Balancing On-Site Engagement via Off-Site History Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Websites face challenges in broadening visitor engagement without over-exposing them to a narrow set of interests, as existing methods primarily focus on presenting supplemental content related to previously viewed on-site content, limiting discovery of new topics that may interest visitors.
Innovation Solution
A recommendation system that analyzes a recipient's on-site and off-site media consumption history to identify 'deficient' topics with an engagement gap, selecting and presenting supplemental content that balances on-site engagement, adapting to changing browsing behaviors and interests, using natural language processing to characterize content topics.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Duration of action of moving object
If supplemental content related to previously viewed on-site content is presented, then visitor engagement is extended, but the discovery of new topics is inhibited
Solution Approach 1:
The patent segments supplemental content into two distinct categories: (1) content related to previously viewed on-site topics, and (2) content related to newly discovered topics identified through off-site browsing behavior analysis. This segmentation allows the system to balance reinforcement of existing interests with introduction of new topics, resolving the contradiction between extending engagement duration and enabling topic discovery.
Solution Approach 2:
The patent introduces a new dimension by incorporating off-site browsing behavior data into the content recommendation system. Instead of relying solely on on-site viewing history (single dimension), the system analyzes off-site browsing patterns to identify emerging interests, adding a temporal and behavioral dimension that enables both engagement extension and novel topic discovery simultaneously.
2Ease of manufacture
If all visitors are presented with the same supplemental content based on aggregate analysis, then implementation is simplified, but customization to individual interests is limited
Solution Approach 1:
The patent implements preliminary action by pre-analyzing visitor browsing behavior patterns and pre-segmenting supplemental content into topic categories before presentation. The system pre-processes both on-site viewing history and off-site browsing data to identify interest profiles, then matches pre-categorized content to visitor profiles. This preliminary preparation enables customized content delivery without complex real-time decision-making, balancing implementation simplicity with customization capability.
3Adaptability or versatility
If supplemental content volume is increased to broaden engagement, then topic diversity improves, but over-exposure to narrow interests occurs
Solution Approach 1:
The patent applies local quality by dynamically adjusting the composition of supplemental content based on the specific visitor's interest profile. For each visitor, the system calculates the proportion of content from different topic categories, ensuring that content volume for each topic is optimized locally for that individual rather than applying a uniform distribution. This prevents over-exposure to any single topic while maintaining overall diversity, as the content mix is tailored to the visitor's demonstrated interests across multiple dimensions.
Data Source
AI summary
Encouraging broader engagement with a target publisher's content by balancing on-site topic engagement. Responsive to receiving notification of a recipient's request of an item of content from the target publisher, the on-site and off-site content consumption history of the recipient is analyzed. For each of a plurality of topics, historic engagement with on-site and off-site content is measured. Deficient topics having better off-site engagement than on-site engagement are detected. Content comprising links to items of content available from the target publisher which are characterized by deficient topics is selected and sent to the recipient. In an embodiment, the supplemental content comprises a plurality of links to other web pages on the target publisher's website which are sent to the recipient browser before the requested item of content finishes loading.