages: question analysis, template matching, information source selection, query execution, and response synthesis. This segmentation allows the system to efficiently manage multiple information sources while presenting a simplified interface to users, maintaining both completeness and ease of operation.

Inventive Principle:

Principle #1Segmentation

Data Source

PatentUS11423232B2Systems and methods for deploying computerized conversational agents

Publication Date: 2022.08.23 NFL ENTERPRISES LLC

AI summary

Embodiments of the disclosed technology relate to systems, methods, and computer-readable storage media for deploying a computerized conversational agent. Some embodiments can receive configuration information comprising an intent and a plurality of possible entities corresponding to the intent. Some embodiments can receive a user request comprising a user entity and a user intent. Some embodiments can determine an actual entity from the plurality of possible entities corresponding to the user intent where the actual entity corresponds to the user entity. Some embodiments can provide to the user a fulfillment corresponding to at least the actual entity.