The Importance of Data Mining and Why Study Data Analytics

One topic that captivated my attention in class was on the topic of Data Analytics. As an advocate of technology, an aspiring entrepreneur and a consumer, I felt the importance to comprehend and understand what is Data Analytics, what it could do or potentially do and the science behind it.

As an aspiring entrepreneur, I truly believe that in this day and age, whichever industry a firm belongs to there is definitely benefits in utilising data. As long as a firm move things, sell things or just communicating with customer, there will be trails of data. A new business model can even be created based on data. As mentioned in “Are you ready for the era of ‘Big Data’”, “Another global company learned so much from analysing its own data as part of a manufacturing turnaround that it decided to create a business to do similar work for other firms.” This is exciting as it tells me the direction, the kind of business that will come to exist in the next decade.

As an advocate of technology, I have always love the evolution of technology. Much of the gadgets that I own are connected to the web as part of Internet of Things which revolves around the basics of data analytics.

As a consumer, I benefited from this era of Big Data. As pricing data proliferates on the web, we consumers are offered price comparison service, which is a disruptive force from a retailer’s perspective but have created substantial value for myself. I can always make sure to be getting the cheapest flights for my next adventure through websites like SkyScanner.

Discovery

While researching deeper into Data analytics and technology of the world today, I stumbled upon the technology of Digital Twin, a technology that I never knew existed. I was blown away. There was data analytics being use at every step of a digital twin. This blew my mind because it made me contemplate about what I think I know about our technological advancement. Maybe my apple watch isn’t as impressive as I really thought it was. Think of the potential to see virtually inside any physical system or assets and know its usage, predict its maintenance and to improve its experience. When digital twins do get commercialise and become easily available to the public, the possibilities are endless. I am sure I will know days before the next time my air condition in my room starts to leak.

Challenge

With a deeper understanding of the science behind data analytics and machine learning, I discovered upon a potential red flag. Our historical data might not be unbiased. Training data in machine learning is what our future are built on, whereby decisions are made and patterns are defined. Ultimately this is how bias can be created. There was a research done by the computer science department of University of Copenhagen whereby a robot read 3.5million books to see how we describe men and women differently. “Beautiful, sexy and gorgeous” were used most often to describe women and “Brave, rational and righteous” were used most often to describe men. Negative verbs about the body cropped up 5 times more often for women and overall adjectives for women were focused on appearance. While for men, behaviour was key. The robot analyse 11 billion words from books dated from 1900 to 2008 and discovered gender biased in both non-fiction and fiction books. This could have a serious impact on machine learning as algorithms works by identifying patterns in data. If these patterns are based on bias language they will naturally produce bias results. This could affect who gets the job if computers are used to sort jobs application. Computer scientist and Assistance Professor Isabelle Augenstein of University of Copenhagen also pointed out that the algorithms used to create machine learning and applications that can learn and understand the human language are fed by data in forms of text from online sources. This is the same technology that allows our smartphones to recognise our voice (i.e. Siri and Google Assistance) and also how google provides keywords and suggestions. This means that the result of biases could skew the language that our smartphone use to interact with us humans. Moving forward, instead of asking ourselves if the machine was the one that made the decision, we should be asking if the decision is right for all.


My Personal Growth

Other than directly learning about how data analytics can impact my learning. I think there is also much to learn from the trends of data analytics (i.e from Data Analytic 1.0 to 3.0) Just like how data analytics can be classified into descriptive, predictive and prescriptive, I feel that my learning experience can also be classified in the same manner. When faced with uncertainties, a typical strategy I would adopt is to be “descriptive and predictive”. For example in preparing for a job interview, what I would have done will be “copying” the good practices used in an interview, learn from past success of interviews and have a rough idea of what are the likely future. 

For instances, the “tough” questions that I might faced. Similar to how data analytics has evolved to be more prescriptive, I think it is imperative that I evolve my way of prepping for interviews. With my current approach of predicting questions and answers in my next interview, perhaps it is better to understand why it would be asked. Why would the interviewer need that information. This could be a better insight to the likely future as compared to just knowing what comes next as I will know why it comes out. Therefore having a better understanding. This will further my preparations by providing the necessary steps I would need with each possible scenario. Which will lead to the next topic on Scenario Planning of different outcomes which I do believe will work hand in hand with this concept of learning.