Referrer-Based Web Personalization via Preliminary Categorization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Web site personalization is challenging for first-time or new visitors with minimal profile information or behavior data, as existing methods rely heavily on explicit or implicit profiling which are not effective in this scenario.
Innovation Solution
A method that personalizes content of an electronic document by retrieving and analyzing a reference document, determining relevant categories, and customizing the requested document based on those categories, allowing for real-time personalization based on the originating page's content, such as selecting relevant advertising banners or highlighting content related to the originating page's categories.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If explicit profiling or implicit profiling methods are used for web site personalization, then personalization accuracy is improved for existing customers, but personalization effectiveness deteriorates for first-time or new visitors with minimal profile information
Solution Approach 1:
The system performs preliminary categorization of the referrer page content before the user arrives at the target page. By analyzing the referrer page's categories in advance and storing them, the system prepares personalization data beforehand, enabling effective personalization for new visitors without requiring prior profiling data.
Solution Approach 2:
The referrer page content serves as an intermediary source of information about user interests. Instead of relying on user-provided profile data, the system uses the referrer page's categorized content as a mediator to infer user interests and apply appropriate personalization to the target page.
2Measurement precision
If referrer page content is retrieved and analyzed in real-time, then personalization relevance is improved, but processing time increases
Solution Approach 1:
The system performs the time-consuming categorization of referrer page content in advance, before the user requests the target page. The categorized referrer data is stored and can be quickly retrieved during the actual page request, eliminating real-time processing delays while maintaining personalization relevance.
3Quantity of substance
If comprehensive visitor information is collected through explicit profiling, then profile completeness is improved, but visitor convenience deteriorates due to questionnaires and data entry requirements
Solution Approach 1:
The system automatically collects and processes referrer page information without requiring any user action. The user simply needs to provide the referrer page URL, and the system autonomously retrieves, categorizes, and applies the referrer content to personalize the target page, eliminating the need for questionnaires or manual data entry.
Data Source
AI summary
Personalization of content of a web site for a user based on a web site that a user arrives from is disclosed. For example, the content of the web site from which the user arrives (i.e., the originating page), as well as the content of the web page the user has arrived to (i.e., the target page), may be categorized as pertaining to particular subjects or topics. Any time a user comes from an originating page, the subject categories for the originating page and the target page may be compared to determine if like categories exist between the pages. In the event that like categories are found, the target page may be personalized based on those categories.