Micro-blog Sorting by User Relevance Score

The patent segments the micro-blog stream into different categories or groups based on user relationships (e.g., friends, followers, topics of interest) and sorts each segment differently. This allows relevant content from specific groups to be prioritized while maintaining overall chronological order, making it easier for users to find interesting content without manually searching through the entire feed.

The patent applies different sorting qualities to different parts of the micro-blog stream. Instead of uniform chronological sorting, it enhances certain sections by boosting micro-blogs from users with higher interaction scores or from topics the user is interested in, making those portions more prominent and easier to discover.

The patent performs preliminary actions by pre-calculating and storing user relationship scores, topic interests, and interaction histories before the sorting process. This allows the sorting algorithm to quickly retrieve and apply pre-computed relevance metrics, improving efficiency without requiring complex real-time calculations when generating the micro-blog feed.

The patent introduces intermediary elements such as topic tags, user profiles, and interaction graphs that mediate between the raw micro-blog data and the final sorted output. These intermediaries enable efficient filtering and ranking by providing structured information about content relevance without requiring the sorting system to directly analyze every micro-blog against every user preference.