Big Data BBQ

Meta­data is crit­i­cal to to help the user find the reports he or she are look­ing for as it adds addi­tional infor­ma­tion about what’s in the reports. What’s meta­data? Strictly it is “data about data” or addi­tional descrip­tions, tags, fields that describe the data or report. Here’s a sug­ges­tion for small report­ing portals:

• Brief descrip­tion [short 140 char descrip­tion of report pur­pose, content]
• Long Descrip­tion [longer text descrip­tion of report pur­pose, content]
• Update date [last data updated]
• Geo­graphic range [geo­graphic hier­ar­chy — con­ti­nents or countries]
• Fre­quency of update [how often report is updated]
• Thumb­nail [image of report]
• Audi­ence [audi­ence of who can access report or who it is tar­geted at]
• Notes [update notes on report, data issues or changes]
• Report­ing alias [who to con­tact with ques­tions or to get access]

It’s pos­si­ble to do this as indi­vid­ual fields or com­bine into one long field (eg Long descrip­tion). One key thing is to stan­dard­ize the vocab­u­lary for each field. For example:

• Fre­quency of update: Daily, Weekly, Biweekly, Monthly, Quar­terly, Annually
• Report for­mat: Spread­sheet, PDF, HTML,PNG
• Geo­graphic Range: North Amer­ica, Cen­tral Amer­ica, South Amer­ica, Europe, Asia Pacific, Mid­dle East, Africa

Some exam­ples of meta­data, using the reports men­tioned in the ear­lier post:
_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​
Online Page View Report
All page views on main web­site dur­ing the cal­en­dar month for North Amer­i­can vis­i­tors.
All page views on main web­site dur­ing cal­en­dar month. Sum­mary and detail with graphs. Refreshed: 4/​15/​2012. North Amer­ica, Monthly update (week 3 of month). Online Team access only, spread­sheet + html for­mat, Data is miss­ing for sev­eral states (NJ, NY) , included in next refresh.

Alias: regrept­team

_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​
Sales_​Units_​Revenues_​Central
Sales report includ­ing rev­enues and units by prod­uct for the Cen­tral Region.
Sales and units for all prod­ucts dur­ing cal­en­dar month for the Cen­tral Region includ­ing pivot tables and graphs by account. 4/​1/​2012. Cen­tral Region. Monthly update (week 1 of month). Sales team access only. spread­sheet for­mat. Data now includes all new accounts added dur­ing prior month.

Alias: regrepteam
_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​_​

Some key best prac­tices to mak­ing the meta­data work correctly:

• Be con­sis­tent in the vocab­u­lary for each meta­da­tum. Plan out the words you will use (con­trolled vocab­u­lary). If you do one long descrip­tion, keep the order of data points con­sis­tent (same order) for each report.
• Doc­u­ment all the reports in this way as it will facil­i­tate use. Min­i­mally doc­u­ment the most used reports with metadata.
• Once you have attached meta­data, you can eas­ily cre­ate sum­maries, fil­ters, or tags for your data to enhance find­abil­ity of data and reports.

Next we will look at how to set up a folder struc­ture to make infor­ma­tion easy to find and use. (3 of 9 Best Practices)