So, here is how the rest of the Association Rules viewer works.

Once the viewer is loaded, the first call is something like:

CALL System.Microsoft.AnalysisServices.System.DataMining.AssociationRules.GetStatistics(’Customers’)

The single parameter of this stored procedure is the mining model name.  The result is a one-row table containing the following columns:

(more…)

- in the browser: keep using http://sqlserverdatamining.com/cloud

- in Excel (add-in): go to the Connections button in the Analyze (in the Cloud) ribbon and type the following URL: https://131.107.181.99/CloudDM/TATServices/

With this change, you Excel add-in will once again use HTTPS to secure the data transmission

Thanks for your patience!

- For the web interface, use the http://www.sqlserverdatamining.com/cloud URL

- For the Excel add-in, please change the services connection URL. To do that, click the Connections button in the "Analyze (in the Cloud)" ribbon and change the destination URL to

http://131.107.181.101/CloudDM/TATServices/

NOTE: This temporary solution does not support SSL. Your data is transmitted in clear

I’ll post here an update as soon as the servers are up again

Soon enough, there will be some content there:

- various data mining related postings from Jamie and me — all the product related postings on this blog will be replicated on the book’s blog . With a bit of discipline, all the postings will be tagged with the relevant chapter number

- errata (well, hopefully that won’t be the main topic of the blog )

- any new downloads or other information that may be relevant for the readers

The new version of the SQL Data Mining book is finally available, at least at Amazon. If you are currently SQL DM user, you have an opportunity to get a free autographed copy  by filling out a short survey about the way you use SQL Server Data Mining. For more details about the survey, check Jamie’s blog post.

More details about the book here: Data Mining with SQL Server 2008

A new SQL Server Data Mining tool is available now for you to use directly from the browser — the Prediction Calculator.

The feature (already present in the 2008 version of the Data Mining Add-ins) is a scorecard generator. The tool learns from your data and produces a simple calculator which can be used to compute a score (between 1 and 1000), score which describes the likelihood of a certain event. If you are not familiar with this kind of tools, think of the personality tests that appear in various magazines or about the FICO score.

When you launch the Prediction Calculator, you need to select a target column and a target value (or range). If you target column has categorical values, then you will need to specify exactly your target value. Example: if you want to predict whether some computer will crash in the next month, you will need to select, say, the “Yes” value of the “Will Crash in the Next Month” column. If your target is numeric, you can specify a range. For example, you can use the prediction calculator to figure out how likely it is that your insurance costs will grow between 20% and 50% in one year.

The result, depending on the columns in your data set, looks more or less like below. As you select different values for various columns, the score goes up or down. If it exceeds a certain threshold then you get a positive prediction, otherwise a negative prediction.

For example, try to play with the calculator below to figure out whether you are a likely Bike Buyer or not:

How it works: each of the attributes gets a score, depending on the state. For example, having 0 cars may score 100 points, while having 2 cars may score 0 points (people with many cars are less likely to purchase a bike). The points are added up and, if the total score (always between 0 and 1000) exceeds a certain threshold, then the prediction is positive, otherwise negative.

The Prediction Calculator tool allows you to determine the threshold that maximizes your “profit”. In the lower part of the prediction calculator result (not visible in the embedded snapshot above) you can specify the costs associated with a False Positive (Type I) or False Negative (Type II) error. You can also specify any profits deriving from a correct positive or negative prediction.

The tool computes the total profit generated by various score thresholds based on your inputs and produces a set of diagrams, one for profit and one for cumulative costs, like below:

You can see in the profit diagram that the profit is maximized for a score around 450 or 500. The costs are also minimized in that area.

The tool detects the scorecard threshold that optimizes the profit (and includes it in the page).

Once you produce and tune a prediction calculator, you can embed it in your HTML page, just copy and paste the HTML fragment — the Bike Buyer calculator above is such a fragment.

Around May this year I practically entered some sort of sabbatical: 3 months to work on an incubation project of my choice (yes, the Microsoft SQL Server organization does this kind of things! if it sounds appealing,  check out our recruiting site or, even better, contact directly our SQL Server Data Mining recruiter, Melsa Clarke - melsac AT microsoft DOT com). With some help from Jamie, various management levels and some nice guys in the SQL Server Data Services team, I gathered the infrastructure for a Software as a Service incubation and set up a web incarnation of the Table Analysis Tools add-in for Excel.

Now it is up and running. The entry page is at sqlserverdatamining.com/cloud, so if I got you bored already and you don’t want to read the rest of this stuff, go ahead and browse that page.

What it is

TAT Cloud is a set of canned data mining tasks that you can use without having SQL Server installed on your machine.  It consists of encapsulations of some common data mining problems, such as detecting key influencers, forecasting, generating predictive scorecards or doing market basket analysis. The tasks can be executed directly from your browser: just go to the web page, upload your data (in CSV) format and run a tool from the toolbar. Even better, the tasks can be executed directly from Excel. For this, however, you will need to have Excel 2007 and install an add-in which can be downloaded from here.

All tasks work on a table (Excel table or a table in CSV format that you upload to the web interface). All tasks produce reports that can be used to learn more about the analyzed data.

Here is a complete list of features:

• - Analyze Key Influencers: it detects the columns that impact your target column. It presents a report of those values in other columns that correlate strongly with values in your target column.
• - Detect Categories (clustering, for data miners) — identifies groups of table rows that share similar characteristics. A categories report is generated, which details the characteristics of each category
• - Fill From Example — to some extent, similar to Excel’s Autofill feature: it learns from a few examples and extends the learned patterns to the remaining rows in the table
• - Forecasting — analyzes vertical series of numeric data, detects periodicity, trends and correlations between series and produces a forecast for those series
• - Highlight Exceptions — finds the interesting (or unusual, or out-of-ordinary) rows in your table
• - Scenario Analysis — What-If and Goal-Seek tools based on a probabilistic model built on top of your data.
• - Prediction Calculator — a tool for generating prediction scorecards
• - Market Basket Analysis — analyzes transaction tables to identify groups of items that appear together in transactions

All features work from Excel. However, not all the features are implemented in the web interface. They will show up eventually!

A really nice presentation of how to use the tools was written by Brent Ozar on SQLServerPedia. Thanks, Brent!

How it works

Your data (CSV file or Excel spreadsheet) is uploaded to the web service site. There, Analysis Services crunches it and produces the reports you get either in the browser or in a different spreadsheet. The data and the mining model used for analyzing it are  deleted immediately after processing. If something bad happens and your session does not conclude successfully (fancy wording for "if it crashes") then both data and models are removed automatically after 15-20 minutes.

You will notice in the Excel add-in (as well as in the web interface) a strange IP address, 131.107.181.99 — this is IP of the service. It will be changed to something cleaner very soon.

Excel uses HTTPS to connect to the service, in an effort to protect your data. However, you should not use this technology on sensitive data.

What it is not

This is not an official shipping Microsoft technology. It is actually less than even a beta. It may crash, it may produce incorrect results, it may be shot down at any time. BTW, I would appreciate if you posted note here or on the Microsoft data mining forums if this happens (particularly about crashes, I should probably know if it gets shut down)!

What next

Well, try the tools: sqlserverdatamining.com/cloud . I recommend downloading the Excel 2007 add-in, rather than using the web application, as it is more functional. Post any questions you might have on the Microsoft data mining forums. And check out this blog periodically for announcements (typically new additions to the web interface) or for more details on how this stuff works

Today I discovered a very nice Analysis Services client tool. Produced by SQLMinds, the tool seems a great addition to Analysis Services. It provides many components, among them a performance tuning service, an OLAP cube browser for the web and a very nice web front end for data mining, the DM Companion tool which can be launched at http://x32.sqlminds.com/dmcompanion.

So, here are a few really nice things about the DM Companion tool (BTW, a fully working demo is running at the aforementioned URL).

The tool works to some extent like a Data Mining -specific SQL Server Management Studio for the web. Therefore, it allows you connect to the server of your choice (through the Analysis Services HTTP pump). The demo seems to allow anonymous connections (pretty safe  as all the interactions offered by the tool are read-only).  Next you get to chose your AS catalog and the mining model you want to use.

For each model you have the option to browse the content or execute predictions.

The prediction feature provides a nice interface for defining singleton predictions, as you can see below:

The interface supports specifying multiple nested table keys as input, so the tool can perform associative predictions as well. It reminds me of the XMLA Thin Miner sample running on SQLServerDatamining,com, however, it looks much better and is nicely integrated with the rest of the application. While prediction functions do not seem to be directly supported, the application is able to predict a cluster for clustering models.

The model browsing features are really nice. Analysis Services includes a set of sample web viewers for Naive Bayes, Trees and Clustering. This application provides some seriously better looking viewers for these algorithms, and extends the suite at least for Neural Networks (a really nice viewer), Sequence Clustering and Association Rules.  The DM Companion viewers offer all the features in the sample viewers, with a nicer implementation which uses AJAX and has better graphics, plus a solid set of new features, the most spectacular being the interactive dependency net browser and the pie chart visualization for decision trees, which you can see below.

Overall, DM Companion looks like a really nice tool for sharing your data mining implementation on the web

Recently, trying to make sense of some prediction results, I realized that the DMX NATURAL PREDICTION JOIN statement has some obvious and some not so obvious dangers, which deserve some clarification.

To summarize, the PREDICTION JOIN operator is used to apply patterns from a trained mining model against new data. The general syntax is

SELECT EducationLevel

FROM MyModel

PREDICTION JOIN

OPENQUERY(MyDataSource, ”SELECT Height, , HairColor, ShoeSize FROM CustomerInfo’) AS T

ON

T.Height = MyModel.Height AND

T.HairColor = MyModel.HairColor AND

T.ShoeSize= MyModel.ShoeSize

In the standard PREDICTION JOIN statement, the ON clauses specify how columns from the new data are to mapped to the input columns of the mining model.

The statement supports a simplified format:

SELECT EducationLevel

FROM MyModel

NATURAL PREDICTION JOIN

OPENQUERY(MyDataSource, ”SELECT Height, HairColor, ShoeSize FROM CustomerInfo’) AS T

In the simplified format, the columns from the new data are mapped to the input columns of the mining model by name. Much easier to type, resulting in a more readable statement!

Now, the natural mapping may end up in three different scenarios. In the first scenario, all the model’s input columns find a corresponding data column in the input. The mapping is performed correctly and everything works as expected. Nothing to blog about in this case!

The second case: some of the inputs are mapped correctly to model columns, some are not, although you would expect them to be mapped correctly. More than a simple user error, this situation is actually “facilitated” by the way the BI Dev Studio works. Assume you start with a table containing 4 columns: Height, HairColor, ShoeSize, EducationLevel and use BI Dev Studio to create a mining structure and a model. BI Dev Studio will generate more friendly names for the columns and your mining model ends up with the structure below:

MyModel

(

Height

[Hair Color],

[Shoe Size],

[Education Level]

)

(more…)

Bill then follows up with a set of tutorials on pivot tables and wraps up with videos on using the Table Analysis Tools from the SQL Server Data Mining add-ins for Excel. His approach is accessible, his style is captivating, his examples are meaningful and very well explained.

Therefore:

1. If you have SQL Server 2005 or 2008 CTP these tutorials give you a great start for BI using Excel and Analysis Service.

2. If you do not have SQL Server, get a free evaluation from http://www.microsoft.com/sql/downloads/trial-software.mspx, then go back to the previous point

3. You don’t have SQL Server, don’t want to try it out, don’t care about the data mining part — fine, just watch some of the tutorials to see what you miss!