Intelligent Text Insight System for Sentiment-Based Comment Summarization

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Conventional comment-review systems often surface outlier comments or those based on narrow filters, hiding common comments and requiring labor-intensive human analysis, which can be inaccurate and inefficient.

Innovation Solution

An intelligent-text-insight system using machine-learning techniques and computational sentiment analysis to summarize and select representative textual responses, generating response summaries and representative comments that capture common sentiments and concepts.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If conventional comment-review systems use verification or voting to select textual comments, then popular or verified comments are surfaced, but outlier or unrepresentative comments are displayed instead of common comments

Engineering Contradiction:

Improveaccuracy of comment representationVSAvoidcommon comments hidden

Core Design Contradiction:

ReliabilityVSLoss of information

Solution Approach 1:

The patent changes the selection parameters from verification status or vote counts to semantic similarity and representativeness metrics. The system uses natural language processing to identify comments that are semantically similar to the majority, thereby surfacing common comments rather than outlier comments that simply have high engagement or verification status.

Inventive Principle:

Principle #35Parameter changes

Solution Approach 2:

The patent replaces the mechanical voting and verification system with an automated natural language processing system. Instead of relying on user votes or manual verification to determine comment quality, the system uses computational linguistics to automatically identify representative comments based on their semantic content and similarity to other comments in the dataset.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

2Ease of operation

If keyword-search functions are used to identify textual comments, then comments matching specific keywords are returned, but more common comments are buried or obscured

Engineering Contradiction:

Improvesearch functionalityVSAvoidcommon comments obscured

Core Design Contradiction:

Ease of operationVSLoss of information

Solution Approach 1:

Instead of searching for comments that match specific keywords entered by users, the system inverts the approach by automatically identifying and surfacing common comments based on their frequency and semantic similarity. The system proactively presents what is most relevant (common comments) rather than requiring users to search for them through keyword filters.

Inventive Principle:

Principle #13The other way round (Inversion)

Solution Approach 2:

The system performs self-service by automatically analyzing the comment dataset to identify representative comments without requiring user input for keyword searching. The natural language processing system autonomously determines which comments are most representative of the overall sentiment and content, eliminating the need for manual keyword-based filtering.

Inventive Principle:

Principle #25Self-service

3Measurement precision

If human reviewers analyze textual comments, then detailed analysis is provided, but the process is labor intensive, expensive, and produces inaccurate or inconsistent analyses

Engineering Contradiction:

Improveanalysis accuracyVSAvoidreview efficiency

Core Design Contradiction:

Measurement precisionVSProductivity

Solution Approach 1:

The patent replaces manual human review with an automated natural language processing system. The system uses computational linguistics, sentiment analysis, and clustering algorithms to automatically identify and categorize representative comments, eliminating the need for human reviewers while maintaining consistency and scalability.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The system creates a computational model that replicates the analytical function of human reviewers. By training machine learning algorithms on labeled data, the system learns to identify representative comments and sentiments in a manner that mimics human analysis but operates at machine speed and with consistent application of criteria.

Inventive Principle:

Principle #26Copying

Data Source

PatentUS11620455B2Intelligently summarizing and presenting textual responses with machine learning

Publication Date: 2023.04.04 QUALTRICS LLC

AI summary

This disclosure relates to methods, non-transitory computer readable media, and systems apply machine-learning techniques and computational sentiment analysis to summarize sentences extracted from a group of textual responses or to select representative-textual responses from the group of textual responses. By using a response-extraction-neural network to extract (and sometimes paraphrase) sentences from textual responses, the disclosed methods, non-transitory computer readable media, and systems can generate a response summary of textual responses based on sentiment indicators corresponding to the textual responses. By applying a machine-learning classifier to generate textual quality scores for textual responses, the disclosed methods, non-transitory computer readable media, and systems select representative-textual responses from a group of textual responses based on relevancy parameters and sentiment indicators corresponding to the textual responses. Such computational techniques generate response summaries and representative responses that provide an efficient snapshot of a group of textual responses analyzed by machine learners.