Front End Server Fallback for User Identification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for understanding user interests and behavior, such as user accounts and cookies, are limited by users accessing multiple sites that do not share information, declining cookies, or browsing in incognito mode, leading to missed opportunities for personalized product recommendations.
Innovation Solution
A network environment system that associates website visits with user identifiers through browser data analysis, including cookie data, device information, and user interactions, to create a unique user identifier record and browser user identifier vector, allowing for probabilistic matching and updating of user records across devices, even without explicit identification.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If user accounts and cookies are used to identify users, then personalized recommendations can be provided, but users may decline cookies, clear cookies, or browse in incognito mode causing identification to fail
Solution Approach 1:
The patent introduces device fingerprinting as an intermediary method to identify users when traditional cookies fail. By analyzing device characteristics (hardware identifiers, software configuration, browser settings), the system creates a unique device signature that serves as a mediator between the user and the recommendation system, enabling identification without relying on cookies that users can decline or clear.
Solution Approach 2:
The system changes the identification parameters from cookie-based data to device hardware and software parameters. Instead of relying on mutable cookie data that users can control, the system uses relatively stable device parameters (screen resolution, processor information, installed fonts, time zone) to create persistent device fingerprints that remain consistent across browsing sessions.
2Loss of information
If cookies are used for user identification, then browsing activity can be tracked across sessions, but users may access sites that do not share account information leading to missed opportunities
Solution Approach 1:
The patent implements a universal identification system that works across multiple websites and devices. The device fingerprinting technology creates a cross-site identifier that can recognize users regardless of which website they are visiting or what browser they are using, enabling the recommendation system to aggregate browsing behavior data across the entire web rather than being limited to single-site cookie tracking.
Solution Approach 2:
The system transitions from two-dimensional cookie-based identification (single site, single browser) to multi-dimensional device fingerprinting that captures hardware identifiers, software configuration, network information, and temporal patterns across multiple devices and browsers, creating a richer identification space that follows users across the digital ecosystem.
3Adaptability or versatility
If probabilistic matching is used to associate visits with user identifiers, then user records can be linked across devices, but false matches may occur reducing accuracy
Solution Approach 1:
The patent implements feedback mechanisms where the system continuously refines its matching algorithms based on confirmed user behaviors and corrections. When users interact with recommendations or provide explicit feedback, the system uses this information to adjust and improve the probabilistic matching accuracy, reducing false positives over time while maintaining cross-device tracking capabilities.
Data Source
AI summary
Browser requests are received and data included in it is added to a vector. If explicit identification information (username, cookie data, etc.) is present, the vector is associated with a pre-existing user record, which is then updated. If not, candidate user records may be identified according to correspondence with values in the vector. Candidate vectors may be eliminated by identifying inconsistency in OS, device, and browser information. Probability assigned to each candidate vector may be adjusted, e.g., reduced, in response to inconsistency in other data relating to a browser. Personalized results are generated based on collected data for a user if possible. If not possible, fallback results are provided by a front end server based on values provided in a browser request matching entries in a fallback database accessed by the front end server. Default results may be provided where the fallback database does not provide a result.