Magazine article Online

Lucene/Solr Now Ready for the Big Leagues

Magazine article Online

Lucene/Solr Now Ready for the Big Leagues

Article excerpt

[ILLUSTRATION OMITTED]

The rich and famous American football player "Neon" Deion Sanders allegedly said, "Sure we're in limos. We're stars. How else is a star supposed to travel?"

Good question. It's one that applies to proprietary enterprise search and retrieval systems. Remember, these systems are frequently the ones online searchers use, even if they don't know it, when accessing library databases, not to mention how they might search for information inside their employers' firewalls.

For decades, the stars of the enterprise were Autonomy Corp.; Convera; Endeca Technologies, Inc.; Fast Search & Transfer; and Verity--what I used to call "The Big Five." Convera went out of business. The others have been acquired by far larger firms. Even the second-tier systems are no longer stand-alone companies. Brainware and ISYS Search Software are part of the struggling Lexmark Corp.'s Perceptive Software unit. Exalead was acquired by Dassault Systemes. Vivisimo, which morphed from a clustering system into a Big Data solution, is now part of the $100 billion IBM, which earlier acquired iPhrase and SPSS Clementine, plus it built its own proprietary information retrieval systems.

In less than 5 years, the landscape of enterprise search has been folded, spindled, and mutilated. The independent search solutions that are available are not yet household names among the Fortune 1000. They are, however, highly global. Examples range from the extremely capable Fabasoft, a Microsoft-centric vendor in Austria, to Funnelback, owned by Squiz, an Australian firm. Based in Bethesda, Md., dtSearch offers a competitively priced solution for the enterprise. Sinequa, with global headquarters in Paris, offers a proprietary information framework, while Transinsight GmbH, based in Dresden, Germany, brings semantics to the search playing field. There are others, but none of the proprietary systems has reached the revenue stature or market impact of the now-faded Big Five.

MAKE ROOM FOR THE ROOKIE

What is pushing aside the aging professionals is a rookie, an upstart--Lucene/Solr. Lucene is the "library," or the plumbing, and Solr is the packaged solution. Today's stars are no longer in limos. Lucene/Solr is behaving like Bo Jackson, another famous American professional athlete who said, "If my mother put on a helmet and shoulder pads and a uniform that wasn't the same as the one I was wearing, I'd run over her if she was in my way. And I love my mother."

Here's the question: "Is Lucene/Solr ready for the professional leagues?"

There are a number of ways to answer this. I want to flip back to the comparison tables I prepared for the first edition of "The Enterprise Search Report." I made a list of what I considered the important functions an enterprise search system should provide to a licensee. As a partial answer to the "big leagues" question, let's look at Table 1, updated for the 2012 search environment, and see if Lucene/Solr delivers on the functions my research suggested were essential.

There are a number of secondary features, what I call "nice to have" capabilities (see Table 2).

OBSERVING LUCENE/SOLR

Some experts may disagree with my Lucene/Solr report card, but I want to make several observations before moving to the next surprising facet of Lucene/Solr as the next big thing in enterprise search.

Lucene/Solr is available as free and open source software. There are some license restrictions, but for most organizations, Lucene/Solr is available without charge. The system can be modified to meet the licensee's requirements. If a special situation arises, the licensee can tap a large pool of Lucene/Solr specialists or turn to one of the commercial open source search vendors that specialized in Lucene/Solr.

Thus, in terms of features, Lucene/Solr is comparable to the Big Five enterprise search solutions. More important, Lucene/Solr is cloud-capable, outfitted with connectors to hook into Big Data management systems such as Hadoop, and equipped with the bells and whistles needed to scale across servers without too much manual fiddling. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed

Oops!

An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.