The Evolution of Analytics


In the technology revolution we are living, I’m sure that you read or heard a story about how data is changing our world. Data may cure a disease, solve a national problem, prevent a disaster, boost a company’s performance, make a team more efficient or enhance our experience.

Data is essentially the plain facts and truth collected during the operations of a business, while you are searching for some articles on the internet, using your mobile for sending a message or finding a location. Data is everywhere and increasing significantly, data is the lifeblood of decision-making and the raw material for accountability and the oil of this century!

Nevertheless, data alone cannot do that without proper understanding, analysis, visualization, transformation, and enrichment to discover the hidden power and the potential of data and reach the top of the Data Information Knowledge Wisdom (DIKW) pyramid.

The figure below describes the data value chain and the importance of analytics to gain the knowledge and wisdom. In this article, we will discuss more the different types of analytics and how they are important for business maturity.

data value chain

Data value chain

What is analytics?

Analytics is transforming data assets into competitive insights, that will drive business decisions and actions, using people, processes, and technologies

Analytics is the discovery, interpretation, and communication of meaningful patterns in data. Especially valuable in areas rich with recorded information, analytics relies on the simultaneous application of statistics, computer programming and operations research to quantify performance.

Analytics is multidisciplinary and has lots of applications, like web analytics, business analytics, security analytics, people analytics, cloud analytics, text analytics and more.

The main challenge of having good analytics that it is mainly performed by information technology experts. The most important that it should be easily done by business users while the business user shall have multi-discipline skills (technologists, mathematicians, statisticians, business deep understanding, and data knowledge as well) and this business user should know how to talk business with business people and technology with technology people. He should never stop learning and should have tough analytical questions to derive the insights.

Recently, I read one of the reports from the Elder research which describes the 10 levels of analytics and their maturity as captured in the below figure. I recommend reading the full report by following this link.

Analytics 10 levels

What are the types of analytics?

Most of the reports and technology vendors are addressing 5 types of analytics, let us explore each one of them.

AnalyticsAnalytics types

Descriptive analytics

It answers some questions, like what happened, when, and where to describe, or summarize raw data about the past at any point of time, whether it is one minute ago, or one year ago. Descriptive analytics is useful because they allow us to learn from past behaviors, and understand how they will impact future outcomes.

Descriptive analytics is based on standard aggregate functions, which require knowledge of math. Most of the social and web analytics are descriptive analytics. They summarize certain aggregations based on simple counts of some events, for example, the number of likes, posts, friends.

Techniques

Aggregation functions and methods

Diagnostic analytics

It answers the question for why it happened to find the root cause of an event. It can be used to provide an understanding of cross-functional data to find the relation between things and can be called discovery analytics.

Techniques

Correlation, drill down, comparison, and data mining

Predictive analytics

It is the type of analytics to answer what might happen in the future to have some forecast and prediction so the organization can take corrective actions. Predictive analytics can be used throughout the organization, for example, forecasting customer behavior and purchasing patterns to identifying trends in activities.

Techniques

Data mining, Machine learning, Regression analysis, predictive modeling …etc.

Prescriptive analytics

Prescriptive analytics attempt to quantify the effect of future decisions in order to advise on possible outcomes before the decisions are actually made, it tries to provide how to do the actions and what are the consequences of these actions to derive a better decision.

Prescriptive Analytics can be used anytime you need to advise users on what action to take.

Techniques

Optimization algorithms and modeling, rules and decision techniques

Cognitive analytics

It tries to mimic the human brain by drawing inferences from existing data and patterns and understand the context, draws conclusions based on existing knowledge bases and then inserts this back into the knowledge base for future inferences to answer what to do for the future how and why?

Techniques

Semantics, Artificial intelligence algorithms, Deep learning and machine learning

Note

So, what is actionable analytics? I think all analytics should be actionable in the five types we discussed above as the main goal of analytics is to discover and understand data to take an action. Without the action, no need to go for it. Moreover, others say that actionable analytics is another name of Prescriptive analytics while I do not think that it should be limited to that.

Examples!

If we observe the insights we might have while applying these types of analytics, we will notice a significant change in the way of asking the analytical questions. Furthermore, it is important that we have the technological capabilities and data breadth and variety to be able to find the answers to those questions.

Analytics type Traffic Healthcare
Descriptive Analytics How many vehicles are in the world today? How many patients have communicable diseases?

When they discovered that?

Diagnostic Analytics Why we have 1.2 billion vehicles in the world?

Is it related to Countries’ population and prosperity?

Why they have communicable diseases?

Is there any correlation between demographics of patients and their diseases?

Predictive analytics How many expected vehicles in the next 5 years? How many patients may have the same disease in that City?
Prescriptive analytics How we can decrease the number of vehicles? What are the choices to prevent others to be infected?
Cognitive analytics What is the best action to decrease the number of vehicles globally without decreasing the production ratio?

You should invest in electric cars to overcome the pollution issue in the big cities, there are millions of people interested and talking about the electric cars.

You should give the patient this medication and isolate them in this area, you should not give them this medication because it may have a negative effect after 10 years due to some studies have been made in these two articles.

After examining these patients faces, the patients’ number x,y, and z are working in the same company which located in an infectious area and this is the main reason for their infection. We found some coworkers who have this disease.

One of predictive analytics examples, the Google Flu Trends which estimates of Flu and Dengue fever based on search patterns.

Another research from Microsoft, the Web-Scale Pharmacovigilance, they wanted to find the pattern in online searches to find the relation between drugs and side-effects, also the combination of specific drug pairing of paroxetine and pravastatin whose interaction to cause hyperglycemia.

I think the online ads engine is a good example for cognitive systems, they understand the context around the user who is searching or exploring the internet and recommend the most relative ads not only based on searches, but based lots of data related to this user, for example, his profile, family, friends, historical purchases, the places he visited, the sites he visited, the place he lives at …etc.

References

Advertisements

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s