Query-Independent Site Quality Scoring for Search Ranking
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Internet search engines face challenges in determining the quality of web sites based on user interaction, as existing methods lack a query-independent measure to effectively rank search results and decide on crawling, refreshing, or indexing resources.
Innovation Solution
A system computes a site quality score by analyzing the duration of user visits to resources, using statistical measures and resource classification to adjust scores, and discarding suspicious or low-value data, thereby providing a query-independent assessment of site quality.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If search engines use query-dependent measures to assess site quality, then they can provide more tailored search results, but they cannot establish a consistent, objective quality metric across different queries
Solution Approach 1:
The patent replaces query-dependent evaluation mechanisms with a query-independent site quality scoring system. Instead of measuring site quality through individual query results, the system uses aggregate user behavior metrics (visit duration, bounce rate, depth) across multiple queries to compute a consistent quality score for each site, thereby establishing objective and repeatable quality measurement.
2Reliability
If search engines rely on user interaction data to determine site quality, then they can capture real user preferences, but existing methods lack a standardized approach to process and interpret this data
Solution Approach 1:
The patent transforms raw user interaction data into standardized quality metrics by defining specific parameters for measurement (visit duration, bounce rate, navigation depth) and establishing thresholds and weightings. This parameter standardization converts complex, unstructured user behavior data into a systematic scoring framework that can be consistently applied across different sites and queries.
3Quantity of substance
If search engines crawl and index all resources, then they ensure comprehensive coverage, but they waste resources on low-quality sites
Solution Approach 1:
The patent implements preliminary site quality assessment before crawling and indexing resources. By calculating site quality scores in advance based on user interaction patterns, the search engine can pre-rank sites and allocate crawling resources proportionally, ensuring comprehensive coverage of high-quality sites while reducing unnecessary crawling of low-quality sites.
4Measurement precision
If search engines use site quality scores for ranking, then they improve result relevance, but they need to balance site-level metrics with query-specific factors
Solution Approach 1:
The patent segments the search ranking function into two independent components: site quality scoring (query-independent) and query relevance assessment (query-dependent). The site quality score provides a baseline relevance indicator, while query-specific factors such as keyword matching, user intent, and contextual information are evaluated separately and combined with the site score to produce final rankings, allowing both comprehensive evaluation and query optimization.
Data Source
AI summary
Methods, systems, and apparatus, including computer programs encoded on computer storage media, for determining site quality scores. In one aspect, a method includes obtaining a plurality of measurements of durations of user visits to resources included in a particular site; and determining a site quality score for the particular site based at least in part on the plurality of measurements, wherein determining the site quality score for the particular site comprises computing a statistical measure from the plurality of measurements.