Automated Blacklisting via Image Feature Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
As the size of catalogs grows, it becomes increasingly difficult for advertisers to detect and identify items that should be blacklisted, leading to inadvertent display of prohibited items in online advertisements and webpages.
Innovation Solution
A computer-implemented method using image feature analysis and collaborative filtering to recommend catalog items for blacklisting by extracting features from digital images and associating them with user purchase and view events, as well as textual descriptions, to identify similarities and correspondences with already-blacklisted items.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual detection and identification of blacklisted items is used, then advertisers can identify prohibited items, but the process becomes increasingly difficult as catalog size grows
Solution Approach 1:
The patent replaces manual mechanical detection with automated image recognition technology. The system uses machine learning models to automatically analyze product images, extract features, and identify blacklisted items without human intervention. This substitution resolves the contradiction by maintaining high detection accuracy while eliminating the increasing complexity associated with manual catalog management as catalog sizes grow.
Solution Approach 2:
The system enables self-service blacklisting by automatically comparing new product images against a database of blacklisted item features. The automated comparison and identification process allows the system to serve itself in detecting prohibited items without requiring advertiser intervention, thereby maintaining detection accuracy while reducing management complexity.
2Productivity
If automated image analysis is implemented, then identification of blacklisted items is facilitated, but system complexity increases
Solution Approach 1:
The patent segments the blacklisting system into distinct functional modules: image feature extraction, feature database maintenance, similarity comparison, and blacklisting decision-making. Each module performs a specific task independently, which improves overall productivity while managing system complexity through modular design. This segmentation allows the system to efficiently process large catalogs without becoming unmanageably complex.
Solution Approach 2:
The system implements a universal image analysis framework that can handle multiple catalog items simultaneously using the same feature extraction and comparison mechanisms. This multi-functional approach improves productivity by processing entire catalogs efficiently while avoiding the complexity increase that would result from creating separate specialized systems for each item.
3Reliability
If comprehensive feature analysis is performed on all catalog items, then accurate blacklisting is achieved, but processing time increases
Solution Approach 1:
The patent performs preliminary action by pre-extracting and storing image features for all catalog items in advance, creating a feature database before the actual blacklisting process. This preliminary feature extraction allows the system to maintain high blacklisting accuracy through comprehensive analysis while reducing processing time during actual blacklisting operations, as the heavy computational work has already been completed beforehand.
Data Source
AI summary
Described are methods, systems, and apparatus for recommending catalog items for blacklisting. For each of a plurality of catalog items: image features are extracted; image features are associated with the catalog item; and user purchase events, user view events, a textual description, and categories are associated with the catalog item by the recommendation system. An identification of a first blacklisted catalog item is received. A catalog item is identified by the recommendation system based on i) a similarity between the image features associated with the first blacklisted catalog item and the image features associated with the catalog item, and ii) a correspondence between at least one of the user purchase events, the user view events, the textual description, or the categories associated with the blacklisted catalog item and the user purchase events, the user view events, the textual description, or the categories associated with the catalog item.