Product Ranking via Comparison Sentiment Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing product search systems struggle to differentiate products with similar rating scores, as the typical 5-star rating scale does not accurately reflect user preferences, and user ratings can be subjective and influenced by different criteria.
Innovation Solution
A method that combines numerical user ratings with opinion scores derived from written comparison sentiments in product reviews, using a supervised learning approach to detect user preferences and adjust ratings, resulting in a final score for ranking products.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If product ranking is based on average numerical rating scores, then the ranking process is simple and fast, but it becomes difficult to differentiate products with similar ratings
Solution Approach 1:
The patent segments the product ranking problem into two independent components: numerical rating scores and textual comparison sentiments. By dividing the ranking signal into these separate segments, the system can process numerical ratings quickly while simultaneously analyzing textual comparisons for finer differentiation, thus resolving the contradiction between ranking speed and differentiation accuracy.
Solution Approach 2:
The patent adds a new dimension to the ranking system by incorporating textual comparison sentiments as a separate feature dimension alongside numerical ratings. This dimensional expansion allows the system to differentiate products not just by score magnitude but also by comparative language patterns, enabling accurate differentiation without sacrificing the computational efficiency of numerical scoring.
2Device complexity
If product ranking relies solely on numerical user ratings, then the system is easy to implement, but user preferences expressed in written reviews are not captured
Solution Approach 1:
The patent merges two distinct data sources - numerical ratings and textual reviews - into a unified ranking framework. By combining these previously separate information channels, the system captures both the quantitative simplicity of numerical scores and the qualitative richness of written user preferences, eliminating information loss while maintaining implementation feasibility through modular architecture.
Solution Approach 2:
The patent creates a multi-functional ranking system that simultaneously processes numerical data and textual data through separate but complementary analysis pipelines. This universal framework can handle both simple numerical comparisons and complex textual sentiment analysis, ensuring no user preference information is lost while keeping the overall system implementation manageable through functional modularity.
3Ease of operation
If the 5 star rating scale is used, then user feedback collection is straightforward, but the scale lacks the granularity to differentiate comparable products
Solution Approach 1:
The patent segments the feedback collection into two parts: the simple 5-star numerical rating system for ease of user input, and a separate textual comparison field where users can express nuanced preferences. This segmentation maintains the operational simplicity of star ratings while adding the precision needed to differentiate comparable products through written comparisons.
Solution Approach 2:
The patent adds a textual comparison dimension to complement the 5-star rating scale. While the star scale provides easy-to-collect quantitative feedback, the textual dimension captures nuanced user preferences and detailed product comparisons, thereby enhancing differentiation precision without compromising the ease of feedback collection through the maintained star rating interface.
Data Source
AI summary
A method of ranking a plurality of products includes obtaining a numerical user score for each of the plurality of products, calculating an opinion score for each of the plurality of products for which a written comparison sentiment applies, determining a final score by combining the numerical user rating and the opinion score for each of the plurality of products, and ranking the plurality of products based on the final score. The opinion score is derived from one or more written comparison sentiments containing a first product name and a second product name. In another aspect, a computer readable medium has instructions stored thereon which, when executed, cause a computer to perform a method of ranking a plurality of products including at least a first product and a second product.