Post reply

DataWarehouse development

  • April 11, 2008 at 1:56 am

    #71828

    Hello everyone,

    I am new to data warehouse development.

    I'll try my best to make you understand my requirement.

    I have a table which is basically a "job request" table wherein I have a column named "description" which contains the job request description raised by employees in the organization which is in turn assigned to service personnels responsible for solving them.

    I have these issues categorized appropriately which will transform to first level of my requirement.

    I need to make analysis based on this description column by extractingappropriate keywords from this column.

    I have done this by using termlookup transformation in SSIS 2005 and have obtained appropriate results.

    Let me give you a example of what I am trying to achieve.

    Suppose the description column contains say

    "Please install softwarexyx on my systemx."

    So according to my requirement the key word which i have extracted is "install" from the description.

    This transforms to second level of my requirement.

    The third level of detail will contain the exact software name to be installed which in this case is "softwarexyz".

    I need to build an "Analysis Service Project(cube)" for the solution where I have my other dimensions like "Location","Time","Category" etc.

    I want a hierarchy structure to build in the cube which is as shown below.

    Category SubCategory ActualName RequestsCount

    Software install Softwarexyz 50

    The problem is that I am unable to get the actualname of the software for the subcategory install because there are multiple job requests for the same category and subcategory:w00t:

    So how do I build my dimension model to make it work.

    I have a model as shown below

    Any suggestions on this is appreciated.:)

  • April 15, 2008 at 10:11 am

    #802674

    IMHO your requirement has very little to do with DW (ie SSAS) and a lot to do with OLTP (ie SQLSvr)

    if you are dealing with help-desk etc you need immediate CRUD access to raw data, and this has nothing to do with DW and cubes which are effectively aggregates (snowflake with dimensions) and need drilldown to get to individual datapoints [if they are even there and havent been abstracted away]

    BTW, your model didn't come out in the thread

Viewing 2 posts - 1 through 2 (of 2 total)

You must be logged in to reply to this topic. Login to reply