Category Recommendation for Product Database Classification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Price comparison engines face difficulties in integrating and managing product classifications from various online stores, which are often different and hierarchical, leading to challenges in maintaining a unified product database for effective price comparison and search functionality.
Innovation Solution
A method and system for category recommendation that extracts queries from foreign identifiers and recommends internal categories with high similarity, allowing for automatic and accurate classification of products by analyzing and matching internal identifiers and categories, thereby facilitating the creation and updating of product databases.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual classification methods are used to integrate product classifications from various online stores, then classification accuracy can be maintained, but the complexity and time required for database management increases significantly
Solution Approach 1:
The system enables automatic self-service classification by allowing the database to autonomously compute correlation values between queries and existing classifications, identify matching patterns, and assign classifications without manual intervention. This resolves the contradiction by eliminating the need for manual classification while maintaining accuracy through automated correlation computation.
Solution Approach 2:
The patent replaces manual mechanical classification processes with an automated computational system that uses query correlation algorithms. The system automatically computes correlation values between product queries and existing database entries, identifies patterns, and assigns classifications programmatically, substituting human effort with automated information processing.
2Productivity
If automated classification systems are implemented to reduce manual work, then processing speed improves, but classification accuracy may deteriorate due to inability to handle diverse classification systems
Solution Approach 1:
The system achieves universality by designing a classification approach that can handle multiple diverse classification systems from different online stores simultaneously. The query correlation method is store-agnostic and can adapt to various classification schemas, enabling the automated system to process heterogeneous data sources accurately while maintaining high processing speed.
Solution Approach 2:
The system dynamically adjusts classification parameters by computing correlation values based on the specific characteristics of each query and existing database entries. This parameter-based approach allows the automated system to adapt to different classification schemes and product types, maintaining accuracy across diverse scenarios while operating at automated speeds.
3Loss of information
If comprehensive product information is collected from all online stores, then database completeness improves, but the time and resources required for data collection and processing increase
Solution Approach 1:
The system performs preliminary actions by pre-computing and storing classification information in the product database during initial data collection. When new products are added, the system uses pre-established query correlation methods to quickly assign classifications without requiring extensive real-time analysis, thus reducing data collection and processing time while maintaining completeness.
Solution Approach 2:
The patent enables skipping time-consuming manual classification steps by implementing automated query correlation that rapidly processes product information. The system rushes through the classification assignment process by computing correlations algorithmically and assigning classifications in bulk, significantly reducing the time required to process comprehensive product data from multiple online stores.
Data Source
AI summary
A category recommendation method includes maintaining a category database including an internal identifier and an internal category to which the internal identifier belongs; externally receiving a foreign identifier and a foreign category where the foreign identifier belongs to from an outside; extracting a query associated with the foreign identifier; and recommending at least one of the internal categories stored in the category database by using the query as a category corresponding to the foreign identifier.