Q: How to mix webcrawl data and external database data...

[josh.jacob]

Hello. I'm working on a search recommendation for a client and had some questions about mixing data gathered through a crawl of a site and data from an external database that drives a dynamic templated page on the site. Let me know if this isn't the correct place to ask.

One of the templated pages is driven by a ZIP code cookie thus has thousands of variations. What are my options for integrating this page with the rest of the site?

Thanks for any comments or pointing in the right direction.

[mark]

Seems like kind of a wide open question to me. Could you provide more details of what you're trying to do?

[josh.jacob]

What I want in the end is the search results for the site to return any relevant content for the site that isn't conditional on a ZIP code plus links to the template pages available in the ZIP code based upon your cookie.

In more detail, this is how the site works right now: If I go to the product section of the site I am prompted to enter my ZIP code. The "index" page of the products section returns a list of the products available in my ZIP code. All the links to the detail page for the specific products go to the same page passing a different ID for the product.

The site in question is for Owens Corning. The URL to the product page is: http://www.owenscorning.com/around/roof ... /index.asp

In terms of a site wide search I could just exclude the product pages and have a search page that listed all the product matches separate from the non-product matches. As I stated above, though, I would like to have everything listed together. In essence I want to rank a product record against Webinators pages and provide a link to the product page template for the correct product.

I know that you can bring in external data to Texis but didn't see anything in the Webinator docs to tell how to get it to work with the page data built from the crawl.

Let me know if I'm still not making sense.

[josh.jacob]

I'm a little fuzzy with how your product works so let me know if I'm wrong here.

For the first option, would a search spawn two searches -- one in the crawl content and one in the ZIP based product data -- and then merge the two results? Is that kinda of how things work? If we had another database driven content search to merge in would we do the same thing -- import the data and do a sub-search based upon some criteria?

For the second option, would I handle the product "filter" in my run-time environment (Java/JSP, ASP, .NET, etc)? Where should I look in the docs for this kind of integration?

One other option I was discussing was making changes to our template page to make special META tags that listed a code (either the ZIP or a distribution code) and then using Webinator to search for that specific META tag. Would Webinator allow for limiting based on a certain META tag?

One last question: We're writing up a quote for implementation for our client and are wondering how to price your product. If we recommend the Thunderstone appliance, do we have to pay extra for a license to Texis and/or Vortex?

Thank you very much for your help!