Apache Tika is a toolkit for extracting metadata and textual content from various document formats. Tika itself provides implementation for parsing some document formats while it relies on external libraries (such as Apache PDFBox and Apache POI) for parsing many more.

Tika provides a uniform Java API for all of the supported document formats to make life easier for the user.  Additionally, Tika provides functionality for detecting document type and content language.

In my earlier …

Read more

eWeek.com recently posted a nice article by Dr. Yves Schabes, founder of Teragram, on how to make enterprise search better through some higher order processing techniques like metadata generation, applying taxonomies, etc. and doing relevance testing on a regular basis.  Naturally, this got me thinking about all the different ways this relates to the Apache Lucene ecosystem (Lucene, Solr, Mahout, Tika, etc.) and Lucid Imagination.

First, by choosing an …

Read more