Product Listing Recognition via DOM Node Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing technologies face challenges in accurately identifying and extracting product listings from diverse and non-standardized e-commerce websites, making it difficult to aggregate information and provide effective user experiences.
Innovation Solution
A method involving the extraction of a document object model (DOM) from e-commerce webpages to identify candidate nodes, which are then processed to determine if they qualify as product listings based on specific characteristics, using image recognition and machine learning algorithms to filter and categorize product listings.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional methods are used to identify product listings on diverse e-commerce websites, then the process requires manual intervention and rule-based approaches, but the accuracy and efficiency of identifying product listings across non-standardized platforms deteriorates
Solution Approach 1:
The patent replaces manual, rule-based identification methods with machine learning models that automatically learn product listing patterns from diverse e-commerce websites. The system uses trained classifiers and image recognition algorithms to substitute human judgment and complex rule-engineering with automated computational approaches that adapt to varying website structures.
Solution Approach 2:
The system changes the approach from fixed rule-based parameters to dynamic learned parameters through machine learning. The model adjusts its identification criteria based on training data from multiple e-commerce platforms, allowing it to adapt to different website structures, layouts, and product listing formats without requiring manual reconfiguration for each site.
2Productivity
If manual methods are used to extract and aggregate product information from multiple e-commerce platforms, then the process is simple to implement, but the productivity and scalability of data aggregation deteriorates
Solution Approach 1:
The system performs preliminary actions by pre-training machine learning models on extensive datasets from multiple e-commerce platforms before deployment. This advance training enables the system to quickly identify and extract product listings from new websites without requiring time-consuming manual configuration or analysis for each platform.
Solution Approach 2:
The patent replaces manual data extraction and aggregation processes with automated machine learning-based extraction systems. The trained models automatically crawl, parse, and extract product information from diverse e-commerce websites, eliminating the need for manual data collection and significantly improving extraction efficiency and scalability.
3Adaptability or versatility
If diverse e-commerce website structures are processed using fixed rules, then the system is easy to maintain, but the adaptability to different website formats and non-standardized layouts deteriorates
Solution Approach 1:
The system transforms fixed, static processing rules into dynamic, adaptive parameters through machine learning. The models learn from training data across multiple e-commerce platforms and automatically adjust their processing parameters to accommodate different website structures, layouts, and product listing formats, enabling high adaptability without increasing system complexity.
Solution Approach 2:
The patent implements a universal machine learning-based processing system that can handle multiple e-commerce platforms with diverse structures through a single adaptable framework. The trained models serve multiple functions across different websites, eliminating the need for separate rule-sets for each platform and reducing overall system complexity while maintaining high versatility.
Data Source
AI summary
In one embodiment, a method includes extracting a document object model (DOM) for a content page, wherein the DOM comprises a hierarchical tree-based data structure. The method also includes identifying candidate nodes in the DOM based on a context of the nodes, wherein the candidate nodes may correspond to listing items. The method additionally includes for each of the candidate nodes, locating its parent and child nodes by traversing the DOM from the candidate node, extracting information from the candidate node and its parent and child nodes, and assessing whether the candidate node qualifies as a listing item based on whether the extracted information fulfills a required set of characteristics for a listing item.