A Short Guide To Installing MDS 2012

Before I can talk about using MDS (Master Data Services), I feel a small responsibility to help you get MDS up and running. Although MDS has been available since SQL Server 2008 R2, the method of installation has changed from the use of a separate installer file (.msi) to be baked into the installation wizard of SQL Server itself. . Fortunately, that change has been to make it easier. However, I will focus on the SQL Server 2012 installation here.

If you did not install all Shared Features, you may want to reinsert your SQL installation disk and step through the screens until you reach the Feature Selection screen as shown in the following figure. If the Master Data Services option is already selected, that means that you previously installed MDS as part of your current SQL Server installation. If it is not selected, you must select this Shared Feature and then proceed through the rest of the installation screens to add the necessary bits to your server.

After MDS has been installed, you must configure it. To do this, open the Master Data Services Configure Manager. You can find this application in the Microsoft SQL Server 2010 folder under the Master Data Services folder. Note, before you can run MDS Configuration, a check for prerequisites is made on the machine. MDS requires Windows PowerShell as well as Internet Information Services. In fact, if IIS is not installed, selecting MDS during the SQL install may not install MDS correctly and you will be prompted to reinstall it.

Assuming that your machine has all of the necessary prerequisites, you can proceed to the next step. The configuration manager consists of two pages of properties. The first page is the Database Configuration. Just like DQS (Data Quality Services) you must select or create a MDS database that MDS can use for its working area. To do this, you must define the database instance and the specific database to be used for Master Data Services. For now, we can leave the other System Settings at their default values.

The second page of the configuration manager prompts you to create the web site to be used with MDS (which is why you need IIS on the machine you are running MDS. You must also identify the database created on the first page of the configuration manager. Finally, you choose whether you want to integration your MDS installation with Data Quality Services (DQS). This feature only exists in SQL Server 2012 because DQS is new to SQL Server 2012.

At this point, you should be able to start up MDS. The main screen shown below adds a feature in 2012 that did not exist in the 2008 R2 version, the ability to use an Add-in for Microsoft Excel to help you load and administer your data.

WOW, it sure seems like Microsoft Excel is getting used throughout more and more of the Microsoft BI stack. First PowerPivot and Power View, then the tabular model in SSAS and now MDS. If you never took the time to learn Excel because you thought you were a SQL guru and did not need Excel, perhaps now is the time to brush up on your Excel knowledge.

Well, I going to stop here for this week so you can get MDS set up on your machine. Besides, it is Labor Day weekend and most of my United States readers probably have better things to do this weekend.

C’ya next time.

Is It The Beginning of the End or the End of the Beginning?

I just read a story about the decrease in the number of amphibians.  Yes, I talking about those frogs, salamanders and lizards.  Although living here in Florida, it seems like there is an over abundance of those tiny lizards that live in your garden and your lawn and sun themselves on the sidewalks until you get close when they scamper away.  In fact, I’ve noticed a lot of little baby lizards, no more than an inch long (full size can be four inches or more).

But the news article specifically pointed to only a few species that appear to have decreased in populations by 50 percent or more within the last decade.  Although scientists are still not sure of the cause of this decline, they cite factors such as disease, new preditors, the increased use of pesticides, and yes, of reason behind all negative change in the world, climate change.

If this story was an isolated report and only about a few frog species, perhaps the majority of people would take little notice of the warning.  However, could it be a part of a larger pattern?  Digging around the internet, it does not take long to find reports on the loss of honeybee colonies, bats, and the decrease in marine animals and especially the shrinking of coral reefs.  You could also look at the loss of many species of plant life especially those that grow only in tropical rainforests which are being cut down at an alarming rate.  How long will it take until the loss of these species directly effect us, our food supplies, our environment?

Of course, the other side of the argument is that species come and go all of the time.  It is called evolution.  The troubling fact is the rate of extinction may be accelerating.  Such mass die-offs of species has happened several times in the past.  Mass extinctions have occurred several times in the past and each time, the extinction seems to be related in some way to climate change, either warming or cooling.  Sometimes the climate change may be initiated by an external factor such as the meteor that led to the extinction of the dinosaurs.  Note that I did not say that the meteor caused the extinction of the dinosaurs.  Most scientists today believe that other than dinosaurs living in the immediate area of the meteor strike somewhere off the Yucatan peninsula most dinosaurs died over hundreds if not thousands of years.  Scientists have modeled the effect of the dust and dirt kicked up into the atmosphere by the meteor and possibly volcanoes that erupted shortly afterward.  While the dust could have lead to cooler temperatures as the sun’s rays were blocked, gasses from the volcanoes could have led to a greenhouse effect much like the one around Venus trapping heat and raising temperatures on the surface.

In any case, most species did not survive.  While the die-out may appear instantaneous from a geological point of view, it could still have taken decades or even centuries for some of the species to die off completely and then slowly be replaced by other species who could cope with the climate changes and adapt.

So if most of the extinction events can be associated with some type of climate change, does it follow that today’s climate change and loss of dozens of species in the last century mean that we are at the start of the mass extinction?

Even if that is true, does that mean that mankind is doomed to become extinct as well?  Keep in mind that with every past mass extinction, some species did survive and even thrive in the new climate conditions.  The more adaptable, the more likely a species might survive.  Mankind, if nothing else, is adaptable.  With our increased use of technology, we may be able to survive any mass extinction even to the extent of manufacturing our own organic food in factories in petri dishes rather than farms.

However, the world could be very different in a few thousand years as more species die and perhaps new ones come to the forefront to fill in the gaps in the ecosystem.  If you could hibernate for the next thousand or perhaps five thousand years, what type of world would you wake up to?  Would you recognize anything?  Yet that new ecosystem could over time become just as vibrant as the one we have today.  In effect, the end of this era might be nothing more than the start of the new era.  The only question I would have is how mankind would fair during this transition.  While we may survive due to technologies we already have today, will mankind fall below a critical population level in which science and technology become unimportant compared to the daily needs of survival?  Will we slowly lose the ability to develop new technology?  Will we even be able to continue to maintain the old existing levels of technology?  Will mankind revert to living in caves?  Will all written record of our great society be lost as the records deteriorate?  Will the knowledge and technology lost over time become the substance of myths and legends?  Will future archeologists ponder over how their primitive ancestors were able to develop the skills to make the strange discoveries they unearth?

Could this have happened before?  Could our current society be just one of several that rose and fell in the past?  Such concerns give more validity to the arguments why we need to explore space and establish colonies elsewhere so that mankind may survive any disaster to this vulnerable planet.

But then just perhaps, this has happened before.  Perhaps the people of Atlantis really did have flying ships.  Perhaps they did explore space and leave colonies elsewhere.  And for the real conspiracy theory advocates, maybe some of the UFOs reported today are nothing more than our ancient ancestors who survived their mass extinction returning to see how the new life on planet Earth survives this one.  So could the beginning of the end of this cycle simply be part of the end of the beginning of the next cycle?

Ok, I would guess you did not expect this story to take that twist.  However, remember when a science fiction fan lives inside a writer, all bets are off.

C’ya next time.

MDS: The New DQS? Not Really!

First, you need to know that the MDS I’m talking about here is Master Data Services introduced in SQL Server 2008 R2. I’ve already talked about Data Quality Services (DQS) which I believe was only introduced in SQL Server 2012.  So the natural question is what is MDS and why do I need it together with the newer DQS or is DQS just a replacement for MDS and can we forget about learning MDS now that we have DQS?


Previously I talked about DQS and showed how you can use it to create individual knowledgebase domains that can then be used to clean data.  The application of DQS on a regular based to clean data going into a production database should improve your overall efficiency compared to manually cleaning the same data.  This is because DQS remember the correction you made previously.  You can also define rules for ‘fixing’ individual domains.  Efficiency improves over time because the knowledge base understands how to handle an increasing number of problems without requiring manual intervention.  However, at the same time, users have to be realistic.  There will probably never come a time when absolute all the possible data issues have been addressed allowing for fully automatic data cleansing.


You could think of DQS as the detergent that helps you clean bad data from a data source getting rid of invalid and misspelled data.  It can also help you reduce duplicate records within that individual data source.  It can even output a data file of cleansed information with duplicate records removed or at least marked. However DQS provides no assistance in merging the clean output data file with another data source.  The ability to combine a freshly cleaned data source with a master data file is what Master Data Services can provide.  To some extent, both products have the ability to cleanse and match data from an input data source.  However, the rules in DQS are domain (column) centric while the rules in MDS are typically more record centric.   Also MDS as the name implies helps you import data from new sources into a common master table.  Another way of saying this is that MDS might be the gatekeeping of your enterprise data system that controls what data can or cannot be added to your data warehouse. 


Because both DQS and MDS provide the ability to define cleansing rules, users may jump to the conclusion that these two products are mutually exclusive.  However, the purpose of the cleansing rules between these two products needs to be clarified.  The purpose of the cleansing rules in DQS is primarily to correct and validate values added to each domain.  DQS rules can determine which values are allowed or not allo0wed in the domain.  These rules are not record based, but column based.


On the other hand, the business rules created in MDS primarily exist to prevent invalid records from being added to a master data table.   In other words, it keeps new records from entering or updating the master set of records which may sit in your enterprise data warehouse. 


Based on this description, you should see that DQS and MDS are clearly different products.  Furthermore, you may not always need to use both of them.  If your purpose is merely to clean bad data from a data entry system and possibly remove duplicate records, you may be perfectly satisfied with using only DQS.  Similarly, if your task is to merge clean data from multiple data sources such as sales data from different countries, you may be more interested in the abilities of MDS to combine records based on business rules that prevent record duplication while allowing data to be merged.  Of course if the data sources that you want to merge have not first been cleaned, you may need to first run them through a DQS project with a corresponding knowledgebase to ‘fix’ any domain issues such as different product codes for the same product in different countries or perhaps even differences in the way customer address information is recorded.


In future weeks, I will take you an exploration of MDS in between some other technical topics that I want to cover.  Before I end today, I want to remind all my readers of two events coming up very soon where I can meet you.

The first is the SQL Saturday Orlando event coming up on September 14th.  SQL Saturday events started right here in Orlando back in November 2007 by Andy Warren.  I’ve been fortunate to have been a speaker at every Orlando event.  This year I will be presenting an introduction to DQS that I call, ‘Putting Quality Back Into Your Data’. 


The second big event coming up is the 2013 PASS Summit that this year is being held in Charlotte, North Carolina during the week of October 15-18.  This will be my second year speaking at the PASS Summit.  If you have never been to a PASS Summit, you are missing the opportunity to meet, talk with and listen to presentations from SQL professionals from around the world.  I feel very honored to be able to give two presentations this year, one on DQS and the other on using SSAS’s Tabular Model to build Cubes that you can manipulate as Excel PowerPivot models.

If you have the opportunity to get to either of these events, please stop by to say Hello.  I would love to meet you all.


C’ya next time.


So Area 51 Exists. Big Deal!

Last week the CIA officially announced that Area 51 exists.  Wow!  Where is the surprise?  Pictures of the area have existed for a long time.  It has been rumored to be the home of everything from advanced avionics research include spy plane and stealth technology to the home of alien spaceships and even captured or voluntarily visiting aliens.  So what did they really tell us that was new?  Nothing.  Of course they hope that everyone believes them just we all believe that the government would never spy on citizens or that universal health care will be cheaper or that they target political groups or that they allow US citizens abroad to die without at least attempting to save them or even the biggest lie, “I’m from the government and I’m here to help you.”

Of course the press made it sound like such a great revelation that Area 51 really exists.  One thing you can believe, if anything ever was beyond top secret at Area 51 it has long been moved to another location and there is nothing left to see in Area 51.  Of course they will still keep up appearances and run off anyone that comes near the area just to keep up appearances.  In all likelihood, any top secret projects (including any alien spacecraft or even aliens themselves) have been relocated to a more secure area that is not under surveillance by UFO hunters or the curiosity seekers.

Of course Nevada, Arizona, and New Mexico are pretty big states with lots of open area with nothing around for miles.  Large military bases are scattered across all of these states and a ‘new’ secret area could be anywhere, maybe even those large bunkers in part of Nevada near where nuclear weapons were tested.  Or maybe those were just bunkers for the testers.  In any case, it might not even be in those three states.  It might be in Colorado, Montana, Idaho or deep in the abandoned mines of Appalachia.

Just because there is nothing to see in Area 51 now does not mean that there never was.  The only real question is whether anything that was ever there came from somewhere other than this planet.  The problem is that the government has lost so much credibility that even if they came out and said they had aliens, people would think that they were just making it up to feed the egos of the UFO hunters and believers.

All this clouds the real issue as to whether there is other life in the universe.  Growing up, anyone who believed that there was life on other planets was considered a little crazy.  One scientist even postulated that even if there was only one other advanced life form in our galaxy and if it had developed a mere million years ahead of us (which is nothing on a cosmological scale) that they would have by now spread throughout the galaxy and would have contacted us or conquered us or eaten us.

Others believe in the Star Trek based Prime Directive which prohibits the Federation from contacting civilizations that have not at least developed warp travel under the theory that until they develop warp travel, they will never get to another star system anyway much less one with life.  But more recently the real world of astronomy has been rocked by the ‘discovery’ of planets revolving around dozens of local stars.  Some of these planets are theorized to be within the star’s habitable zone.  Our abilities to determine the chemical signature of these planets is still not refined enough to tell us whether any of these newly discovered worlds have gases in their atmosphere that would indicate some type of life on them.  Still, the ability to detect those chemical signatures is not too far off.

So what happens if (or when) we discover other life in our galaxy?  What will that do to society?  Will religion collapse as we know it or will religion adapt to the fact that “my father has many rooms in his house” (John 14:2) meaning that there are many habitable planets in the universe.  In fact, the Vatican has stated that aliens do exist on other planets and they are all part of God’s greater plan.  HHhmmm, wonder how they know that?

So if aliens exist, the next big question is where are they which is quickly followed by what will they do?  Again science fiction has given us a lot of different possibilities which we do not have time to go into here.  Check out the Internet on this topic for the various possibilities.  Maybe we are not interesting enough.  Maybe we have nothing to offer the galactic confederacy.  Maybe they are waiting to see if we can get past our violent stage.  Maybe compared to them, we are as unintelligent as we think animals are compared to us.  So why aren’t we their pets?   Or is that what the alien abductions are?

As you can see, it is too easy to come up with conspiracy theories about what the government isn’t telling us when they wait years to tell us a secret that we already all know about.  So, to the CIA, thanks for nothing.  When lies are built on top of lies surrounded by lies and filled with deception, where do we go from here?  And maybe that is the real reason the aliens will not welcome us into their little group.

C’ya next time.

Promoted Links – SharePoint 2013

Promoted Links is a new list type in SharePoint 2013. In some ways it is very much like the link lists available in previous versions of SharePoint. However, it adds a couple of special new features such as the ability to include a background image with a link and the ability to control the link target to open the linked content in a separate window.

To begin using a Promoted Links list, I navigate to the site where I want to use the list and open the Site Contents page from the Settings icon in the upper right of the page (the new Site Actions button). Next I click the ‘Add an App‘ button at the top of the Lists, Libraries and other Apps section of the page.

On the Your Apps page, look in the Apps you can add section for the app titled Promoted Links and click on it. SharePoint then prompts me to name the list or I can click the Advanced Options link to continue, but this only adds the additional metadata field of a description for the list. In this example, I am going to create a Promoted Links list of all the schools from a school test project. Therefore, I will name my list: Test_Schools. (Note, while spaces are allowed, I prefer to use underscores in their place.)

Once I create the list, it appears in my Lists, Libraries, and other Apps list. Initially it displays the ‘New‘ icon immediately beneath the object name. This feature makes it easier to find the newly created object. While it may not seem that important when I only have a few objects in my site, as my site grows and I create dozens of libraries and lists, this feature is worth its weight by quickly highlighting the new objects I created and am most likely to use.

When I click on the new list, SharePoint shows me the default message telling me that the list is empty, but that I can add tiles from the All Promoted Links view. At first, this statement may seem a bit confusing especially if I was expecting it to say something about adding a new item to the list. This statement however is just referring to the fact that there are two ways to display links in this list. I will come back to illustrate these options after adding a few links to the list. For now, I click the link All Promoted Links and then click New Item to display a form to enter the link information into the form shown below.


Note that only three of the properties of the promoted link are required (have an ‘*’ at their end). Let’s look at those first.

The link property Title is the text that appears to the user when displaying the link on a page. It is required and can consist of any text you like up to 255 characters in length. Of course I can change the default maximum length by editing the list settings. For this example, I will use Edom Middle School.

The second required property is the Link Location. This property is the URL or address that I want to navigate to when the user clicks the link. Note that this field is divided into two components. The first component is the actual link which in my example is: http://schoolcollaboration.fl.net/SitePages/Home.aspx. The second component is a description for the link which I will call Edom Middle School Site.

The third required property is the Launch Behavior. This property tells SharePoint how to open the link. For example, the default is to open the linked page in the same window/tab that I am currently in. I can also open the link in a dialog window, a separate window that allows the user to simply close the window to return to my site without losing their place. The third option opens the link in a new tab in the current browser instance. I generally recommend In-page navigation for links to pages within my site and dialog or new tab to launch linked pages in other sites. The theory behind this rule is that I want to minimize the number of open windows while still giving the user the option of quickly returning to the page containing the link.

If I were to click the Save button at this point to save my link and then display the Promoted List on a page, it would look like the following figure:


This view is essentially the standard list view that I am familiar with from past versions of SharePoint. However, if I edit the page and then edit the web part, I can change the Selected View in the web part dialog to the Tiles view as shown in the figure to the right. The standard list view is the All Promoted Links view. Of course, just like any other list, I can create my own customized list view based on either the Tiles or All Promoted Links views.

So what does the Tiles view look like? It reminds me of the Modern (Metro) interface style of Windows 8 as shown to the left. Note however that the top area is blank (or in this case a green square) while the title appears at the bottom in a darker green square. Furthermore, when I hover over the tile with my mouse, the title moves up to the top of the tile and the dark green color covers the entire tile.

The reason the top square is blank is that I did not include all of the available list options. Specifically, I did not include a background image to use with the link. To do this, I first have to capture the image I want to appear on the tile and save it to my SharePoint site. For this example, I’ve taken the school logo and loaded it into my Site Assets library.

After loading the image into my Site Assets library, I can return to the Test_Schools promoted list and edit the entry for Edom Middle School adding the Background Image Location value as shown in the following figure.

I can also add a description for this image. However, like the description for the Link Location, this information does not appear on the final tile view.

Now when I return to my page to view the pilot schools, the link looks like the following figure.

Hover —>

In a similar fashion, I can add the other tiles for the other pilot schools in the project using a logo when available for each school. After saving my work and displaying the page, I might see something like the following:

The first thing to note is that the Promoted Link list only appears as a single line of icons across the top of the page (or content area). It does not have the ability to wrap to multiple lines to create a grid of icons like I see in the All Site Content page. However, I do get a pair of navigation buttons in the upper right corner when necessary that allow me to scroll through the icons to find the one I want to click on.

Another issue is that the default order of the icons appears to display the most recently added links starting from the left to the right. If I want to change the order of the icons, I must use the Order property of the list to define the order that I want the links to appear.

That’s about all you need to know about the Promoted List other than that you may want to resize your background images to 150×150 pixels before uploading them to improve performance.

C’ya next time.

Is There A Better Way?

Today I want to talk about a news item I pulled off the Internet.  It should come as no surprise to anyone that the number of U.S. citizens spending time in our prison population has been growing.  Today U.S. Attorney General Eric Holder said in a speech in California that our prison system is “outsized and unnecessarily large”.  Most people don’t realize just how out of control our prisons have become though.  According to statistics on Wikipedia (take that for what it is worth) the number of adult in our U.S. prisons was 2,266,800 or about 0.7% of the adult population.  When you add in 4,814,200 more adults that were on probation or parole, the total comes out to 2.9%, almost 3% of the adult population.

The real surprising number is that the total has increased by 4 fold  from about 500,000 in 1980 while during the same time, the population did not increase by 4 fold.  There is a ton of additional statistics on the Wikipedia site (http://en.wikipedia.org/wiki/Incarceration_in_the_United_States) but my point is not to go through the statistics.  Rather, my point is more aligned with Eric Holder’s question as to whether we are incarcerating more offenders just to simply “warehouse and forget” about them or if we should have a judicial system that focused on the cause of crime, the just punishment for the crime, the deterring of criminal offenses, and the rehabilitation of those found guilty.  Mr. Holder also states that we cannot prosecute or incarcerate our way to a safer nation.


We cannot simply turn every criminal loose just because we want to keep the size of our prisons down.  Without punishment as a deterrent, what would cause people to hesitate before committing a crime?  Some say that many criminals are not deterred by punishment anyway.  So maybe, we as a nation need to focus on determining why people turn to crime.  Is it due to lack of good job opportunities?  Are those opportunities reduced because our education system has resulted in workers that are inferior to foreign workers?  What affect, if any, does the challenge to our work ethics by all of the government programs that promise free money, free services, and free food, destroy the very fabric of our society?

Then once a person is labeled as a criminal, where is the rehabilitation?  Where is the opportunity to start over?  Would you want a criminal living next door to you or your children?  Why is the immediate reaction, ‘No’!  Is it because we all know that criminals in prisons are not rehabilitated as much as prisons are a training ground to how to commit other crimes?  Even if a person was sent to jail for a non-violent offense such as unpaid debt, stealing money from their employer, or simple shoplifting which they did because it was the only way they could provide for their family, what can they do when they get out?  They now have a record.  That record is worse than a scarlet ‘A’ on their chest.  No one will hire them.  No one will trust them.  No one will give them a chance to show that they will work an honest job for a fair pay that will support their families.  Where is the rehabilitation?

Again, not everyone charged with a crime should be released.  Mass murderers, kidnappers of children, and dozens of other crimes sometimes need to be warehoused and forgotten.  However, how much better for everyone would it be to provide non-prison sentences that may include community service, work projects, or other activities be for low-level nonviolent crimes?  Perhaps even having minor offenses expunged from the records for some low-level crimes to give that person a chance at a new law-abiding life.

It will be interesting to see if this speech made to the American Bar Association’s House of Delegates will make a difference or whether like so many other grand speeches it falls on deaf ears. Just for the record, the United States has the highest incarceration rate per 100,000 people of any other country in the world.  That number is 716.  Even Cuba is only 510, Russia is 484 and England is a mere 148.  The current estimate is that it costs around $22,000 a year to keep a person in prison.  According to CNN, we spend an average of only $10,615 per student per year to provide public education.  Granted, variations exist by state and of course there are far more students than criminals.  However, it is not the total spent, but the amount spent per person that seems out of line.

Anyway, I just wanted to give you something to think about thanks to Eric Holder’s speech.

C’ya next time.

Using Microsoft Word to Write Your Blogs

The first thing to know about creating a blog in SharePoint 2013 is that blogs are separate sites, not just web parts on an existing site. If you haven’t created a blog site yet, you need to go to the Site Contents page for the site that will be the parent site for the blog. In the second half of the Site Contents page under the Subsites section, click New Subsite. On the following page, supply a title and web address for your site and then select the Blog template from the Collaboration tab in the Template Selection group. I recommend inheriting parent permissions initially. You can break inheritance and create custom permissions later. You can also decide on this page whether you want the blog name to appear in the top link bar. If not, you must supply another way to navigate to the blog site on your own. The following image shows a completed blog site form.

After creating the blog site, SharePoint automatically opens it for you and you should see a welcome blog entry. In the left navigation area, SharePoint also provides a set of four categories which you can use to group your blog entries. Of course you can easily remove these and add categories of your own that may be more appropriate for your blog site. SharePoint also keeps an Archive by month of your blogs to make it easy for readers to go back in time to find an older blog entry.

In the top-right column SharePoint displays a set of blog tools that allow you to do everything from create a new post, manage existing posts, manage comments to the posts, manage the categories used to group posts, change the post layout, and even launch a blogging app. What is a blogging app you ask? Let’s click the link and find out.

After a few seconds, a dialog appears asking you to enter information, a URL, to register your SharePoint blog account. Fortunately, you do not have to figure out what the URL is as SharePoint automatically fills in the Blog URL text box for you. Simply click the OK button to continue.

You should get another dialog informing you that the account registration for the blog site was successful. In the title bar of this dialog, you get your first clue that the blogging app referred to by SharePoint is Microsoft Word.

At this point Microsoft Word opens, not with a blank page, but with a blank blog template as shown below. Since you are creating the blog entry from within Word, you have all the standard

Word functionality from formatting your text to inserting tables, pictures, hyperlinks, and more. In fact, if you want to create a blog with images in it, using Microsoft Word can actually save you time.

How? Normally you have to upload any images you want to use on a web page into an image library before you try to display the image on a page. That is because the image must be on either the same server as SharePoint or on another web server that your SharePoint Server can reference. SharePoint will not display an image on a page that only exists on your local computer. After all, what would happen if your computer was not turned on or perhaps not even connected to the network anymore? Therefore, you have to first upload the image to a picture or image library. Then you need to capture the shortcut link to the image to the clipboard. Returning to the page where you want to display the image, you first must edit the page, then insert an image web part or perhaps a content editor web part into which you can paste the link for the image now stored in a library. That is a lot of work.

However, using Word, you can simply create your blog entry and include the image directly in the text as shown below:

Next, click the Publish option from Publish dropdown menu in the Blog group of the Blog Post ribbon shown to the right.

Finally, returning to my blog site and refreshing the page, I see my most recent blog entry at the top of the blog that displays not only the blog text, but also the embedded image.

What did SharePoint do with the image? Open the Site Contents and look in the Photos library. More next time on advanced permissions for blogs and how to create a private vs. public blog site.

As an interesting side note, I can also use Microsoft Word to publish to other common blog sites such as WordPress. To do this, begin a new document using the Blog Post template. If I did not see this template, I could download it from Microsoft’s template site by using the search box to search for: Blog Post.

After opening a new document with the Blog Post template, click the Manage Accounts button in the Blog Post ribbon.

In the Blog Accounts dialog, click the New button to display the New Blog Account dialog. The dropdown for the blog provider already contains several common blog sites. If my provider is there, simply select it (i.e. WordPress). If not, I would select: Other and Word prompts for additional connection information.

For the common providers, Word already knows the Blog Post URL and will automatically supply it. I will have to enter my specific blog URL along with my user name and password and any other information my provider requires.

After clicking OK, Microsoft Word attempts to connect to the blog provider. If all the information is correct, I should see the following dialog after a few moments.

I can now set the new blog provider as my default provider.

Or I can simply change the account when I begin a new blog post. Note in the following figure that I can also insert a category for my blog. However, I do not have the ability to select multiple categories like I may do when logged into my actual blog provider.

As with a SharePoint blog, the main advantage of creating my blog in Word, especially if I have embedded pictures like this example and all of my technical blog posts, is that Word handles uploading not only the blog, but also the images to the appropriate media library. (You are looking at the final result.)  I don’t have to first upload the text, then upload all the images and finally reinsert the images back into the text.  This is a huge time saver.  Using the dropdown for the publish button in Word, I can either publish immediately, or I can publish as a draft. Publishing as a draft allows me to connect to my provider and select additional categories for my post as well as schedule the publication for a different time.

That’s it for this time. C’ya.

Going Digital

Are you ready to go all digital?  Is your workplace ready?  Over the past several years, various organizations have taken the plunge to go all digital.  The publishing industry has probably been the most successful.  I don’t think I bought more than 1 or 2 paper books in the last year or so but I have bought at least two dozen electronic books and I carry them around with me all of the time.  I also use the electronic version of best sellers from our local library which I can check out on-line rather than having to physically go down to the library.  What a fantastic resource.

Newspapers and magazines are starting to go the same way.  At first they offered both printed and digital versions of their publications, but now some of them are cutting out the printed versions.  For example, the June 24th issue of Information Week, a free magazine that many of us in the IT industry have subscribed to for years, was the last paper version of the magazine.  In fact, for the last year I have only subscribed to the electronic version of SQL Server Magazine because it is much more convenient to carry around not only the most current issue, but the complete last year of issues on a tablet.  Furthermore, there are interactive features available in the digital versions that do not and can not exist on the paper magazines.  Stories can be linked to similar stories, to further information, and to vendors of products described in the article.  Advertisements can be interactive with links to videos to demo the product.  You can even email the editors and authors sometimes directly from the digital pages of the magazine.

Many of the national newspapers have digital subscriptions available today.  Even our local newspaper started offering digital subscriptions.  For those people who like to clip coupons to save money at the store, I’m starting to see more of the coupons go digital as well so that you only have to display the barcodes on your smartphone to get money back from restaurants, department stores, and other places.

I use to do all of my writing first on paper and then type in what I created for editing and formatting.  There was always something about the feel of a smooth flowing pen on paper as you write that created a sense of enjoyment.  Ok, maybe that is a little weird for some of you, but anyone who is a writer and began their love of writing more than 10 years ago knows exactly what I mean.  However, now I’ve switch to writing all of my blogs, documentation, newsletters, etc. directly in a word processor.  It is not always Microsoft Word either.  Any word processor will do because the basic functions are the same.  I learned how to type in high school and I find that I can type much faster than I can write the same text out freehand.  Thus I save time not having to first write it down and then type it up.  I still haven’t gotten use to writing on a tablet and I much prefer a standard keyboard, but hey, who knows.  After all, I learned how to use a mouse with my left hand so that I could write with my right hand without having to keep switching what I was holding.  (Yes, I occasionally would try to navigate around the screen with my pen and write with my mouse so it was the best solution I could come up with.)

I guess the only problem I have with writing electronically is the ‘darn’ spell correction in most word processors these days.  Sometimes when I misspell a word, the program tries to correct it with what it thinks I was going to say, but while it gets a correctly spelled word, it is often the wrong word.  Therefore, I sometimes send out text with what looks like stupid mistakes.  Yes, I know you are suppose to proof read your document before sending it out, but sometimes I even miss the most obvious wrong words and then have my friends question how smart I really am.

Anyway, my point is that soon you will be able to forget about recycling paper, because paper as we know it today will become ancient history like impressing characters into a wet clay tablet.  I can hear the trees celebrating.  But equally important will be the fact that we no long have to store paper.  Physical paper takes a lot of space and it weighs quite a bit.  (Pick up two copies of War and Peace and use it for your arm exercises.  It may be cheaper than a set of barbells.)  You can easily build a large library at home with practically no required space.  You can also carry that library with you and read on the commuter train to work or while waiting in line somewhere or even during that dull staff meeting that you have to go to every week.

Yes, sometimes there is just something comforting about holding a real book with real pages.  But with the improvements in Kindles, Nooks, iPads and even most smart phones, I find that I miss real books less and less.

I’ve always been a Star Trek fan and many of the things used in Star Trek have become possible over the last several decades since the original series.  Ok, we don’t have transporters or warp drives yet (although wouldn’t that be fun), but the flip phone style communicators exist as do the pads that were used and are now called tablets or even the iPad.  Some of the medical devices for imaging a person’s body to diagnose disease exist or are on the horizon.  However, never once in all of the episodes did I see Kirk, Picard, or Janeway need to fall back on sheets of paper for information.  If you haven’t tried it, try going a week without using paper, or at least a day.  At the very least cut back on your paper usage and rely on digital versions of the information.  It is not as hard as you may think.

C’ya next time.

Create a Matching Policy – 2

Last week I left you with the thought that an exact match can be either a prerequisite or just one of several domains used to match records.  The purpose of using an exact match as a prerequisite is that the domain is so crucial to determining whether two or more records match that if the value of this domain is different between two records, there is no possibility for the records to match.  Therefore, there is no need to check any of the other domains when this prerequisite domain is different.  This may seem counter-intuitive at first, but we must remember that we are not looking for unique records here, but matching records and eliminating records by using a prerequisite will speed up the matching process.


While you can have more than one prerequisite domain, not every domain that you are trying to match on should be a prerequisite field.  In fact, DQS requires that you have at least one domain which is not a prerequisite.  Fortunately, fields like business name and business address should not be prerequisites because there is a high probability that these could be misspelled or formatted differently.  For example, Adani & Baez Computers Inc. is probably the same business as Adani and Baez Computers Inc. and the same as Adani & Baez Inc. even though they are not exact matches.  If we used the Business Name field as a prerequisite exact match, these three names would not be considered a match.  A similar argument could be made for Business Address and even Business City.

Therefore, we would add these additional domain elements in our matching rule.  Each domain element other than any prerequisite domains must have a weight factor.  The weight factor determines how important matching domain values are in determining whether two records are a match.  Initially, you may have no real preference for how to weight the additional domain values.  A good place to start however is to make all the domain weights equal.  Therefore, if you have five additional domains, make each one 20%.  Note the total of the weighting factors must add to 100%.


After evaluating all the domains for each set of records, DQS calculates a number between 0 and 100% to represent the chance that the two records are a match.  Of course, a score of 100 would indicate an exact match across all the domains used in the rule.  Of course, most pairs of records will probably have a value less than 100.  But how much less than 100 still indicates a match?  DQS uses a Minimum Matching Score to filter the potential matches.  For reasons beyond the scope of this blog, DQS places a minimum score limit of 80% on matching records.  While I can change the Minimum Matching Score to any value between 80 and 100, I cannot set it any lower than 80.  This value can be found below the Rule Name and Description as seen in the following figure.

Another option you must set is whether a record can appear in only a single matching set of records or whether it can belong to more than one set of matching records.  The first option is called non overlapping clusters and the second option is called overlapping clusters.  With non overlapping clusters, each record appears in only one cluster or is considered a unique record.  This is the option I would usually select if my goal was to eliminate matching records.  However, using overlapping clusters might be a good way to analyze for other relationships between records.  For my analysis here, I will use non overlapping records.


Next, I click the Start button shown in the previous figure to begin the analysis.  This process can take a substantial amount of time if I am working with a large database table.  Therefore I recommend creating a smaller representative table to test out my matching rule definition before running the final rule against my full table.  After the matching process completes, a grid appears at the bottom of the screen.  In this grid, which is filtered on the matching records, you can see how DQS has grouped similar records based on the matching rule.  Each matching groups is called a cluster and is given a cluster id that has the same value as the first record id added to the cluster.


Clicking on the Profiler tab provides some interesting results.  In the figure below, I can see that my test table had nearly 8000 records.  In the grid to the right, I can see the number of new and unique values for each of my domains.  New values are domain values that I may want to go back to add into the knowledge databases for the domains if they truly are unique and not just bad data values.  The Unique column tells me how many of the 8000 records had unique domain values.  I can see that there are only 5129 unique business names.  That means that I should have nearly 3000 potential duplicate records based only on the Business Name.  Looking at Business Address, there are 6683 unique addresses.  This tells me that there are at least some business records with the same business name, but different addresses.  This is entirely possible if the business has more than one location and may not be an error.  On the other hand, of 8000 records, there are only 224 unique city values, 32 unique states, and 465 unique zip codes.


The low rate of unique values for these three fields tells me that their contribution in determining matching records is low.  At the same time, it makes each of these domains candidates for an exact match or even a prerequisite match.  In fact, that is why I selected Business States to be an exact prerequisite match.

If I previously cleaned my data elements to remove inconsistent abbreviations and spelling problems, I would want to look at the fields with the most unique values to be potentially more important when determining matching records.  However, just because a domain has a high uniqueness value, does not make it a good matching value.  Take Business Phone as an example of a poor domain to determine matching records.  Because a business could have more than one phone number at a given address, different phone numbers do not necessarily mean that the records are not a match.  Even a different address may not be as important as having the same business name.  Therefore, I use the uniqueness values as a guide, not an absolute to determine how I might want to reassign weights to each domain in the rule instead of weighing them all equally.

Another thing to consider is to examine each cluster and determine why some of the matching scores are less than 100%.  In the figure below, I have highlighted a record with only a 96% match score.  To find out why this score is not 100% I can right click on the row to display the details.


The Matching Score Details dialog shows the two records that are being compared at the top with the overall matching score.  In the second grid, I see a grid of the other domain with their contributing score.  In the figure below, I can see that phone number score was only 79%.  However, looking at the two phone number I can see that difference was most likely a key entry mistake.  In any case, the 79% score on this domain is only worth 20% of the total matching score.  Since the other domains are exact matches, I have:

0.2 * 100 + 0.2 * 100 + 0.2 * 100 + 0.2 * 100 + 0.2 * 79 = 96

But perhaps more importantly, this tells me that the phone number domain may not have been cleaned.  If I do not have a reliable way to do this without spending a lot of time or money, I may simply want to deemphasize the importance of the Business Phone domain and rerun the matching rule.


Finally when I am done, I can proceed to the next page of the defining the matching policy where I will run the rule one last time.


Clicking Finish will allow me to publish the matching rule back to the Knowledge Base as shown in the following dialog.


After the publishing process completes, the final dialog informs me of its success as shown below.


I now have a matching policy saved with the business address domains that I can use in a project with the full database table.

C’ya next time.