Textual Review Rating via Sentiment Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for rating mobile applications and other items rely heavily on numerical scores, which can be biased and do not fully capture user opinions, as textual reviews often provide valuable but unutilized information.
Innovation Solution
A computer-implemented method that collects and analyzes textual user reviews to calculate ratings for specific features and domains of an item, using a database to aggregate ratings for feature descriptors and words, providing a more comprehensive and accurate reputation score.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If numerical rating scores are used to evaluate items, then the rating process is simple and quick, but the rating accuracy and comprehensiveness deteriorate due to bias and inability to capture detailed user opinions
Solution Approach 1:
The patent segments the textual review into multiple sentences and further into individual words, analyzing each component separately to extract sentiment and meaning. This segmentation allows the system to process complex textual information in manageable units, improving rating accuracy without overwhelming computational complexity.
Solution Approach 2:
The patent introduces an intermediary database that stores pre-analyzed word ratings and sentiment information. This intermediary layer acts as a mediator between the raw textual review and the final rating calculation, enabling accurate sentiment analysis by referencing pre-computed word meanings and associations without requiring complex real-time analysis.
2Loss of information
If textual reviews are analyzed in detail to capture user opinions, then the comprehensiveness of the rating improves, but the processing time and computational resources increase
Solution Approach 1:
The patent performs preliminary analysis by pre-computing and storing word ratings, sentiment scores, and associations in a database before they are needed for actual review analysis. This preliminary action allows the system to quickly retrieve and utilize pre-analyzed information during the rating process, reducing processing time while maintaining comprehensive information utilization.
Solution Approach 2:
The patent creates a simplified representation of the textual review by extracting key words and sentences and copying their pre-analyzed ratings from the database. Instead of re-analyzing the entire text, the system copies relevant pre-computed information, significantly reducing processing time while preserving the essential information content.
3Reliability
If only numerical scores are collected from users, then the data collection process is straightforward, but the reliability of the rating deteriorates due to bias and lack of contextual information
Solution Approach 1:
The patent replaces the mechanical approach of directly using numerical scores with an information processing system that analyzes textual content. By substituting the simple numerical collection mechanism with text analysis processing, the system extracts reliable sentiment information from user reviews, improving rating reliability through contextual understanding while managing system complexity through automated natural language processing.
Data Source
AI summary
A computer-implemented method to determine a rating of an item is described. At least one user review with textual information relating to the item is collected. A rating for at least one sentence of the textual information is calculated. An overall rating for the item is determined based on the calculated rating for the at least one sentence of the textual information.