Buckle up for Big Data
There is no doubt today that big data has arrived. More than a few organizations are experimenting with the kind of insights they might be able to get with so much information. These organizations must understand the intricacies of foundation technologies used to build the infrastructure to make sense of the data.
We are already experiencing a gap between companies that understand and exploit big data and companies that are aware of it but do not know what to do with it. Facebook and Google are leading the charge when it comes to big data, but other industries such as finance, retail, telecom etc. are not far behind and are trying to follow new trends.
The Semantic Data Model in Big Data
The key to taking unstructured data, including audio, video, images, unstructured text, events, tweets, wikis, forums and blogs, and extracting useful data from them is to create a semantic data model as a layer that is on the top of data stores. This is necessary to help in making sense of all the data. The semantic data model is a relatively new approach, based on semantic principles that result in a data set with inherently specified data structures. Mostly, singular data or a word does not convey any meaning to humans, but paired with a context, this word inherits more meaning. The context of data is often defined mainly by its structure in a database environment, for example, its properties and relationships with other objects. Therefore, the vertical structure of the data is defined by explicit referential constraints in a relational approach. However, in semantic modeling, this structure is defined in an inherent way, which means that a property of the data itself may coincide with a reference to another object. A semantic data model can be illustrated graphically through an abstraction hierarchy diagram. It shows data types as boxes and their relationships as lines, represented hierarchically. Therefore, types that reference other types are always listed above the types that they are referencing. Yeah, read that sentence a few times for it to really sink in. This hierarchical representation makes the model easier to read and understand.
Securing Big Data
The collection and accessibility of so much data also means that businesses implementing it must become vigilant about the security of the data. Think about security architecture from the very beginning. Companies that collect and leverage big data often find some “toxic data” on their hands. For example, take a company which collects machine data that can be used to provide insights to user behavior. They might also have data which is not of any direct use to them, such as credit card numbers, patterns of usage information etc. Although the capability to correlate that data and draw inferences might prove to be valuable, it can also be toxic. If this data goes outside the company and ends up in hands of someone with ulterior motives, it can be devastating for customers as well as the company. Companies would have to develop a solid and impenetrable security structure before employing any big data solutions. Big data is a technology that can be applied to every field and every industry. But enterprises need to prepare properly for it as it may not be easy for them to handle so much data with so many security concerns. You have to be fully aware of the risks involved with big data before you dive into it.
#BringItOn
Leadership and Keynote Speaker and member of the Data Science Research Centre at University of Derby
8 年No doubt a semantic model will be helpful. But first you need some really good business questions, the answers to which will help the organisation. Smart questions will most likely lead to smart insights. Dumb questions will mostly lead to useless and wasted analysis.
Strategic Sales Leader | Expert in SaaS Growth & Lead Generation | Master of Outbound Strategies | Skilled in CRM Optimization & Team Development | Proven Track Record in Driving Revenue & Enhancing Customer Success
8 年Big data can be utilized when constructs are built. I used to think that big data, as having some type of rhyme or reason. It doesn't. It it sand on the sea. Now, if a company want to use it, they need to figure out the why. Why do they want to use their data. Once they figure that out it is easier to find a solution. Then we can move to what and how. I have learned a lot talking to my data scientists and how VP of sales at fortune 50 companies are using this treasure trove of data. It's amazing!
Executive Brand & Marketing Consultant - CBO/CMO/CPA - Branding/Marketing/Strategy/Finance - Starbucks / Disney / Universal Studios / Noodles & Co. / BDO
8 年Nice article Naveen. Like what you're getting at here. Let's face it data and the ability to learn from it has been a business advantage as far back as anyone can remember. In fact, three of the four buckets (enterprise, public, transactional) have been around for quite a while...albeit the amounts and ease of access were limited. It's that fourth bucket (social) that's really brought dimension to the enterprise, public and transactional data...suddenly we begin to be able to put a "why" around all the other data and that allows us to build a better "how" we talk with Customers! The challenge and opportunity is being able to utilize the new insights to better "serve" the Customer rather than just "manipulate" the "customer." Brands leveraging the data to build meaningful and deeper connections are winning the long race (think Starbucks, Apple, BMW). Brands using the data to simply manipulate the Customer may have generated nice ROI in the short term. However as of late it seems those brands are struggling to deliver sustainable results (think fast food discount-driven brands) and are pressured by the enterprise to focus only on the Individual channel ROI. Using all the data collectively to build better connections and deeper relationships with your Customer (vs just extracting value) is not surprisingly emerging as the long term winning formula. I wonder if selling the idea of investing in digital media/marketing on the ease of measuring an direct ROI (compared to traditional media) might be coming back to haunt us. Perhaps a better way to view the new digital frontier is the added benefit it brings to all the marketing tools available to better build the brand, connect with the Customer and deliver long term business success!
Sr. System Engineer, Vanderbilt University Medical Center
8 年Part of the difficulty is in trying to define big data. There is the semantic model, but there are others as well. It seems that the ability to figure out good ways to transform raw data into useful bits that can be processed for good purpose. The actual integrated technologies that do this are all secondary to the skills of the person or team analyzing the data. The irony here is that the jargon will be throw around without any recognition that these techniques are not new. Just their implementation methodologies.