This information page will seek to guide you in your research into Microsoft
Office SharePoint Search products. We say products because there exists a
separate search SKU from the MOSS portal which can be installed as a standalone
product and integrated into existing portals or intranets. The standalone
SKU or product is basically MOSS portal with all of the MOSS features accept Search
removed, that also means no BDC so custom protocol handlers are required to
connect to your line of business applications.
There are many resource links on this page and some of it can be
overwhelming. A great place to start off is this guide from Microsoft:
SharePoint Search Evaluation Guide just print it out and read it
thoroughly. Then take a close look at this diagram below to see all the pieces
and where they fit together.
Indexing
This is how content is added to the search databases: both the SQL Server
Metabases and full text index files (referred to as the Content Index above). This
process is handled by the Index Engine in SP and utilizes
protocol handlers and
ifilters to connect to the different content sources. They extract the search
text and metadata which are them processed by Wordbreakers and Stemmers before
being added to the index. Metadata are properties like title, author, client,
matter etc. and are used to further identify and classify the content to make it
sortable, groupable and advanced searchable. In order to index proprietary
content types that are not already indexable by the standard SP protocol handlers, you
need to develop or acquire a custom protocol handler that understands how to
connect to and work with your data source. See our
list of SP Products. Alternatively you can also utilize
the BDC to index your business data - for more on this see:
BDC Information Page . For even more details
on indexing see
Indexing Explained
Querying
The SP Query engine is accessible through the SP object model and SP web
services to allow custom development against it. For most implementations you
will never have to worry about how to connect to either as the MS Search web
parts in the Search Center or 3rd party ones like
Ontolica take care of that. These search interfaces are responsible for
generating the query that achieves the best relevancy : see
Writing SQL Syntax Queries for Relevant Results. For those that what to
write their own web parts that utilize the query interface, check out these links:
Building Custom Enterprise Search Web Parts,
Query Web Service Tool and
Enterprise Search SQL Syntax Reference
Results
This is the part of the Enterprise Search system that your users will only
care about, the rest is behind the scenes. It is how your users interact with
the system and requires the most thought. Your options on user interface varies
from customizing the "Out of the Box" Search Center(Customizing
SharePoint Search Results Display), to implementing a 3rd party search
interface product like Ontolica
Search for MOSS 2007 or
Longitude. Of course if you are really ambitious you can write your own
interface:
Building Custom Enterprise Search Web Parts