http://www.ailerontechnology.com/
more to come ….
AileronTechnology.com
Phone: 716-997-2399
Email: Info@AileronTechnology.com
John Indelicato – Founder
Our daughter Victoria Stasiewicz graduated today and with your husband Brandon Stasiewicz. Theresa and I are very proud of Victoria and her incredible accomplishments.
The primary take away from this article will be that you don’t start your Machine Learning project, MDM , Data Quality or Analytical project with “data” analysis, you start with the end in mind, the business objective in mind. We don’t need to analyze data to know what it is, it’s like oil or water or sand or flour.
Unless we have a business purpose to use these things, we don’t need to analyze them to know what they are. Then because they are only ingredients to whatever we’re trying to make. And what makes them important is to what degree they are part of the recipe , how they are associated
Business Objective: Make Desert
Business Questions: The consensus is Chocolate Cake , how do we make it?
Business Metrics: Baked Chocolate Cake
Metric Decomposition: What are the ingredients and portions?
2/3 cup butter, softened
1-2/3 cups sugar
3 large eggs
2 cups all-purpose flour
2/3 cup baking cocoa
1-1/4 teaspoons baking soda
1 teaspoon salt
1-1/3 cups milk
Confectioners’ sugar or favorite frosting
So here is the point you don’t start to figure out what you’re going to have for dessert by analyzing the quality of the ingredients. It’s not important until you put them in the context of what you’re making and how they relate in essence, or how the ingredients are linked or they are chained together.
In relation to my example of desert and a chocolate cake, an example could be, that you only have one cup of sugar, the eggs could’ve set out on the counter all day, the flour could be coconut flour , etc. etc. you make your judgment on whether not to make the cake on the basis of analyzing all the ingredients in the context of what you want to, which is a chocolate cake made with possibly warm eggs, cocunut flour and only one cup of sugar.
Again belaboring this you don’t start you project by looking at a single entity column or piece of data, until you know what you’re going to use it for in the context of meeting your business objectives.
Applying this to the area of machine learning, data quality and/or MDM lets take an example as follows:
Business Objective: Determine Operating Income
Business Questions: How much do we make, what does it cost us.
Business. Metrics: Operating income = gross income – operating expenses – depreciation – amortization.
Metric Decomposition: What do I need to determine a Operating income?
Gross Income = Sales Amount from Sales Table, Product, Address
Operating Expense = Cost from Expense Table, Department, Vendor
Etc…
Dimensions to Analyze for quality.
Product
Address
Department
Vendor
You may think these are the ingredients for our chocolate cake in regards to business and operating income however we’re missing one key component, the portions or relationship, in business, this would mean the association,hierarchy or drill path that the business will follow when asking a question such as why is our operating income low?
For instance the CEO might first ask what area of the country are we making the least amount of money?
After that the CEO may ask well in that part of the country, what product is making the least amount of money and who manages it, what about the parts suppliers?
Product => Address => Department => Vendor
Product => Department => Vendor => Address
Many times these hierarchies, drill downs, associations or relationships are based on various legal transaction of related data elements the company requires either between their customers and or vendors.
The point here is we need to know the relationships , dependencies and associations that are required for each business legal transaction we’re going to have to build in order to link these elements directly to the metrics that are required for determining operating income, and subsequently answering questions about it.
No matter the project, whether we are preparing for developing a machine learning model, building an MDM application or providing an analytical application if we cannot provide these elements and their associations to a metric , we will not have answered the key business questions and will most likely fail.
The need to resolve the relationships is what drives the need for data quality which is really a way of understanding what you need to do to standardize your data. Because the only way to create the relationships is with standards and mappings between entities.
The key is mastering and linking relationships or associations required for answering business questions, it is certainly not just mastering “data” with out context.
We need MASTER DATA RELATIONSHIP MANAGEMENT
not
MASTER DATA MANAGEMENT.
So final thoughts are the key to making the chocolate cake is understanding the relationships and the relative importance of the data/ingredients to each other not the individual quality of each ingredient.
This also affects the workflow, Many inexperienced MDM Data architects do not realize that these associations form the basis for the fact tables in the analytical area. These associations will be the primary path(work flow) the data stewards will follow in performing maintenance , the stewards will be guided based on these associations to maintain the surrounding dimensions/master entities. Unfortunately instead some architects will focus on the technology and not the business. Virtually all MDM tools are model driven APIs and rely on these relationships(hierarchies) to generate work flow and maintenance screen generation. Many inexperienced architects focus on MVP(Minimal Viable Product), or technical short term deliverable and are quickly called to task due to the fact the incurred cost for the business is not lowered as well as the final product(Chocolate Cake) is delayed and will now cost more.
Unless the specifics of questionable quality in a specific entity or table or understood in the context of the greater business question and association it cannot be excluded are included.
An excellent resource for understanding this context can we found by following: John Owens
Final , final thoughts, there is an emphasis on creating the MVP(Minimal Viable Product) in projects today, my take is in the real world you need to deliver the chocolate cake, simply delivering the cake with no frosting will not do,in reality the client wants to “have their cake and eat it too”.
Note:
Operating Income is a synonym for earnings before interest and taxes (EBIT) and is also referred to as “operating profit” or “recurring profit.” Operating income is calculated as: Operating income = gross income – operating expenses – depreciation – amortization.
You get what you inspect, not what you expect!
This post will combine business needs with current capabilities for scoring data sets/tables large or small consistently. The methodology and approach we’re going to discuss was implemented with AWS Athena and QuickSight.
A few years ago we put together an article on simple data profiling in a simple sql environment.
We’ve had the occasion to use this method at every client we,ve been to in the last decade.
A very important concept in big data, data mining IT integration and analytics projects is the idea of analyzing, surveying profiling your data and knowing what you’re going to have deal with before, you have to deal with it in a one off fashion
Why Score data , first a discussion of the business value of why this matters.
Let’s assume that you have a business problem, a business question that needs answers and you need to know the details.
There is a predisposition today to call a data scientist.
The problem is you don’t need to know the science behind the data, you need to know, the information that can be derived from the data
Business is based on Concepts.
People thinks in terms of Concepts.
Concepts come from Knowledge.
Knowledge comes from Information.
Information comes from Formulas.
Formulas determine Information relationships based on Quantities.
Quantities come from Data.
Data physically exist.
Data needs to be deciphered into information in order to apply formulas to determine relationships and validate concepts.
My point is proving these low level concepts is probably worthless to the business and maybe even destructive, unless they are associated with a actual set of Business Goals or Measurements and they are linked directly with understandable Business deliverables.
And moreover prior to even considering or engaging a data scientist or attempt data mining you should process your datasets through a ‘Scoring” process.
When we say linking we mean creating an “information value chain” relating Business Goals to Business Questions and breaking them down(decomposing) them into the following:
There are two things necessary to Score a data set or file universally.
File Scoring: Overall combined score of all columns based on standards data profiling measures(count of nulls, minimum value, maximum value, mode value, total records, unique count, etc…)
Column Scoring: measures for each column or domain including but not limited to frequency of distinct values, patterns.
These two measure will combine to provide a “Score” and the necessary detail to analyze the results in detail and predict corrective methods and fitment for possible feature engineering for machine learning models as well as general data quality understanding.
Back to how we got here:
Our projects were standard data warehouse, integration, MDM old school projects.
Nonetheless they required a physical or virtual server, a database and usually they were limitations on the amount of data that could be profiled due to these constraints.
However the problems we faced then and that our clients faced are the same problems they face today only exponentially bigger.
Fast forward 10 years. No Database, No Server, No Limitations sizewize, and pay for only what you use.
This has only been possible since last year it was not due virtual machines, or big data.
It is due to the advent of serverless sql, as well as other serverless technologies
We created a few diagrams below to try to illustrate the concepts both from a logical and physical perspective
As you can see below most projects begin with the idea of ingesting data transforming it or summarizing it in some manner and creating reports as fast as possible.
What I find most often is that the source analysis, the matching of source data content contextualization to business requirements and expected outcome is usually last, or discovered during the build. This requires the inclusion and creation of a meta-data mart and the ability to provide scoring and analysis on the contextual of data agnostic way, put simply scoring the data or content , Purely on the data, prior to trying to use it for business intelligence, reporting, visualizations and or most importantly data mining.
As you can see below we need to insert simple processes Conceptionally the same as outlined in my original blog in order to gain information and insight and predictability in providingi business intelligence output , visualization or featuring engineering for possible data mining.
So let’s explore what we can accomplish now without a server without a database.
We have implemented my own version of the capabilities for profiling using Athena the new AWS offering as well as a few other tools.
AWS and in particular Athena will provide the capability to accomplish all these things on demand and pay as we go
We plan to follow this article up with detailed instructions on how to accomplish this however as you can see it is using all the existing standard tools for the S3, Athena and Glue,Lambda.
Well actually I spent some time in the QuickSight tool and below example is a pathetic example of a dashboard it’s just to show you real quick the idea of looking at aggregate patterns of data and details of data visually and literally after only a few minutes of effort.
Here a quick example looking at Phone1 column patterns for my customer dataset. Notice the majority of records is in the 999-999-9999 pattern for 89%.
Then we add a Field and do a quick drill into the pattern representing 14% or 9999999999.
Now we can see the actual numbers the make up the 14%
Another fast change and we are in a Bar chart.
I know these are very simple visuals however the last point I’d like to make is that this experiment took us about 4 hours and we did on an ipad Pro, including signing up and creating an AWS account.
Last word on connecting the business objectives to the data after scoring it. This may seem like a cumbersome process, however unless you link the informational and abstract concepts directly to data columns or fields you will never have a true understanding of information based on your data. This mapping is the key, the Rosetta stone for deciphering you data,
(Father & Daughter) Data Governance
As proud parents Theresa Whiteside and I would like to spread the word to our daughter Victoria Stasiewicz will be speaking at the MDM Summit in San Francisco and add to that our congratulations on her recent promotion as “Manager of Data Governance” for HealthNOW.
Jim Stagnitto thanks for your part in this journey
DNA and the concept of MDM( Master Data Management) have many similarities.
“There is a subtle difference between data and information.”
We in IT have complicated and diluted the concept and process of analyzing data and business metrics incredibly in the last few decades. We seem to be focus in on the word data.
And if you consider the primary business objectives of MDM is provide consistent answers with standard business definitions and an understanding of the relationship or mappings of business outcomes to data elements.
DNA vs MDM or IT’s version of DNA.
The graphic I’ve chosen for this post symbolizes the linkage or lineage of human beings to DNA. What I’d like to do is relate the importance of lineage of data to examples of human how DNA communicate lineage and then discuss it in our IT as it relates to business functionality
Living Organisms are very complex as is a company and its data or information. The genetic information of every living organism is stored inside these nucleic the basic data.
There are two types of nucleic acids(NA) namely:
DNA– Defines Traits, Characteristics
RNA – Communication, transfers information and synthesis.
Lets examine them.
DNA- Defines Traits, Characteristics
DNA-Deoxyribonucleic acid – In most living organisms (except for viruses), genetic information is stored in the form of DNA.
RNA – Communication, transfers information and synthesis
.
RNA – can move around in the cells of living organisms and serves as a genetic messenger, passing the information stored in the cell’s DNA from the nucleus to other parts of the cell for protein synthesis.
So here goes this is a bit of a stretch but if you consider DNA is the “Data” and a person the “Information” is created from the communication through the RNA.
To continue the analogy the DNA or chromosomes in and by themselves are out of context. It’s only once they been passed from one person to the next driven by RNA and result in a human being in that they become contextually realized as a human.
Again if we break down human DNA and inspected it, we can tell many things origin, or ancestory , traits of the person possible, diseases of the person, but it needs to be processed for us to understand the actual person.
So my point in relation to various approaches to traditional MDM or Master Data Management yis that this(DNA) is how life is created, it’s science, it’s not a methodology or product in approach a vendor guess work it’s real and the main point is “lineage” is the key
“There is a subtle difference between data and information. Data are the facts or details from which is derived. Individual pieces of data are rarely useful alone. For data to become information, data needs to be put into context.”
Examples of Data and Information as it relates to MDM
The history of temperature readings all over the world for the past 100 years is data.
If this data is organized and analyzed to find that global temperature is rising, then that is information.
“Mislesading” Data”
Data needs to be interpreted and analyzed, it is quite possible — indeed, very probable — that it will be interpreted incorrectly.
When this leads to erroneous conclusions, it is said that the data are misleading. Often this is the result of incomplete data or a lack of context.
For example, your investment in a mutual fund may be up by 5% and you may conclude that the fund managers are doing a great job. However, this could be misleading if the major stock market indices are up by 12%. In this case, the fund has underperformed the market significantly.
Comparison charts
“Synthesis: the combining of the constituent elements of separate material or abstract entities into a single or unified entity ( opposed to analysis, ) the separating of any material or abstract entity into its constituent elements.”
Synthesis in MDM
Communication, transfers information and synthesis, like RNA several companies are prevelant in terms of data movement and replication, in essence data logistics
Defines Traits, Characteristics like DNA many companies have developed and refined products and the techniques required for during these. They are data profiling , domain pattern profiling and record linkage, the basis of transforming data into information .
However lineage is the key, and without this to serve as a connection between data and information, in essence there is no information. And in this case the “information” is the business term from the business glossary.
The integration of data movement capability and the linking of data profiling capabilities can result in providing a business the capabilities answering business question with certitude through the transparency of lineage.
To translate this into “business terms” it’s very similar to providing an audit trail in that a business can I ask this is questions like what customers are the most profitable look at those customers and then drill into wood products what areas with those profits coming
What lineage accomplishes is to lay a trail of cookie crumbs for data movement but for your business questions, it’s simply makes sure that you can connect the dots as data gets moved and/or translated and/or standardize and or cleansed throughout your enterprise.
And with the simple action of linking data file metadata of files, columns , profiling results to a businesses glossary or business terms, will result in deeply insightful and informative business insight and analysis.
“Analysis the separating of any
material or abstract entity into its
constituent elements”
In order for a business manager for analyze you need to be able to start the analysis at a understandable business terminology level. And then provide the manager with the ability to decompose or understand lineage from a logical perspective.
There are three essential capabilities required for analysis and utilizing lineage to answer business questions via a meta-data mart and these are very similar to the pattern that exist in DNA.
1. Data profiling and domain(column) analysis as well as fuzzy matching processes that are available in many forms:
a. – Scan all the values within each column and provide statistics(counts) such as minimum value, maximin value, mode(most occuring) value, number of missing value etc…
b. Frequency or Column value Patterns – determine the counts of distinct values within a column and also identify the distinct pattern occurring for all values within a single column and associated counts ie(SSN = 123-45-6789 , SSN PATTERN = ‘999-99-9999’
c. Fuzzy Matching( similarity algorithms) – This capability enables the find the counts of duplicate or similar text values.
2. The results need to be stored is a “Metadata-mart” in order to see the patterns, results and associations providing lineage and retatingraw data to business terms and hierarchies.
3. Visualization and analysis capability to allow for analysis, drill down into data mart aggregated and statistical contents and associated business hierarchies and businessterminology
Underlying each of these analytical capabilities is a set of refined processes, developed and proven code for accomplishing these basic fundamental task.
In future post I will describe how to implement these capabilities with or without vendor products, from a logical perspective.
References:
http://www.diffen.com/difference/Data_vs_Information
 | John Indelicato on DNA and the concept of MDM(Mas… |
 | Data Profiling and D… on Creating a Metadata Mart via T… |
 | James Searle on Creating a Metadata Mart via T… |
 | irawarrenwhiteside on Creating a Metadata Mart via T… |
| James Searle on Creating a Metadata Mart via T… |
