Everything in our digital universe is connected. Every single day, you wake up and start a series of interactions with people, products and machines. Sometimes, these things influence you, and sometimes you play the role of the influencer. This is how our world is connected, in a network of relationships and influence among people, products and technologies. Understanding connections is the realm of graph analytics and its what’s next in big data.
The better we can understand connected networks, the more knowledge we have to proactively enrich numerous aspects of everyday life and work – to better understand things like traffic patterns, consumer behavior, weather anomalies, and to make predictions about business, crime or disease trends. It is no longer enough just to consider the content of a specific analytic entity, whether that’s a customer or a network; we also have to understand the context of connections between customers and among networks to unlock the insight and value that comes from charting relationships between entities that influence outcomes.
Sounds great, right? So, why hasn’t graph analytics caught on like wildfire, like you’d expect. Typically, the problem has been complexity and scale. As you can imagine, once you move beyond statistical analysis of single entities as they relate to a mean, and start looking at relationships between hundreds, or even thousands of individual entities, the data processing load increases by several magnitudes.
Thus far, graph analytics has been difficult and expensive, requiring specialized systems, unique, hard-to-find skillsets and a patchwork of algorithms to discover these intertwined relationships. And, as a result, the use cases have been relatively limited. But, all that is changing. And, today’s engineered solutions offer a set of advanced, contextual analytics delivered economically and at massive scale on large multi-structured data sets. And, as that happens, these systems for analyzing connections in networks of data will open up new ways of looking at big data, and many new use cases.
Analysis of connections in networks will combine traditional statistics, machine learning and sentiment analysis with influencer analysis to examine customer satisfaction and pinpoint the customers who hold the most influence within a network of people. The ability to track influencers who have the most direct and indirect impact on product purchases can also help inform things like proactive viral marketing campaigns. It will allow companies to detect, in near real-time, the cyber-threats hidden in the flood of diverse data generated from IP, network, server and communication logs – a huge problem, as we know, that exists today.
The ability to track relationships between people, products, processes and other “entities” remains crucial to breaking up sophisticated fraud rings. If you only use a content-based decision model, you can get stuck the minute a fraudster creates a new identity, or changes their behavior. We’ll be able to understand the behavior of connected devices, using sensor data to more efficiently modulate electricity, and power smarter cities. We’ll analyze traffic patterns and driving behaviors, particularly as more connected and self-driving cars start hitting the roadways. In healthcare, providers will be able to quickly scan through network graphs of patients to discover therapies used with other patients with similar characteristics (such as age, clinical history, associated risk factors, etc.) that have the most positive outcomes.
Our own imagination, and those of our data scientists only limit the possibilities when it comes to analyzing relationships and influence among people, products, processes and devices for new insights. As I alluded to earlier, graph analytics is also the future of big data analytics, because everything is connected.
0 comments:
Post a Comment