Serverless Search Using Prefix-Based Index Database Sections
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
As the volume of indexed information grows, traditional index databases become too large to be efficiently stored in a single computer system, especially in mobile devices or small servers, necessitating a solution to reduce storage space while maintaining search functionality.
Innovation Solution
The index database is partitioned into smaller sections based on prefixes, with each section containing terms that share a common prefix, allowing for efficient loading and searching of specific sections, and utilizing serverless functions to manage these sections, reducing storage requirements and enabling efficient search operations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Volume of stationary object
If the index database is stored in a single computer system, then complete search functionality is maintained, but the storage space becomes too large for mobile devices or small servers
Solution Approach 1:
The index database is divided into multiple sections, each containing a subset of the total index terms. Each section can be independently loaded and searched, enabling the system to handle large volumes of data while maintaining reasonable storage requirements on individual devices.
Solution Approach 2:
The patent introduces a new dimension of distribution by partitioning the index database across multiple sections that can be stored and processed in different locations or time periods, transforming the single-dimension storage problem into a multi-dimensional solution space.
2Speed
If the entire index database is loaded into memory, then search speed is maximized, but the memory requirements become prohibitive for resource-constrained environments
Solution Approach 1:
The index database is segmented into multiple smaller sections that can be loaded into memory as needed. This allows the system to maintain high search speeds by keeping relevant sections in memory while avoiding the prohibitive memory requirements of loading the entire database.
Solution Approach 2:
The patent implements dynamic loading of index sections based on search queries. Sections are loaded into memory when needed and can be unloaded when not in use, creating a dynamic memory management system that adapts to varying search requirements while maintaining fast search performance.
Data Source
AI summary
One embodiment included a non-transitory machine-readable medium containing instructions that when executed carry out a method of searching for a search term. The method uses, instead of an index database of search terms, an index database section of only search terms that have the prefix of the search term, such that execution can occur on an improved processing system that is relatively small. The index database section is arranged as a prefix database of terms that start with the prefix, e.g., a trie, a radix trie, or a ternary search tree of the terms. The method may be implemented as a serverless function triggered by a user entering a search term or part thereof.