Contact Us: (617) 237-6111 / info@DataBoiler.com / P.O. Box 181, Weymouth, MA 02191                             © 2013-2018 Data Boiler Technologies - All rights reserved 
FinTech Pioneer | Risk Strategist
Big Data, did you get the sight? Scene from movie “The Matrix”, second creation and article by Kelvin To, Founder and President of Data Boiler Technologies, LLC What a scene it was that everything surround Neo are just bits and bytes in the movie, "Matrix"! Getting insights from Big Data is absolutely important. Yet, not everyone has the sight of how to use Big Data for meaningful purposes. The struggle to come up with Big Data use cases is real. First, everyone may define Big Data differently. Some sees it as a Hadoop open-source revolution. Others see it as anything to do with the social media. As for the data management folks, they would like to throw in the SOA, ESB, etc. to build the Big Data ecosystem, while the advanced analytic group focuses on the predictive and/or machine learning aspects of Big Data. All these different views are good as long as you can put them in the right contexts. Second, how to generate big data use cases depend on the problems you are trying to solve. Some problems are about optimization, such as using algorithms to forecast and determine the bank's optimal level of securities inventory for short-term demands. Others involve pattern recognition, such as machine learning to spam filter the prohibited proprietary trading activities for Volcker rule compliance. In addition, Big Data can enhance liquidity in the market if the credit team may access where else and in what conditions they can accept the risks. Furthermore, Big Data can unveil niche sales opportunity, help craft channel distribution strategy, avoid a "me-too" proposition in portfolio design, and efficiently generate client proposals. In Big Data, what works for B2C markets won't bring the same benefits in B2B, the key is about "fit-for-purpose". Once you get this right, the creativity in Big Data applications is unlimited. Third, anyone can judge the usefulness of Big Data, but not everyone can master "the usefulness of useless knowledge". This legendary article on Harper's magazine by Abraham Flexner in 1939 still holds some of the greatest principals to inventing. As critics commented University's training being too short-term focused, our modern society is indeed obsessed with practicality. Most people have overlooked curiosity and tenacity as key success factors for inventors. Abraham pointed out that "the real enemy is the man who tries to mold the human spirit so that it will not dare to spread its wings." In other words, self- imposed boundaries and/or overthinking of Big Data's practical use will work against our creative minds for natural curiosity. Free the environment by offering opportunities, not duties. It will bring freshness to new discoveries in a rapid accumulation process. Forth, beware of the pros and cons. Different NoSQL databases (column store, document store, key-value, graph DB, etc.) are designed to serve different purposes. For example, semantic graph and/or some document databases are meant for superfast OLAP based on data's relationships in a network using triples logic. It is a great method to provide meanings for data, especially in resolving nuisances with OTC ontologies and LEI reference data problems. Graph database in particular has OLTP drawbacks, while using triples logic to infer meanings on data aren't absolutely 100% accurate (i.e. based on likelihood of data correlations and affiliation, there exist the possible issues of false positive and false negative). In any case, all the NoSQL, Hadoop, Flume, SQOOP, HIVE, PIG, B-Trees, LSM, etc. are respectable inventions for Big Data. The more you learn, the farther you can go to find the right applications. Lastly, don't judge a book by its cover. Take the above picture for example, you can choose to laugh about it or consider the analogy to big data machine learning. Imagine a blind person in the men's room using pictorial sensors to recognize the urine bowls and check for the unoccupied. The technology sends a digital signal to the brain and enables the blind to see. This is an example of machine learning in Big Data. Big Data use cases can be limitless while Big Data talents are limited in supply. Seek them out and interview on their characters. You will be amazed by their depth of knowledge and insights to solve your problems.
Next >       Next > < Previous < Previous DATA BOILER TECHNOLOGIES, LLC 212 Home Big Data Big Picture Big Opportunities Volcker Compliance About Us