Level: Intermediate Michael Baessler (mbaessle@de.ibm.com), Software Engineer, IBM
Andrea Elias (baader@de.ibm.com), Software Engineer, IBM
Thilo Goetz (tgoetz@de.ibm.com), Software Engineer, IBM
Sebastian Nelke (snelke@de.ibm.com), OmniFind and UIMA developer, IBM
21 Dec 2006
Walk through an end-to-end scenario for creating and deploying
custom text analysis in IBM® OmniFind™ Enterprise Edition 8.4. Find out everything you need to
know from writing and testing a UIMA annotator, deploying it in OmniFind, and
creating semantic search queries, to writing a custom search application
for your custom content. Along the way, find pointers to other
documentation and discover common traps and pitfalls. Be sure to check out the rest
of this
series as well.
Objectives - Develop a UIMA annotator.
- Test and deploy custom text analysis in an OmniFind context.
- Use XML search in the standard OmniFind search UI.
- Write a custom search application using OmniFind's search and indexing API.
Prerequisites
Prior experience with OmniFind helps, but is not required.
System requirements
You need a working installation of OmniFind 8.4. Please note that this tutorial does
not cover the installation of OmniFind. Please see the OmniFind installation
guide for more information. This tutorial makes no assumptions about the OS platform
OmniFind is installed on -- this tutorial should work for all platforms.
You also need to install the UIMA development tooling and the OmniFind base
annotators on a development machine. Please note that there are
sometimes undesirable interferences between the UIMA SDK and an OmniFind
installation, particularly on Windows. We therefore recommend
not
installing the UIMA SDK on your OmniFind server.
Formats html, pdf
Tutorial overview
IBM OmniFind Enterprise Edition (henceforth: OmniFind) is more than a
simple text search engine. You can use OmniFind to build powerful,
so-called semantic search applications with little effort. Using IBM's
open-source text analysis platform UIMA (Unstructured Information
Management Architecture), you can add your own custom text analysis to
the OmniFind indexing process. In a search application, you can use an
XML metaphor to search for the semantic content your custom analysis has
discovered.
This tutorial is a step-by-step introduction to the fascinating world of
semantic search with UIMA and OmniFind. This tutorial starts at the beginning
with the installation of the UIMA tooling you need for developing your
custom analysis, then writing and deploying an annotator in OmniFind,
right up to writing a custom semantic search application.
Please note that this tutorial covers version
8.4
of OmniFind. However, virtually all of this tutorial applies to
OmniFind 8.3 as well. This tutorial points out any significant differences along the way.
| 
