 Ask Our Expert: October 2006
|
|
To do fine grained query, is it helpful to have metadata, or is full text search of the content chunks good enough?
-
MarkLogic Server supports XQuery over native XML - that is, there's no need to define "chunks" of content up-front that you want to be able to search. In effect, every XML element (and attribute) at every level in the XML tree is treated as a "chunk", and its content and structure are indexed as the content is loaded.
In the webinar we discussed the "XML ladder" - the idea that XML is flexible enough that you can start simple and add structure, metadata, semantic markup, etc. over time. So, while it's very useful to have rich metadata - and including metadata constraints in your search will certainly make the results more accurate - it's possible to build effective content applications without metadata, and enrich your content (and your application) over time.
I already own a search engine. Why not just use my search engine to build these content applications? What advantages does Mark Logic really offer over my search engine?
-
With an XML content server such as MarkLogic you can do fine-grained query, manipulation, and rendering of your content. That is, you can look inside each document or page, making full use of whatever XML structure and markup is in the content. Then you can manipulate and re-assemble parts of each document or page - at arbitrarily-fine levels of granularity. You just can't do that with a search engine.
A lot of people try to combine a relational database management system such as Oracle with a search engine, to create rich content applications. Often it is possible to create content applications this way, but only with a tremendous amount of development effort, and the end result is brittle and difficult to change. That's a direct result of the need to define up-front what can be queried and how, when building these hybrid systems. With MarkLogic Server, all content and structure is indexed automatically at load time, and applications are built with XQuery. So applications are built very quickly, and are very flexible.
MarkLogic Server was built from the ground up to enable content applications - it offers the benefits of a relational database, a native XML database, and a search engine, all tightly integrated by design. That's why major content providers such as Elsevier, Wolters Kluwer, O'Reilly, and Oxford University Press use MarkLogic Server as a platform for their content applications.
|
|
|