Уровень 0 · материалов: 8
В кластер входят документы, описывающие технические методы, инструменты и архитектурные решения для организации и ускорения полнотекстового поиска в базах данных или специализированных системах.
Общие признаки: реализация полнотекстового поиска, повышение производительности поиска, индексация данных, инструменты полнотекстового поиска
Группа выше: Полнотекстовый поиск
Смысл: The main idea is to introduce the Sphinx search engine as a viable, open-source alternative for implementing high-performance full-text search and attribute filtering within private corporate databases, specifically using MySQL.
A technical guide explaining how to install and configure the Sphinx open-source search engine to perform efficient full-text and attribute-based searches on a MySQL database.
Смысл: The text is a technical tutorial explaining how to set up and integrate the Sphinx search engine with a PHP and MySQL environment to replace basic full-text search.
A step-by-step technical guide on installing, configuring, and integrating the Sphinx search engine with PHP and MySQL for improved full-text search capabilities.
Смысл: The main idea is to demonstrate how a custom-built, bit-oriented indexing and execution system can outperform general-purpose search engines like Sphinx for specific high-load dating site queries.
Mamba replaced general search tools with a custom bit-indexed 'demon-searcher' to reduce search times for millions of profiles from 100ms to under 1ms.
Смысл: The text explains the mechanics, implementation, and advanced use cases of full-text search in MySQL, highlighting its superiority over basic pattern matching for large datasets.
A technical guide on using MySQL's FULLTEXT indexes and MATCH...AGAINST construction to implement efficient search and recommendation systems.
Смысл: The text explains how to create a functional, self-hosted full-text search system for websites using PHP and the phpMorphy library, emphasizing the importance of lemmatization and indexing for efficiency and accuracy.
A technical tutorial on developing a custom PHP-based search engine featuring morphological analysis, ranking, and database-driven indexing.
Смысл: The text explains the architectural decisions and technical implementation of a high-performance, resource-efficient full-text search engine tailored specifically for email mailboxes at Mail.Ru.
Mail.Ru developed a custom search daemon using a snapshot-plus-transaction-log architecture to provide fast, low-resource email searching.
Смысл: The main idea is that while PostgreSQL provides native full-text search capabilities, a naive implementation is prohibitively slow; performance can be drastically improved by storing pre-computed `tsvector` data in a separate table and using GIN indexes, though ranking results remains a computational challenge that may require specialized extensions like RUM.
The author details the transition from Sphinx to PostgreSQL for full-text search, demonstrating how moving from naive queries to pre-computed `tsvector` tables with GIN indexes reduces query time from minutes to milliseconds.
Смысл: The main idea is to explain the internal mechanism, advantages, and use cases of the GIN index in PostgreSQL, highlighting its efficiency for full-text search, arrays, and JSONB, while noting its update-performance costs.
A detailed technical guide on PostgreSQL GIN indexes, explaining how inverted indexing accelerates searches for complex data types like text, arrays, and JSONB.