Network Resource Ranking via Cross-Resource Context
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for customizing network resource content and style for users are limited by their reliance on session data and interactions, making it difficult to determine the most relevant presentation for users on their first visit and failing to utilize context information from distinct network resources.
Innovation Solution
A method that utilizes context information from past interactions with other network resources to rank elements of a first network resource, involving a second server that retrieves and analyzes this information to determine the most relevant content or style for a user, even if no direct interaction data is available, using machine-learning techniques to identify hidden correlations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If content providers use generic presentation styles to appeal to broad audiences, then user satisfaction is maintained across diverse groups, but user engagement and conversion rates are limited
Solution Approach 1:
The system performs preliminary actions by collecting context information from multiple distinct network resources before the user visits the target resource. This advance data gathering enables the content customization system to prepare personalized content presentations in advance, rather than relying on generic styles or waiting for session data accumulation.
Solution Approach 2:
The invention transitions from traditional single-resource session data to multi-resource context information, adding a new dimension to user profiling. By incorporating data from distinct network resources beyond the immediate web page, the system creates a more comprehensive user profile that enables better content adaptation and higher conversion rates.
2Measurement precision
If content providers rely on session data from the current web page, then they can customize content based on user interactions, but they cannot determine relevant presentation for users on their first visit
Solution Approach 1:
The system performs preliminary data collection from multiple distinct network resources before the user visits the target resource. This advance preparation eliminates the need to wait for session data accumulation on the current page, enabling accurate user preference assessment even on first visits.
Solution Approach 2:
The system uses context information from distinct network resources as intermediary data sources to infer user preferences. These external resources serve as mediators that provide indirect but valuable information about user behavior patterns, enabling preference accuracy without direct interaction with the target resource.
3Loss of information
If content providers use traditional web analytics focusing on current page interactions, then they can track user behavior, but they miss hidden correlations from other network resources
Solution Approach 1:
The system applies multi-functionality by using a unified approach to collect and analyze context information from multiple distinct network resources. The same data collection and analysis mechanisms are leveraged across different resources, enabling comprehensive user profiling without proportionally increasing system complexity.
Solution Approach 2:
The system introduces context information from distinct network resources as intermediary data sources. These intermediaries provide valuable user behavior insights that bridge the gap between isolated web page analytics and comprehensive user understanding, revealing hidden correlations without requiring direct integration of every possible data source.
Data Source
AI summary
Method of and system for ranking elements of a first network resource for a first user, first network resource being hosted by a first server, method comprising, at a second server in communication with first server via a communications network: receiving an indication of elements from first server; receiving an indication of first user; based on at least one received indication, retrieving context information from a first database in communication with second server, context information being at least partially indicative of a relative relevance of elements to first user, context information including information about a past interaction of at least one of first user and a second user with a second network resource, second network resource being distinct from first network resource; and based at least in part on context information, determining at least one of a ranking of elements by relevance to first user and a most relevant element.