cancel
Showing results for 
Search instead for 
Did you mean: 

Too Big or not Too Big, That is the Data

0
0
Big Data & DIKAR Model

Big Bang: Noise at birth

While you are reading this article, somewhere a baby cries. You can’t hear him. Nobody tells you. So, you don’t know this fact. Anyway, it is not your problem.

Suddenly, the baby next door starts screaming. This time, you are forced to listen. But what can you do? Life goes on ... The situation is different when it is your own baby. As soon as he begins to cry, you rush into his bedroom!

To be able to act quickly, you haven’t skimped on the necessary paraphernalia: the baby-monitor never leaves you. You have carefully chosen the channel so as not to interfere with the baby next door. You have adjusted the sensitivity of the device, so that you don’t jump at every beep.

The DIKAR model

There is noise, all the time, everywhere. Only noises that affect you are likely to be labelled to form a given data. Give it a meaning (one data only, or an accumulation, or comparison or overlap…) and it becomes a piece of information.
This information will be integrated (or not) into your global knowledge and urge you (or not) to make a decision, to take an action. This is the DIKAR model from Venkatraman [1996]: Data> Information> Knowledge> Action> Result

the DIKAR model

Any transformation is based on an interpretation of the environment model. This model depends on your business, your goals, the state of your organization, your capabilities...

The more you operate in a competitive market, the more relevant the model must be. The greater the need for sharing is, the greater the requirement for formalization is. The faster the environment changes, the more agile the modeling process must be.

From Big Data to Big Result

As is often with "buzz words", we put a lot of things under the name Big Data. This ranges from the use of new technology (Hadoop, Cassandra, ...) which support existing practices (a more flexible, faster, bigger, … Business Intelligence), to the launch of intrinsically innovative practices.

Of these, the best known are intended to identify correlations in the statistical processing of very large amounts of data, whether structured as forms, databases, ... or unstructured as traces of browsing or Internet searches, expressions in emails or on social networks ....

Without neglecting the problems of selection and data quality, the hope to find correlations is not in vain. In fact, it would be surprising not to find correlation in the mass of these sources! But are these correlations new sources of knowledge?

Does analysis of these correlations allow it to refine the perception of the environment that we already have, and to lead to a competitive advantage? Are these new technologies and the only "data" sufficient, or is it necessary to make another step in the DIKAR model?

That's one small step for your information model, one giant leap for your market position. Like GAFA, NATU, Uber ... that extra step may be the cause of your next business model. Maybe, it will be more modest but just as "disruptive" as was "Why-Cry", the baby monitor that analyzes baby cries [1].

Should we still architect information?

Not so long ago, we did not plan to manage information systems without knowing information that forms the heart of the system. Since then, there has been a great wave of outsourcing initiatives, software packages and integration projects that have mobilized a lot of energy.

When the information model is embedded in a software package, the contribution of IS on competitiveness is based on excellence in implementation and marginally on its design, on its originality.

In our digital world where the slightest "click" can make a major difference, one may be tempted to establish a "two-speed" IS, with an agile one for the front office and a structured one for the back office.

A simple map of "end to end" process shows the limits of a compartmentalized approach because the relevant information must permeate all activities to benefit from the system effect.

Architecting information is to identify the business concepts and their articulations, describe their semantic and chart their implementation across organization and software, and to the finest level when necessary.

At MEGA, we know that the development and the governance of this double information traceability - horizontally along the process and business capabilities, and vertically through the architecture layers - can only be envisaged as part of a tooled information architecture approach.

1. www.why-cry.com

 

Article co-written with Rebecca Schofield

0
Comment
MEGA

Big Bang: Noise at birth

While you are reading this article, somewhere a baby cries. You can’t hear him. Nobody tells you. So, you don’t know this fact. Anyway, it is not your problem.

Suddenly, the baby next door starts screaming. This time, you are forced to listen. But what can you do? Life goes on ... The situation is different when it is your own baby. As soon as he begins to cry, you rush into his bedroom!

To be able to act quickly, you haven’t skimped on the necessary paraphernalia: the baby-monitor never leaves you. You have carefully chosen the channel so as not to interfere with the baby next door. You have adjusted the sensitivity of the device, so that you don’t jump at every beep.

The DIKAR model

There is noise, all the time, everywhere. Only noises that affect you are likely to be labelled to form a given data. Give it a meaning (one data only, or an accumulation, or comparison or overlap…) and it becomes a piece of information.
This information will be integrated (or not) into your global knowledge and urge you (or not) to make a decision, to take an action. This is the DIKAR model from Venkatraman [1996]: Data> Information> Knowledge> Action> Result

the DIKAR model

Any transformation is based on an interpretation of the environment model. This model depends on your business, your goals, the state of your organization, your capabilities...

The more you operate in a competitive market, the more relevant the model must be. The greater the need for sharing is, the greater the requirement for formalization is. The faster the environment changes, the more agile the modeling process must be.

From Big Data to Big Result

As is often with "buzz words", we put a lot of things under the name Big Data. This ranges from the use of new technology (Hadoop, Cassandra, ...) which support existing practices (a more flexible, faster, bigger, … Business Intelligence), to the launch of intrinsically innovative practices.

Of these, the best known are intended to identify correlations in the statistical processing of very large amounts of data, whether structured as forms, databases, ... or unstructured as traces of browsing or Internet searches, expressions in emails or on social networks ....

Without neglecting the problems of selection and data quality, the hope to find correlations is not in vain. In fact, it would be surprising not to find correlation in the mass of these sources! But are these correlations new sources of knowledge?

Does analysis of these correlations allow it to refine the perception of the environment that we already have, and to lead to a competitive advantage? Are these new technologies and the only "data" sufficient, or is it necessary to make another step in the DIKAR model?

That's one small step for your information model, one giant leap for your market position. Like GAFA, NATU, Uber ... that extra step may be the cause of your next business model. Maybe, it will be more modest but just as "disruptive" as was "Why-Cry", the baby monitor that analyzes baby cries [1].

Should we still architect information?

Not so long ago, we did not plan to manage information systems without knowing information that forms the heart of the system. Since then, there has been a great wave of outsourcing initiatives, software packages and integration projects that have mobilized a lot of energy.

When the information model is embedded in a software package, the contribution of IS on competitiveness is based on excellence in implementation and marginally on its design, on its originality.

In our digital world where the slightest "click" can make a major difference, one may be tempted to establish a "two-speed" IS, with an agile one for the front office and a structured one for the back office.

A simple map of "end to end" process shows the limits of a compartmentalized approach because the relevant information must permeate all activities to benefit from the system effect.

Architecting information is to identify the business concepts and their articulations, describe their semantic and chart their implementation across organization and software, and to the finest level when necessary.

At MEGA, we know that the development and the governance of this double information traceability - horizontally along the process and business capabilities, and vertically through the architecture layers - can only be envisaged as part of a tooled information architecture approach.

1. www.why-cry.com

 

Article co-written with Rebecca Schofield