Microsoft's three views on big data

It seems that the tool of big data makes us very frightened. In addition to the cost of infrastructure, it employs scientists who are able to find actionable ideas from a wealth of unstructured information from a variety of sources.

But maybe big data isn't as scary as you might think. According to Microsoft, it is ready to position itself as a large data enterprise, leveraging Business Intelligence (BI) tools to serve the enterprise.

For example, the company recently released a preview version of 2013 Excel Data Explorer, a self-service bi plug-in that better serves the day-to-day business staff (not just data scientists), makes it easier for them to find value in data from a variety of sources such as Hadoop, a large data platform.

Microsoft also recently announced the launch of the SQL Server 2012 Parallel Data Warehouse (PDW), a large-scale parallel processing data Warehouse device designed for the integration of Hadoop.

This brings us to the confusion about Big data: what are they? How will Microsoft use it?

The first point: Microsoft's SQL Server General manager wrote in his recent blog that it is hard for an enterprise's it stack to support large data because of the inevitable infrastructure and scalability dilemma.

An organization may not need the server and other hardware racks needed for large data solutions. "Instead, an affordable solution should provide the ability to store and process large amounts of data while eliminating the cost of any upfront infrastructure, because you only pay the storage and computing power you need," Kelly said. ”

We are not surprised that Microsoft's Windows Azure hdinsight service fits this description. But Kelly's comments are still useful for companies that are around and want to get big data solutions from other suppliers.

Microsoft's second view is the looming shortage of data scientists: companies cannot find large, qualified data masters, and find forward-looking insights from unstructured sources of information such as social media content and machine sensor data.

"The industry needs more data scientists, although most organizations now have the staff they need to help them make valuable insights from the data to improve their business," Kelly said. ”

In other words, large data tools and applications are sufficient today. Microsoft's claim is based primarily on the democratization of the so-called data movement. Popular tools such as Excel data Browser plug-ins (Excel with the add-in) enable end users to conduct BI analysis directly (business intelligence) without the help of it.

The third point: Big Data is both an opportunity and a challenge.

OK, this may sound like marketing nonsense, but it's actually a bit real. Achieving a large data platform is both a big problem and an important opportunity. But don't let the former stop you from pursuing the latter.

Campbell, a technology researcher at Microsoft, said: "I often ask myself where the final value of large data is and how I should enter it." In my mind, these two key points are nothing more than time insights and the rate of return on accessible data. These two key points, in turn, help the process of information production. ”

Campbell says that the production of information is the process of transferring data from one domain to another. Good information tools enable businesses to gain business insights in a shorter period of time. With the addition of a large data tool, you have more extensive structured and unstructured data to explore. "These tools allow you to jump from the intuitive phase to the validation phase very quickly," he says. ”

(Responsible editor: The good of the Legacy)