Big Data Must Be Multifaceted [Dr. Cheong's Talk Part 2]

The term "Big Data" has been trending on Google Trends for several years, undoubtedly becoming a buzzword in the information age to describe the exponential growth of data. It is estimated that since 2010, the amount of data generated globally each year surpasses the total data produced throughout human history up to that point. This immense volume and popularity of big data have made it both highly sought after and somewhat intimidating.

In the previous article, I, Dr. Cheong, briefly discussed the four V's of big data (Volume, Variety, Velocity, Veracity). This was mainly to align with the current popular discourse; otherwise, even the credibility of discussing the topic might be challenged. Some believe that all four V's must be present to qualify as big data, while others think that handling large or complex data analysis constitutes big data. Some even repurpose business intelligence (BI) dashboards with flashy statistical charts and call it big data. I prefer using the term "data," but to go with the flow and due to the lack of a better alternative, I continue to use "big."

Whether one is fascinated by or resistant to big data, the key is to understand what kind of data can demonstrate its value (Value) under what circumstances.

Numbers, Text, or Images Accumulate Every Second

We often hear about how big data can precisely calculate consumer preferences, leading to targeted marketing and customer relationship management (CRM). This type of big data usually refers to behavioral or transactional data that can be quantified. For example, a consumer searches for a specific product on an e-commerce platform at a particular time (say, 9 PM), searches for the product multiple times (e.g., 3 times), eventually purchases the product (5 units), and makes an online payment (totaling 777 yuan).

Another example is ride-hailing apps. From the moment a passenger searches for a vehicle using the app, to the driver accepting the request, and finally, the passenger reaching the destination and the driver earning a fare, the process generates data that includes the passenger's identity, order time, driver and vehicle information, geographic location, route, fare, etc. Some of this data is historical, while some are real-time records, and most of it exists in numeric form.

Additionally, in both scenarios mentioned above, if we include post-transaction activities where consumers/passengers and e-commerce merchants/drivers rate and review each other (e.g., giving a positive icon, 5-star rating, or a few comments), this data is no longer purely numeric but also includes substantial amounts of text or images.

From these examples, we can see that big data is multifaceted, encompassing numbers, text, images, and even streaming audio or video. They all share common characteristics: they can be recorded in real-time, accumulated, calculated, tracked, and reused. This is where the value of big data lies.

Dr. Angus Cheong Chairman of the Asia-Pacific Internet Research Alliance and Chief Data Consultant at uMax Data Technology Ltd.

(Originally published in Hong Kong Economic Times, reprinted with permission)