Regularized Sentiment Score Generation via Dynamic Text Block Resizing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional sentiment analysis engines provide inconsistent and unreliable sentiment scores, deviating significantly from manual interpretations, necessitating a more reliable method for monitoring sentiment progression in conversations.
Innovation Solution
The approach involves identifying a resizable text block within a conversation data object and performing successive regularized sentiment profile generation iterations, using an aggregation model and stage-wise penalty factor to determine a quantifiable regularized sentiment score, which is updated until it exceeds a threshold, ensuring accurate sentiment analysis.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If conventional sentiment analysis engines are used to analyze conversation transcripts, then sentiment scores can be generated automatically, but the scores vary significantly between different engines and deviate from manual interpretations
Solution Approach 1:
The patent combines multiple sentiment analysis engines into an ensemble system that aggregates their outputs. By merging the results from multiple engines and comparing them against manually interpreted sentiment scores from training data, the system achieves more accurate and consistent sentiment predictions than any single engine could provide alone.
Solution Approach 2:
The patent introduces an intermediary layer between automated sentiment engines and final sentiment scores. This intermediary consists of trained machine learning models that learn the mapping between engine outputs and manual interpretations, effectively mediating the discrepancy between automated and human sentiment assessment.
2Device complexity
If text blocks are analyzed in fixed sizes for sentiment analysis, then processing is simpler, but the analysis may include irrelevant context or miss important sentiment shifts
Solution Approach 1:
The patent implements dynamic text block resizing where the size of text blocks analyzed for sentiment is adjusted based on the detected sentiment progression. When sentiment shifts are detected, the system dynamically modifies block boundaries to capture relevant sentiment transitions, rather than using fixed-size blocks throughout the conversation.
Solution Approach 2:
The patent segments conversations into variable-sized text blocks based on sentiment boundaries rather than fixed time or word count intervals. This segmentation strategy divides the conversation at points where sentiment changes occur, ensuring each block contains coherent sentiment information without including irrelevant context from different sentiment states.
Data Source
AI summary
Various embodiments of the present disclosure performing conversation sentiment monitoring for a conversation data object. In various embodiments, a text block that can be resized is identified within a conversation data object and successive regularized sentiment profile generation iterations are performed until a regularized sentiment score of the block exceeds a regularized sentiment score threshold. A current regularized sentiment profile generation iteration involves determining a regularized sentiment score for the block based on an initial sentiment score, a subjectivity probability value, and, optionally, a stage-wise penalty factor. A determination is then made as to whether the score exceeds the threshold. If so, then a regularized sentiment profile of the conversation data object is updated based on the regularized sentiment score. If not, then the text block is resized and a subsequent regularized sentiment profile generation iteration is performed based on the resized block.