Analytics — deep, predictive, operational, (insert preferred flavor here) — has climbed to the top of business executives wish lists in the past few years. The explosion of big data from social media sources and the coming supernova from the Internet of Things promises complete understanding of customer needs as well as the prediction/influencing of future behavior. With sufficient data, best of breed algorithms, faster computers, and emerging deep learning approaches — statistical correlation will become a largely exact science. Understanding causation will become an unnecessary luxury. Welcome to the analytics nirvana.
Of course, inspiration and implementation often diverge. The day-to-day practicality of big data analytics continues to raise ongoing challenges. The “P-words” — preparation, people, prediction, and production point to four key areas where such challenges are encountered today and will arise in the future:
* Preparation of incoming data. Sourcing, cleansing, and contextualizing data before analysis can make the difference between valid discoveries and rabid nonsense, but preparation is taking 80% of the time and effort. Is this sustainable? Can it be reduced?
* People. Data scientists have been equated to unicorns due to their elusive nature. Where can they be found, what are their mandatory skills, can they be grown internally or bought in? What are the necessary roles and responsibilities? How does this fit with existing BI structures, such as the BI Competency Center?
* Prediction of future outcomes. What are the rules regarding algorithms? How can they be made trustworthy or futureproof? How can we understand and control self-improving algorithms? What are the possibilities and implications of nudging people towards desired behaviors?
* Production. How do we move effectively and seamlessly from the analytic laboratory to the production environments of manufacturing, operations, and sales?
But there’s more. How does big data analytics impact on the entire business-IT landscape? Which strategies are effective in managing its implementation? What tools and technologies are appropriate to big data analytics? What is its relationship to cognitive computing? What are the ethical and economic implications of big data analytics at its most extreme?
An upcoming issue of Cutter IT Journal with Guest Editor Barry Devlin seeks your insight into and experience of big data analytics in the four areas above, and beyond. We welcome articles from practitioners who can share real experiences, both positive outcomes and challenges that proved implacable. We would be delighted to hear from academics involved in application-oriented or industry-based research. In brief, we want to hear from those who can share practical experience, present empirical evidence, and provide recommendations on successfully harnessing the potential of big data analytics. Furthermore, we seek reasoned opinions and innovative predictions—both upbeat and down—on how this field may evolve over the next five years.
Topics of discussion maybe include — but are not limited — to the following:
- What are the challenges in analyzing big data and how can they be addressed?
- How can the time it takes to prepare data for analysis be reduced?
- Can we learn lessons from the experience of building data warehouses?
- What mandatory skills should a data scientist possess and what role should he/she play in the current BI structure, such as the BI Competency Center?
- What types of algorithms should be employed? How susceptible are they to gaming? Could they become self-fulfilling prophecies?
- What production roles will be augmented? How far can production be automated?
- Which strategies are effective in managing a big data analytics implementation?
- What is the role of cognitive computing in big data analytics?
- What tools and technologies will maximize analytic efforts?
TO SUBMIT AN ARTICLE IDEA: APRIL 15, 2016
Please respond to Barry Devlin at barry[at]9sight[dot]com and with a copy to cgenerali[at]cutter[dot]com no later than April 15, 2016 and include an extended abstract and a short article outline showing major discussion points.
ARTICLE DEADLINE: MAY 20, 2016