A revolution is occurring in modern analytics, driven by our ability to capture new sources of information at a detail previously too complex and costly to imagine. As more data comes from new sources (from machines to social media) and is applied to new applications, data is evolving into greater diversity, including every variation of data type from unstructured to multistructured. Even as new tools to analyze and manipulate this newly available resource come online, it is not enough to look at the data manipulation layer alone.
Philip Russom, Ph.D.
Organizations of all sizes are in competition to realize value from data – and to realize it faster. To do so, they increasingly need flexible and agile business intelligence(BI), analytics, and data infrastructure, not systems that take too long to develop and do not give users the dynamic, iterative, and interactive access to data that they need. Fortunately, technology developments are trending in a positive direction for organizations seeking to accelerate their path to value with BI, analytics, and the critical supporting data infrastructure. These include self-service BI and visual analytics, self-service data preparation, cloud computing and software as a service(SaaS), and new data integration technologies.
Cambridge Semantics, Looker, Modemetric, SAP, SAS, Tableau Software, Unifi Software, Zoomdata
An increase in data maturity correlates to an increase in business success. Yet though organizations gladly allocate budget to business projects, they neglect data maturity—even to the point of allowing it to deteriorate.
As BI and analytics become more mainstream, organizations are realizing that it makes sense to both enrich and augment their data in order to gain more insight. Successful companies realize that utilizing traditional structured data only for analytics is a non-starter. Organizations are more often adding ‘new’ data sources to the mix, including demographic data, text data, and geospatial data to their data sets. They are also looking for external data, such as social media data, weather data, and other third-party sources. The demand from data consumers has also driven many new organizations to pursue sharing their data. Many of these data sources are cloud-based.
Fern Halper, Ph.D.
One of the strongest trends in data management today and into the future is the development of complex, multi-platform architectures that generate and integrate an eclectic mix of old and new data, in every structure imaginable, traveling in time frames from batch to real time. The data comes from legacy, mainstream enterprise, Web, and third-party systems, which may be home grown, vendor built, open source, or a mix of these. More sources are coming online from machines, social media, and the Internet of Things. These data environments are hybrid and diverse in the extreme, hence the name hybrid data ecosystems (HDEs).
Philip Russom, Ph.D.
Machine learning is the analytics buzz word of the day. While some of the techniques have been around for decades, what has changed is the volume and diversity of data as well as the compute power to find insights in that data faster. That means that machine learning against disparate and big data can be used to get to insight – and fast. Machine learning is being used in predictive analytics in numerous use cases from customer behavior analysis to predictive maintenance to image recognition and more. The value is real and growing.
Fern Halper, Ph.D.
Want to become a data engineer but aren’t sure which technologies are the right fit for the job? People switching into big data are faced with a difficult decision—should you learn MapReduce or Spark? The answer seems simple, but requires more information and insight. Answering this and other questions correctly places you on the path to becoming a data engineer.