ISO/IEC 13249-2:2003

Overview

ISO/IEC 13249-2:2003 - "Information technology - Database languages - SQL multimedia and application packages - Part 2: Full-Text" defines the full‑text component of the SQL multimedia and application packages family. The standard specifies full‑text user‑defined types and associated routines for SQL environments, together with notations, conventions and concepts needed to support advanced text handling inside database systems. It targets robust, database‑integrated full‑text search and ranking (tokenization, pattern construction, scoring) without redefining core SQL.

Keywords: ISO/IEC 13249-2, SQL full-text, full-text search, user-defined types, tokenization, stemming, proximity search.

Key Topics

• Full‑Text Data Model and Concepts: text model, text identification facilities, multilingual handling, stop‑word treatment and word normalization.
• User‑Defined Types (UDTs) and Routines: a family of FullText types (e.g., FT_Pattern, FT_Phrase, FT_StemmedWord, FT_Fuzzy, FT_Proxi, FT_Soundex) and their methods for contains, score, tokenize, tokenizePosition and segmentize.
• Structured Search Patterns: single words/phrases, sets, Boolean operators, context patterns, proximity, fuzzy and thesaurus‑based expansions.
• Search & Ranking Facilities: methods for identification of matching text, text scoring, number‑of‑matches and ranking semantics.
• Tokenization and Stemming: token position tracking, tokenization with or without stemming, and related helper routines.
• Interoperability and References: uses definitions and references from ISO/IEC 13249 (all parts), ISO/IEC 9075 (SQL), and ANSI/NISO Z39.19 for thesaurus handling.

Applications and Who Uses It

ISO/IEC 13249-2 is intended for implementations where SQL databases need integrated, standards‑based full‑text search and ranking. Practical applications include:

• Library and archival systems - precise search across large text collections with thesaurus support.
• News/media & multimedia - indexing and retrieving textual metadata, captions and transcripts.
• Scientific research repositories - advanced pattern matching, proximity and phrase queries.
• Enterprise content management, decision support, data mining and data warehousing - integrated text analytics, search and ranking inside relational environments.

Primary users:

• DBMS vendors and product architects who implement standardized full‑text UDTs and routines
• Application developers building SQL‑based search applications
• Data architects and information managers specifying searchable text models
• Search engineers implementing proximity, fuzzy and thesaurus‑driven expansions

Related Standards

• ISO/IEC 13249 (all parts) - SQL multimedia and application packages framework
• ISO/IEC 9075 - the SQL standard (referenced for SQL definitions and conventions)
• ANSI/NISO Z39.19:1993 - thesaurus standard referenced for terminology and thesaurus expansion

This standard helps ensure interoperable, SQL‑centric full‑text capabilities-useful when precise, standards‑based search, ranking and text pattern constructs are required in database systems.