Response Determination via Correlation-Based Paragraph Selection

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing artificial intelligence systems for question answering face limitations in providing accurate responses, particularly for subjective questions and require extensive preliminary work in creating templates and processing diverse queries, leading to inefficiencies and incomplete answers.

Innovation Solution

A method and apparatus that determine responses for user input data by obtaining search results, calculating correlation degrees, building candidate content, and selecting relevant paragraphs from search results to provide accurate and concise answers, leveraging internet search and machine learning algorithms for improved efficiency and accuracy.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If traditional AI question answering systems use template matching and semantic analysis methods, then they can provide structured responses, but they require extensive preliminary work in creating templates and processing diverse queries, leading to inefficiency and incomplete answers

Engineering Contradiction:
Improveprocessing efficiencyVSAvoidtemplate creation complexity
Core Design Contradiction:
ProductivityVSDevice complexity

Solution Approach 1:

The patent extracts the core relevant information from search results by calculating correlation degrees between search result content and user input data. Instead of relying on pre-defined templates, the system extracts and selects paragraphs that have the highest correlation with the user's question, thereby eliminating the need for extensive template creation while maintaining response quality

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

The system performs self-service by automatically obtaining search results, calculating correlation degrees, and selecting candidate content without requiring manual template creation or extensive preliminary processing. The algorithm autonomously identifies and selects the most relevant paragraphs from search results to form the response

Inventive Principle:
Principle #25Self-service

2Reliability

If AI systems process all search results to ensure comprehensive answers, then answer completeness improves, but the number of webpages needed increases, reducing processing efficiency and user experience

Engineering Contradiction:
Improveanswer accuracyVSAvoidprocessing time
Core Design Contradiction:
ReliabilityVSLoss of time

Solution Approach 1:

The patent applies local quality by calculating correlation degrees for different paragraphs within search results and selectively processing only those with high correlation values. This allows the system to focus computational resources on the most relevant content rather than uniformly processing all search results, thereby maintaining answer accuracy while reducing processing time

Inventive Principle:
Principle #3Local quality

Solution Approach 2:

The system performs partial action by selecting only the top K paragraphs with the highest correlation degrees to form candidate content. Instead of processing all search results exhaustively, it takes a partial approach by focusing on the most relevant portions, which efficiently balances comprehensiveness with processing speed

Inventive Principle:
Principle #16Partial or excessive action

Data Source

PatentUS11238050B2Method and apparatus for determining response for user input data, and medium
Publication Date: 2022.02.01 BAIDU ONLINE NETWORK TECH (BEIJIBG) CO LTD
  • US11238050B2 patent drawing
  • US11238050B2 patent drawing
  • US11238050B2 patent drawing

AI summary

According to exemplary embodiments of the present disclosure, a method and an apparatus for determining a response for user input data, a device, and a computer-readable storage medium are provided. The method includes: obtaining search results associated with the user input data. The method also includes: determining correlation degrees between the search results and the user input data. The method also includes: building candidate content for the user input data based on the search results and the correlation degrees. The candidate content includes a paragraph associated with the user input data in the search results. The method also includes: determining the response for the user input data from the candidate content.