Digital Catalog Augmentation via Review Sentiment Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing solutions for sentiment analysis in e-commerce fail to effectively process unstructured customer reviews to identify new product features, sentiment towards these features, and consumer preferences, leading to biased views and inadequate product catalog updates.
Innovation Solution
A digital catalog augmentation system that includes a review scraper, text normalizer, and sentiment extractor to analyze customer reviews, identifying new criteria and sentiment, and a feedback module to update product descriptions, prioritize offerings, and generate reports based on negative sentiments.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If sentiment summarizers use pre-identified features from product descriptions, then the system can maintain structured data and perform sentiment analysis, but the system fails to accommodate new features and provides biased consumer views
Solution Approach 1:
The system performs preliminary text normalization and feature extraction from unstructured reviews before sentiment analysis. The review normalizer pre-processes raw text to identify and standardize feature mentions, enabling the sentiment analyzer to work with structured data while discovering new features automatically.
Solution Approach 2:
The patent introduces an intermediary feature extraction layer between raw reviews and sentiment analysis. This intermediary component identifies and structures feature mentions from unstructured text, transforming them into a format suitable for sentiment analysis while enabling discovery of new features not in the pre-identified set.
2Loss of information
If merchants manually read and analyze all customer reviews, then they can gain comprehensive insights, but it is beyond most merchants' capability in today's fast pace and large volume e-commerce environments
Solution Approach 1:
The system enables automated self-service processing of customer reviews. The review scraper automatically collects reviews from multiple sources, the normalizer automatically processes and structures the text, and the sentiment analyzer automatically generates insights, eliminating the need for manual merchant intervention while maintaining comprehensive analysis.
Solution Approach 2:
The patent replaces the mechanical manual process of reading and analyzing reviews with an automated computational system. The system uses text normalization techniques and sentiment analysis algorithms to process large volumes of reviews automatically, substituting human cognitive effort with machine-based processing.
3Manufacturing precision
If the system updates product catalogs with new features from reviews, then product descriptions become more accurate and comprehensive, but the complexity of processing and integrating unstructured data increases
Solution Approach 1:
The system segments the complex task of catalog augmentation into distinct modular components: review scraping, text normalization, feature extraction, and sentiment analysis. Each module handles a specific aspect of processing, making the overall system more manageable while achieving high accuracy in catalog updates.
Solution Approach 2:
The review normalizer applies parameter changes to transform unstructured review text into structured data. It normalizes text formats, standardizes feature representations, and transforms raw opinions into structured feature-sentiment pairs that can be directly integrated into product catalogs.
4Ease of operation
If merchants limit the number of reviews to a few, then analysis becomes manageable, but it results in a biased consumer view
Solution Approach 1:
The system replaces manual review selection with automated sentiment analysis. The computational system processes all available reviews objectively, using algorithms to identify genuine sentiment patterns rather than relying on subjective human selection, thereby maintaining measurement precision while managing operational complexity.
Solution Approach 2:
The sentiment analyzer acts as an intermediary between the large volume of reviews and the merchant. It processes all reviews systematically and generates summarized insights, serving as a bridge that preserves comprehensive consumer views while making the results manageable and actionable for merchants.
Data Source
AI summary
Disclosed are methods, circuits, devices, systems and functionally associated computer executable code for digital catalog augmentation. A digital catalog interface module reads from a digital catalog data storage, directly or indirectly, one or more catalog data records constituting an offer listing within a digital catalog, wherein the offer listing may include a description of a specific product or service offering and/or links to execute a transaction relating to the offering. The system includes a Review Criteria and Sentiment Extractor (RCSE) to identify and convert one or more reviews posted on a review forum into one or more data records used to augment the offer listing within the digital catalog.