Cos Should Take A Holistic Approach To Big Data

Big Data is still in a phase where organisations need to understand its business value. Mitesh Agarwal, CTO & Director, Systems Solution Consulting, Oracle in conversation with Biztech2.com stresses on the need for a holistic approach towards Big Data that addresses the challenges across different phases.

How should the enterprises approach Big Data?

When we say Big Data, we are referring to a massive amount of data which traditional systems can’t understand. It is going to be the next big thing for enterprises where data volume is huge and majorly unstructured. CIOs need to figure out the three Vs of Big Data before devising a future strategy - ‘Velocity’ - the speed at which the data is coming; ‘Volume’ - amount of data, whether structured or unstructured, which is getting stored; ‘Value’ - what significance this data holds for the company.

Organisations should take a holistic approach that looks into every aspect, right from the phase of acquisition of data, storing it, analysing it, and finally deriving value out of it. This end-to-end approach will justify the V-model and will help organisations take all the data centric decisions more effectively.

More from Biztech

Future Group - Reliance Retail Deal approved by CCI RBI ban on cryptocurrencies takes effect; prohibition could force investors to tap the black market

What are the possible challenges across these different phases that you mentioned?

The challenge with the acquisition phase is that everyone knows how to acquire structured data but what if the data is coming from something like a Twitter feed? Do we have a mechanism which can probably understand a Twitter hashtag?

After acquisition comes another challenge of storing this unstructured data. For efficient usage of the data available, it is important for the meaningful data to be filtered first. For example, if there are a million tweets on an online campaign, it might be possible that only a few hundred are useful. So, the data storage should be intelligent enough to do the filtering. After analysing the filtered structured and unstructured data, it should be available for critical business decision process in a simplified interface.

How can organisations best optimise the analytics part?

The CIO plays a very significant role here. Firstly, the CIO needs to ensure that all the steps before arriving at the analysed data are taken correctly. The final data should be a proper representation of the entire data collected, and the analytics tools used should be able to define the desired value out of it. For example, in the context of telcos, if the decision making authority wants to know the number of call drops in a particular area, the tool should be able to understand, interpret and distinguish call durations of dropped calls from the regular ones. To maximise interpretation, organisations need software utilities within their appliances, like for instance, tools built within the mobile tower itself to understand network signals from mobile towers.

Talking of telcos, another challenge they face is management of subscriber data. How can Big Data help deal with this effectively?

Traditionally, all the data management challenges were dealt with through separate systems for both BSS (Billing Support System) stack and network stack. However, with customers increasingly demanding specific services and coming up with specific complaints, the challenge now is to know precisely which stack is causing the problem. For example, if there is a persistent conference call drop, the network stack will only tell how many users were connected at that time. But, to solve this issue there is a need to integrate BSS also to tell how many of them were prepaid and postpaid, how many were having active 2G or 3G data during calls. And, that is where Big Data analytics comes into picture.

Big Data analytics helps to interpret data coming from the combined stack, and the result is we know exactly which layer was responsible for those call drops and solve the issue accordingly. With Big Data analytics it is no more ’subscriber data management’ but only ‘subscriber management’ as specific concerns of a customer can be addressed seamlessly.