User List Expansion via Feature Correlation Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for selecting and delivering targeted content, such as advertisements, on computer networks are inefficient in identifying and reaching users with specific interests, often failing to accurately expand user lists to include users with similar features, leading to irrelevant content exposure.
Innovation Solution
A computer-implemented method and system that analyzes user Internet activity data to select a subgroup of features, identify clusters of users, and determine correlations between supplemental users and existing user features, expanding user lists by adding users with similar interests, thereby enhancing the relevance of content delivery.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional user list selection methods are used, then the implementation is simple, but the accuracy of identifying users with specific interests deteriorates
Solution Approach 1:
The patent segments the user list expansion process into distinct functional modules: feature extraction module that analyzes user attributes, correlation analysis module that computes relationships between features, user clustering module that groups similar users, and list expansion module that generates the final expanded user list. This segmentation allows each module to specialize in a specific task, improving overall accuracy while making the complex system more manageable and maintainable
Solution Approach 2:
The patent introduces several intermediary components: feature vectors that mediate between raw user data and correlation analysis, similarity scores that mediate between user comparisons and clustering decisions, and weighting mechanisms that mediate between different feature importances. These intermediaries transform complex data relationships into manageable computational forms, enabling accurate user identification without requiring direct complex comparisons of all user attributes
2Quantity of substance
If user lists are expanded to include more users, then the reach of content delivery is improved, but the relevance of content to each user deteriorates
Solution Approach 1:
The patent dynamically adjusts multiple parameters during the expansion process: similarity thresholds that control how strictly users must match existing list users, feature weighting parameters that emphasize or de-emphasize specific user attributes based on content type, and correlation coefficients that determine the strength of relationship required for user inclusion. By optimizing these parameters, the system expands user lists while maintaining content relevance through quantitative control of inclusion criteria
Solution Approach 2:
The patent employs a multi-stage expansion approach where users are added to the user list in progressive stages rather than all at once. In each stage, users meeting a certain similarity threshold are added, then the threshold is adjusted for the next stage. This partial action approach allows the system to expand the user list progressively, evaluating relevance at each stage and preventing irrelevant users from being included, thus maintaining content relevance while increasing user quantity
3Measurement precision
If correlation analysis is performed between supplemental users and existing users, then the accuracy of user selection is improved, but the computational time deteriorates
Solution Approach 1:
The patent extracts only the most relevant features from complete user profiles for correlation analysis, rather than analyzing all user attributes. The feature extraction module identifies and extracts key discriminative features such as demographic characteristics, browsing behavior patterns, and purchase history indicators that are most predictive of user interests. By extracting only these critical features, the system maintains accurate user selection while dramatically reducing the computational burden of correlation calculations
Solution Approach 2:
The patent performs preliminary user clustering and feature extraction before conducting detailed correlation analysis. Users are first grouped into clusters based on broad similarity criteria, and only users within relevant clusters undergo detailed correlation analysis with existing list users. This preliminary action filters out obviously irrelevant users before expensive correlation computations, maintaining selection accuracy while reducing overall computational time through hierarchical processing
Data Source
AI summary
Systems and methods of selecting consumers to receive content on a computer network are provided. A user list identifying a first plurality of users having a group of features corresponding to internet activity of the first plurality of users can be obtained at a computing device. A subgroup of features can be selected from the group of features, and a cluster of users of the first plurality of users can be identified. The users of the cluster of users can each have at least one feature of the subgroup of features. A supplemental user having a supplemental feature can be identified. A correlation between the supplemental feature and at least one feature of the subgroup of features can be determined, and an expanded user list that includes at least one of the first plurality of users and the supplemental user can be generated.