How do I set Microsoft Clustering to do full text comparisons on strings?

Answered How do I set Microsoft Clustering to do full text comparisons on strings?

  • Tuesday, November 13, 2012 8:40 AM
     
     

    Hi guys. I'm trying to find clusters of people and addresses in a table of companies and people.

    An address varchar column may contain:

    '8 oak avenue'

    'Oak avenues'

    How do you set the clustering to cluster those together, i.e. partial string comparison?

    Does it do it automatically?

    Do I need to create a full text index on every column that I want to do a partial text match on?

    Thanks,

    Richard.

All Replies

  • Wednesday, November 14, 2012 10:30 AM
     
     

    Instead if full text search you can use Querying Alternatives in SQL Server,For more information see below URL

    http://www.mssqltips.com/sqlservertip/1342/full-text-search-querying-alternatives-in-sql-server/


    Please remember to click Mark as Answer and Vote as Helpful on posts that help you. This can be beneficial to other community members reading the thread.

  • Thursday, November 15, 2012 6:39 AM
     
     
    Any updates?

    Please remember to click Mark as Answer and Vote as Helpful on posts that help you. This can be beneficial to other community members reading the thread.

  • Thursday, November 15, 2012 2:38 PM
     
     

    Thanks, but I specifically want to know if an Anlysis Services Data Mining model can use partial\fuzzy matching and how to do it (possibly through a full text index). 

    I'm not interested in normal SQL statements.

  • Thursday, November 15, 2012 3:30 PM
     
     Answered

    Analysis Services does'nt do any fuzzy matching as far as I know. You need to prepare the data and present to it.

    One possible way you can cluster on a given scenario is to create Case and Nested tables stuctures. As a same address can exist in multiple cities, state, counties, you might want to create Case/nested tables as given:

    e.g.

    Casetable<- PersonID,

    Nestedtable<-PersonID,

                       PartialAddress

                       State

                       City

    You can accomplish creating this structure using SSIS term extraction and term lookup components. Or simply using string functions and row_number() over partition function in SQL.

    hth,

    rok


    please remember to mark as answered if the post helped resolve the issue. ------------------------------------------------------ "Essentially, all models are wrong, but some are useful" - George E.P.Box

    • Marked As Answer by RichardJECooke Saturday, November 17, 2012 10:03 PM
    •