During the demo, you presented me with questions about LucidWorks Enterprise. I was able to answer some of them immediately following the demo; however, our timeslot for the demo caught up with us, so some were left unanswered. The team here at Lucid Imagination helped me capture your questions, so let’s dive in:

Q: I have multiple domains, each domain should have its own search content and result presentation layer. For example: car search and job search, how could LucidWorks Enterprise handle multiple domains?

A: For complete separation of schema, relevancy and client UIs, use multiple collections. LucidWorks Enterprise is able to handle unlimited numbers of collections. (Available Dec 2010)

Q: Will LucidWorks Enterprise search content within SWF files?

A: In order to search content of SWF, you’ll have to provide the SWF metadata and text to LucidWorks Enterprise. This can be done with SWF file filter, available via commercial packages. LucidWorks Enterprise does not contain one out of the box.

Q: What about language sensitive search?

A: Even though the LucidWorks Enterprise admin UI is currently English only, LucidWorks Enterprise can support multiple languages both for content and for Search UI. There are multiple ways to do so; my favorite one is to organize different language content under different fields, each field can be associated with its own field type and its own analysis chain.

Q: Is Lucid Works Solr (not the enterprise version) still under the Apache License?

A: Yes. The LucidWorks for Solr Certified Version is still licensed under ASF license v2.

Q: When LucidWorks Enterprise connects to DB how does it update the indexes if user has crud multiple records?

A: When indexing content from DB, you can consider the following options: 1. Use SQL logic for delta updates (manage dirty flag, last update field or other DB logic) 2. Use the delta SQL statement when running updates 3. In some cases, use unique IDs from your database to manage duplicates in LucidWorks Enterprise indexes.

Q: Does Solr support personalized search?

A: No. Solr does not support personalized search. Personalized search is implemented in LucidWorks Enterprise only.

Q: Is there an online demo site that allows us to try out?

A: Search.lucidimagination.com, which indexes the mailing lists and other relevant community content of Lucene, Solr, Tika and many other projects is based on LucidWorks Enterprise. You can apply for the developer access program of LucidWorks Enterprise here.

Q: Does this software support spatial searches?  e.g. ‘Show me records with 100 miles of zip code 90210′

A: The out of the box distribution of LucidWorks Enterprise does not include spatial search, this can be added on top of the Solr included with LucidWorks Enterprise.

Q: What kind of document type do you support?

A: Out of the box, we support HTML, XML, PPT, PPTX, XLS, XSLX, DOC, DOCX, ODF, PDF, Quattro, RTF, TXT, ZIP, TAR, MP3 and JPEG

Q: Can you show how to add users to roles?

A: Users can be added either via the ReST API (See: http://lucidworks.lucidimagination.com/display/LWEUG/Users) or via binding to an LDAP server (see: http://lucidworks.lucidimagination.com/display/LWEUG/LDAP+Integration)

Groups are only supported via LDAP integration.

Configuring users and groups to use the search filters (filtering documents for specific users and groups) can be configured via LWE UI, at admin->roles

Q: Has LucidWorks Enterprise been used on any Drupal CMS applications?

A: To the best of my knowledge, not in production yet.

Q: In a database import, how will the index be updated and in sync with database changes?

A: LucidWorks Enterprise supports 2 SQL queries for database importing – one for a full import, and one to be used for consecutive delta imports. The 2 SQL statements configuration is available only when accessing  the ReST API.

Gee.  That almost sounds like a marketing guy wrote it.  Let me try again: LucidWorks Enterprise is software that let’s you easily build great search applications.  You can install it, index some content, and search that content with just a few keystrokes and clicks of the mouse.  Or, you can build a full-blown, integrated search application with custom plugins and your own user interface, all running on a fifty-node cluster.  And if you really need it, the full flexibility of Solr is right there, always accessible.

LucidWorks Enterprise extends Solr to include some enterprise-grade features, just as Solr extends Lucene to provide server functionality.  By extending and embedding Solr, we get to use all of its powerful features without having to do what many folks do: fork the code base.  You see, we are really embedding Solr without changing it appreciably.  Where we do change it, we contribute our code back to the open source project.  A great example of that is SolrCloud, a feature we developed specifically for LucidWorks Enterprise.  The problem was that it was such a fundamental change to Solr that we would have to maintain a forked code-base forever, or simply contribute the code to the open source project.  We chose the latter, and that’s the choice we will continue to prefer for such fundamental and powerful features.

So, what’s in LucidWorks Enterprise?  Here’s a quick summary of the main features:

SolrCloud:

provides simplified configuration management, load balancing and failover in a distributed environment.  It’s the key feature if you have a cluster of search nodes.

User Interface:

we added a web-driven user interface to simplify the process of configuring the system, indexing content, and trying out searches.

CLICK Scoring:

ever wonder how to get better results on searches over your own content?  The CLICK scoring framework provides feedback from searchers back into the relevance ranking of documents, improving relevance for all.

ReST API:

all of the power of LucidWorks Enterprise is available programmatically, so you can build more thorough applications and automated administration.

Smarter Defaults:

LucidWorks Enterprise is configured out-of-the-box to be a great search application, with good relevance and a full list of features like spell checking, autocomplete, and unsupervised feedback.

Data Acquisition:

we have added a crawler and scheduler that simplifies the process of getting data into LucidWorks Enterprise.  Just point it at a Web site or set of files, or import data from a database.

So that’s LucidWorks Enterprise.  We’ll be talking more about it over the coming days.  This is the beginning of an exciting time for Lucid Imagination: we’re really excited to raise the bar with search and get our customers using the latest and greatest in search technology: LucidWorks Enterprise. It’s free for developers, so download LucidWorks Enterprise now! ]]> http://www.lucidimagination.com/blog/2010/10/14/introducing-lucidworks-enterprise/feed/ 1 http://www.lucidimagination.com/blog/2010/02/02/new-lucid-imagination-beta-website-now-up-feedback-welcome/ http://www.lucidimagination.com/blog/2010/02/02/new-lucid-imagination-beta-website-now-up-feedback-welcome/#comments Wed, 03 Feb 2010 01:59:47 +0000 Ken Hoyle http://www.lucidimagination.com/blog/?p=1752 You are invited to visit the new LucidImagination.com, now up for public beta.

What’s new? We have improved the navigation with a simplified, streamlined hierarchy; provided more relevant links and content across the site; and  added new content to help those interested in using Lucene/Solr open source search to drive business growth.

Search Application Developers:
We have developed a focused resource area which will be at developer.lucidimagination.com. Here you will find loads of technical …