The Unstructured Information Management Architecture (UIMA) framework is an open, industrial-strength, scalable and extensible platform for building analytic applications or search solutions that process text or other unstructured information to find the latent meaning, relationships and relevant facts buried within. It enables developers to build analytic modules and to compose analytic applications from multiple analytic providers, encouraging collaboration and facilitating value extraction for unstructured information.

The UIMA Java open source framework provides:

• a build time environment for creating UIMA-compliant analytic components (from scratch or by wrapping existing analytic modules) and for developing composite analytic modules
• a runtime infrastructure for executing the composite analytic processes.

UIMA is supported by a diverse community. UIMA is now being standardized. OASIS has formed a UIMA Technical Committee to standardize the architecture and Apache has accepted a UIMA podling project to develop the UIMA framework in an open community. UIMA was originally developed by IBM in collaboration with numerous government, commercial, scientific, and academic institutions and is deployed in production at numerous sites. Carnegie Mellon University's Language Technology Institute is hosting a UIMA Component Repository intended to become the central hub for finding and posting UIMA compliant components.

Open source analytics modules from OpenNLP and General Architecture for Text Engineering (GATE) team at the University of Sheffield can run within the UIMA framework. In addition, many ISVs have developed or are developing UIMA-compliant modules.

UIMA is transitioning its open source development to Apache, licensed under the Apache 2.0 license.

This SourceForge site will continue to host the source code for the Java versions prior to the first Apache version. On SourceForge, the source is licensed under the Common Public License (CPL) and may be freely included in commercial and non-commercial implementations. Open source contributions cannot be accepted for the versions on SourceForge; however we are growing an active, diverse community of developers for UIMA at Apache, where we are accepting any and all contributions and operating Apache UIMA as a full open source community development model under the Apache guidelines.

Pre-Apache versions of the UIMA framework are available as a downloadable SDK designed to assist developers in building and testing UIMA-compliant analytic modules. This SDK includes additional complimentary components beyond what is being delivered in open source, but may not be re-distributed.

IBM has embedded the UIMA framework within the IBM WebSphere Information Integrator OmniFind Edition enterprise search and analytics product, enabling users and vendors to deploy their text and multi-modal analytics in larger integrated solutions.

• Visit the Apache UIMA website.
• UIMA Java Framework pre-Apache versions of the source code
• The UIMA SDK on IBM’s alphaWorks - includes examples, documentation, and the C++ Enablement layer
• Version of the UIMA SDK on IBM’s developerWorks that corresponds to the version of UIMA embedded in recent releases of the OmniFind product.
• Overview of UIMA and many links to related resources.

• November 15, 2006:IBM aids development of Open and Standard Content Analytics Software
• January 23, 2006: IBM Delivers Advanced Search and Business Insight Framework to Open Source Community
• December 19, 2005: GATE 3.1 beta 1 released, includes a UIMA interoperability layer
• November 7, 2005: IBM Software Tracks Blogs, Web Content to Capture Buzz, Spot Trends Around Companies, Products and Marketing Campaigns
• August 30, 2005: IBM Offers Advanced Early Warning for Automotive Industry
• August 8, 2005: IBM Enterprise Search Software to Enable Discovery and Analysis of Information On DemandSoftware Vendors to Support New Open Source Technology for Extracting Knowledge From Unstructured Information
• August 8, 2005: IBM to Open Source Technology for Analysis of Unstructured Information