By now we've all realized that data has value. Some people are even calling it the "new oil". But there's a gap that hasn't quite been bridged to truly obtain this value.
Fundamentally and historically, the value of data has been realized through structure and relationships, which in turn serve either functional or analytical workflows. You have core business applications, data warehouses, reporting engines, etc. that serve this up and drive business transactions, workflows, decisions, etc.
More recently, we've been demanding more from data, seeking to better inform us of the present, predict future outcomes and do so in such ways that it goes beyond reporting to actionable business workflows and even monetization of the data itself. The challenge, however, is that most organizations struggle with the fundamental aspects of data, structure and relationships, and with digital proliferation matters have become even more overwhelming. Consistent data models, various technologies, data quality, integration architecture, etc. have made it very difficult to forge meaningful strategies and initiatives that can truly tap into this value we seek. Perhaps the strategy is to embrace "structureless and let ML do the work", however, NLP doesn't quite cut it, and you'll likely need a custom ML model (i.e. via TensorFlow), which could be promising but nothing mature has developed yet. Either way, it would seem structure and relationships continue to be a significant impediment. One way or another, you'll need to lay out a strategy (i.e. governance) that can, ultimately, bridge the gap and tap into "value".
