Relay Apparatus Keyword Search Cache Optimization

The patent segments the content retrieval process into two phases: first utilizing local cache with keyword matching for immediate results, then supplementing with server requests only for missing content. This segmentation allows the system to leverage cached keyword-based content without requiring full content name matches, thereby improving bandwidth utilization while maintaining search convenience.

The patent implements preliminary caching of content data with associated keywords before retrieval requests occur. When a keyword search is performed, the system first checks the cache for matching keywords and returns cached content directly if found, avoiding unnecessary server transmissions. This preliminary preparation resolves the contradiction by enabling efficient keyword-based retrieval that leverages existing cache resources.

The patent changes the matching parameter from exact content name comparison to keyword-based partial matching. Instead of requiring the entire content name to match for a cache hit, the system extracts keywords from the requested content name and matches them against cached content keywords. This parameter change increases cache utilization flexibility while maintaining transmission efficiency by still returning cached data when matches are found.

The patent makes the cache system multi-functional by enabling it to serve both exact content name requests and keyword-based searches. The same cache database structure is used for both precise content retrieval and flexible keyword matching, allowing the cache to adapt to different retrieval scenarios without requiring separate systems. This universality resolves the contradiction by making the cache more versatile while maintaining its efficiency benefits.

The patent applies partial action by requesting only the specific content data that is not already cached, rather than retrieving all content data from the server. When a keyword search is performed, the system identifies which cached content matches the keywords and only requests the remaining missing content from the server. This partial retrieval approach ensures data completeness for the requested content while significantly reducing bandwidth consumption compared to fetching all content.

The patent implements a feedback mechanism where the system first checks the cache status before making server requests. The cache lookup results provide feedback that determines whether server communication is necessary and what specific data should be requested. This feedback loop ensures that complete content is delivered to the user while avoiding redundant bandwidth consumption by leveraging available cached data whenever possible.