MetaMap is a highly configurable application developed by the Lister Hill National Center for Biomedical Communications at the National Library of Medicine (NLM) to map biomedical text to the UMLS Metathesaurus or, equivalently, to identify Metathesaurus concepts referred to in English text. MetaMap employs a knowledge-intensive approach, natural-language processing (NLP), and computational-linguistic techniques, and is used worldwide in industry and academia. At NLM, MetaMap is one of the foundations of NLM's Medical Text Indexer (MTI), which is applied to both semiautomatic and fully automatic indexing of biomedical literature. For more information on MetaMap and related research, see SKR Research Information, MetaMap Portal, and Indexing Initiative (II).
NOTE: MetaMap Transfer (MMTx), our legacy Java-based version of MetaMap, is no longer supported. We strongly urge any remaining MMTx users to switch to the downloadable version of MetaMap2011.
Common uses of MetaMap include:
- Information extraction
- Text summarization
- Question answering
- Data mining
- Literature-based discovery
- Text understanding
- UMLS concept-based indexing and retrieval
- Natural-language analysis of biomedical literature and clinical text
- Downloadable binary and full sources available
- Downloadable UMLS-based datasets for various UMLS releases
- DataFileBuilder suite, which allows users to create their own data sets
- MetaMap Java API to a local MetaMap installation
- SKR Web API to our Batch and Interactive scheduling facility (currently 120 ~3GHz processors)
- MetaMap UIMA Annotator, which encodes MetaMap named entities in a format utilizable by UIMA components.
- MedPost/SKR part-of-speech tagger server and Word Sense Disambiguation (WSD) server
MetaMap Tutorial Slides Available:
- MetaMap 2010 AMIA Tutorial Description - PDF (38kb) or Text (3kb)
- MetaMap 2010 AMIA Tutorial Slides – PowerPoint (3.2mb) or PDF (1.6mb)
MetaMap is intended to be easy to download, install, and run out of the box. Site-specific customization of MetaMap is expected; however, users desiring to create their own data files will need a moderate amount of programming knowledge to use our DataFileBuilder suite; users wishing to modify the MetaMap source code will need programming expertise in Prolog and a SICStus Prolog license, available from the Swedish Institute of Computer Science.
October 21, 2010 - With the 2010 Release of MetaMap, we are retiring three previous versions of MetaMap, namely MetaMap07, MetaMap08, and MetaMap08V2. Only the MetaMap binary executables are being retired; the MetaMap UMLS datasets corresponding to these releases (2007AA, 2008AA, and 2008AB) will remain available.
MetaMap 2010 includes less new functionality than previous releases because the bulk of our development efforts since MetaMap09V2 have focused on converting MetaMap from Quintus Prolog to SICStus Prolog, which will henceforth be the principal implementation vehicle of MetaMap. We also converted MetaMap 2010 to version 4.8.24 of Berkeley DB, as recommended by SICS.
New functionality and enhancements delivered in MetaMap 2010 include the following:
* De-Normalized Data Tables,
* Minimum Concept Length,
* Changes in Numerical Output Format,
* Silent Mode, and
* Variants Bug Fix.
MetaMap UIMA Annotator - April 2010. The MetaMap UIMA Annotator encodes MetaMap named entities in a format utilizable by UIMA components. The annotator is based on the MetaMap UIMA Wrapper (http://sourceforge.net/projects/metamap-uima/) authored by Kai Schlamp. The annotator utilizes classes from the 2009v2 version of the MetaMap Java API which is required for use of the annotator.
Differences from Kai Schamp's MetaMap UIMA Wrapper include a modified UIMA type system which includes Acronyms and Abbreviations and the addition of a MatchMap structure to the Candidates and Mappings. Use of MetaMap XML and the SKR API are currently not supported.
MetaMap 2009 v2. This release provides an updated version of MetaMap consisting of mostly bug fixes and enhancements to our implementation of NegEx, the AA-detection logic and lexical processing.
This release includes also three more substantive changes:
New Phrase-Breaking Method,
MatchMap Consolidation, and, most significantly,
Improvements in XML generation.
Please see the Release Notes for more information.