Comments for The Data Quality Chronicle

Comment on Justin Bieber abuses Twitter but proves how similar phone numbers are by William Sharp

William Sharp — Sat, 31 Mar 2012 14:17:08 +0000

They can be used as a confirmation of accuracy, however, if you consider how multiple customers can share the same phone number this may not be as accurate as you’d think. In business situations different people can share a phone number or have very common phone number. In retail type situations multiple customer households complicate this in the same way.
I have used phone number as confirmation of an accurate match, however, that can be misleading.

Comment on Justin Bieber abuses Twitter but proves how similar phone numbers are by Prashanta C

Prashanta C — Sat, 31 Mar 2012 12:15:43 +0000

Good one William,

Using phone number(s) in matching may not be a great idea as you mentioned, but I still tend to take it as an affirmative criteria when most of the other critical elements end up being good matches. Say you have a name, address, DOB, gender and identifier match, a matching phone number kind of confirms a good match. Usually we apply edit distance on phone. In above scenario we might give a good boost to overall score if it’s exact phone number match and lower(significant) if the match is 1 or more edit distance.

My 2 cents

-Prashant

Comment on How To Integrate Social Media Into Your WordPress Site by Prash Chan (@MDMGeek)

Prash Chan (@MDMGeek) — Tue, 24 Jan 2012 21:59:25 +0000

Can’ wait 2 x’plore RT @dqchronicle How To Integrate Social Media Into Your WordPress Site | | The Data Quality Chronicle bit.ly/wKj81m #li

Comment on Can Big Data and Social Media help HR in efficient/quality hiring? by Andy Wenzel (@AndyWenzel)

Andy Wenzel (@AndyWenzel) — Mon, 23 Jan 2012 02:05:48 +0000

Can Big Data help with HR in hiring? http://t.co/eZyyfUNc

Comment on Too much push, not enough pull by Loraine Lawson (@LoraineLawson)

Loraine Lawson (@LoraineLawson) — Sat, 21 Jan 2012 15:16:37 +0000

Poll on data quality. http://t.co/Re9PNkz3 by @dqchronicle

Comment on Big Data … Little Data Quality by DATAVOTE (@contactdata)

DATAVOTE (@contactdata) — Wed, 18 Jan 2012 06:23:58 +0000

Interesting post http://t.co/jgUCgSN6 by @dqchronicle

Comment on The Seven Habits of Highly Effective Data Quality by David Pratt (@DataMgmtWonk)

David Pratt (@DataMgmtWonk) — Wed, 18 Jan 2012 02:28:17 +0000

Just read "The Seven Habits of Highly Effective Data Quality" via @dqchronicle http://t.co/EfC5vs7N #dataquality #7Habits #covey

Comment on When using data quality tools, what do you include? by (@AxelTroike) (@AxelTroike)

(@AxelTroike) (@AxelTroike) — Tue, 17 Jan 2012 13:09:13 +0000

My comment on post by W. Sharp ( @dqchronicle ): When using data quality tools, what do you include? http://t.co/FMoZGgsI #DataQuality

Comment on When using data quality tools, what do you include? by Axel Troike (@AxelTroike)

Axel Troike (@AxelTroike) — Mon, 16 Jan 2012 19:46:19 +0000

Commented post by W. Sharp ( @dqchronicle ): When using data quality tools, what do you include? http://t.co/FMoZGgsI #DataQuality

Comment on When using data quality tools, what do you include? by Axel Troike

Axel Troike — Mon, 16 Jan 2012 19:39:21 +0000

First and foremost, people in any organization need to have a documented common understanding what are the data objects of interest for their business and what are the describing attributes. To create this common point of reference,
a data modeling tool is required.

In a second (better: parallel) step, the processes that create, update or delete those data objects need to be identified. Each process that manipulates data is a possible factor to influence the data quality for better or for worse.

To document the identified processes and their specification, a process modeling tool is required. (I recommend to use a process modeling tool that is integrated with the data modeling tool in the way that processes can be linked to the data and vice versa.)

Having such a fundament, you can examine process after process to eliminate possible quality risks, by adding integrity checks to your transactions, by training your staff to double-check data at any point of entry, by additional batch tools that correct / clean-up data on a regular basis etc.