November Happy Hour will be moved to Thursday December 5th.

3rd party search engines - what to consider

Vote:
 

We are about to bye and install an 3rd party search engine.

 There is some engine that has some sort of integration with episerver (for example siteseeker) and other that doesnt (like google mini).

What is there to consider when installing a search engine that doesnt have any episervermoduls or integration packs? One thing that I think of is the security. Can I filter the search result so that you only see what you have access to? Is there other aspects to have in mind?

#28484
Mar 11, 2009 9:48
Vote:
 
In the case of Google Mini I'm fairly sure that you can solve the access rights for search results problem by doing some proper metadata tagging on the page, i.e. adding the authorized groups to the metadata. There may be other ways of solving it, but that is my take on it.
#28681
Mar 18, 2009 10:36
Vote:
 
Hi David,

When I justify EasySearch (on CodeResort EPiCode, an integration between EPiServer and Lucene), there are two main reasons for an integrated search solution:

1) If you don't have any form of integration then your admins and editors either have to learn how to adapt to the search engine (putting in meta tags or other such weird and wonderful constructs). Administrators can't control the index without going to another application.

2) Unintegrated search engines will typically use a crawler than the pages that are indexed are indexed as they appear to a user, which is not necessarily useful.

For example: imagine you have a page that includes content from several other pages. Maybe a "Top 10 News Articles". When someone searches on the text from a news article, you don't want to get back the "Top 10 News Articles" page because, chances are, it's gone from that page now. What the user wants is the specific news article page. I see this so often when using google search. Links to blog page 3; "Today's Postings" and so on and so forth.

An integrated search engine *should* allow you to define what is indexed *in terms of EPiServer content*.

Also, of course, crawlers don't instantly update. An integrated search solution *should* make pages appear in search results the instant they are published; AND remove them again the instant they are deleted.

In addition, as you said, security filtering is important; you have to have the search engine crawler act as the most restricted level of user, or you risk exposing confidential information.

Finally, creating the search page itself is yet more work you'll have to do. Some integrated search products include controls and sample pages that mean you can get search working in a few minutes ;-)

Hope this helps,

Andy

 

#28688
Mar 18, 2009 13:23
Vote:
 

Thanks allot for your comments!

#29405
Edited, Apr 24, 2009 15:11
Vote:
 
  I have been working with several unintegrated search engines
  • 1) Google GSA/Mini - is easy to implement with its XML interface, and does what Google user suppose I should. But you have to host an extra server only for the search functionality. Expensive.
  • 2) Microfoft Search server - pretty easy to implement and works really good. Here you have to host it on a server to; it takes no CPU so I think the web server is able to host it (any comments?).
  • 3) Euroling SiteSeeker - Extremely good support but hard to implement if you don't want standard installation , and the one I used 2008 doesn't support Unicode, so its not possible to indexing Chinese and Korean sites.

And at least a search engine I implemented today on my site http://epiwiki.se/  
Google AJAX Search API, my "search as you type" implementation take about 4hours to do. And if it's working on Google your site gets the same result.  To make just an organic search result page is even easier. BUT you don't have control when googlebot craws your site (it can take weeks).

#29512
Apr 29, 2009 23:40
Vote:
 

Hi, we are also looking into an integration between a third party Search Enginee and EPiServer CM6. It will be helpful to know more about the products you are using and your recommendations.

 

Thank you

#53221
Sep 01, 2011 11:49
Vote:
 

I've integrated following search engine with different episerver sites

Microsoft search server

EasySearch

Forward Search

Google mini (with non - episerver site)

To my best experience I would receommend the Micrsoft Search Server ,as I found this most easier to install , config, and customised the results.Its very much depends on your requirements as well , for instance, if you want to trigger the index update in case of content change (page create, publish, delete etc)  than forward search provides event service to achieve this but it has some issues with it.

Muhammad

#53229
Sep 01, 2011 23:07
* You are NOT allowed to include any hyperlinks in the post because your account hasn't associated to your company. User profile should be updated.