Types of Problems Suited to Big Data Analysis

Today, we covered the problems that arise with analysis from Big Data. We discussed why Big Data might be less predictive and powerful than first thought. For example, Big Data that has been collected over the past 5 years sounds like it could be trustworthy enough to form strong predictions from; however, the longer the Big Data collection has gone on, the better. If data has been collected for 20 years, it will hold much more accuracy than data collected over 5 years. This is because it can pick up trends that have appeared, then disappeared, then perhaps reappeared. 

The issue that follows on from that is if you want to create accurate predictions, you have to find data that has been collected at a wide scale for decades, which is still rare in todays world but also expensive, especially if you want to start collecting it today. You would require large expensive data collection methods, long term, backed up storage facilities and a strategy successful enough to last decades.

Furthermore, even after a long and difficult Big Data collection process, it will not be able to predict everything. For example, data collected from the 1970s until 2000 would struggle to indicate the sharp rise of smartphones in the 2010s and 2020s. 

In conclusion, I think Big Data is one of the most powerful tools in the modern world, however, its expenses and inaccuracies shouldn't be underestimated.