Inter-site Browsing Attribute Clustering for User Intent
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Modern product recommendation and digital advertising methods primarily focus on intra-site browsing behaviors, neglecting inter-site browsing attributes, which limits their ability to accurately understand user intentions and preferences.
Innovation Solution
A method that clusters websites based on user browsing history, calculates similarity between website groups, and classifies browsing modes into sojourner, resident, inter-wanderer, or intra-wanderer types to determine user browsing preferences, enhancing product recommendation and digital advertising effectiveness.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If modern product recommendation methods focus only on intra-site browsing history, then the analysis process is simple, but the accuracy of understanding user intentions is limited
Solution Approach 1:
The patent segments the analysis process into multiple components: extracting browsing sequences from web browsing history, clustering websites into groups based on similarity, identifying browsing attributes (sojourner, resident, inter-wanderer, intra-wanderer types), and combining these with product information. This segmentation allows comprehensive inter-site browsing analysis while maintaining manageable complexity through modular processing steps.
2Measurement precision
If browsing attributes are extracted from inter-site browsing history, then user preference accuracy improves, but computational complexity increases
Solution Approach 1:
The patent creates simplified representations (copies) of complex browsing behaviors through clustering. Instead of analyzing every individual website visit, the system clusters websites into groups and represents browsing patterns as sequences of cluster visits. This copying approach maintains the essential characteristics of user behavior while significantly reducing computational complexity.
Solution Approach 2:
The patent transforms raw browsing data into structured parameters including browsing sequences, cluster identifiers, and attribute classifications (sojourner, resident, inter-wanderer, intra-wanderer types). By changing the parameter representation from raw URLs to structured behavioral attributes, the system enables more accurate user preference analysis while facilitating efficient computation through standardized data formats.
3Measurement precision
If comprehensive browsing history analysis is performed, then product recommendation accuracy increases, but processing time increases
Solution Approach 1:
The patent performs preliminary clustering of websites into groups before the actual recommendation process. By pre-processing the browsing history data and organizing websites into clusters, the system prepares structured information that can be quickly queried during product recommendations. This preliminary action reduces processing time during actual recommendation operations while maintaining comprehensive analysis accuracy.
Data Source
AI summary
Disclosed is a method of producing browsing attributes of a user, and the method includes: searching for a web site group in a web browsing history; acquiring a tag of the website group according to a percentage of a web category of the website group; obtaining a browsing preference attribute of the user by calculating a similarity; and obtaining a present browsing mode attribute of the user by analyzing a purity by the web browsing history of the user.