A Data Analysis View on Technology Trends
?brahim Mutlay
Consultant | Innovation & Business Development | Deep-Tech Commercialisation | Digital Strategy | Materials Science | Process Engineering | Sustainable Manufacturing | Marketing | 20 years | Industrial Problem Solving
Predicting future technological trends is a crucial task for industry, investors and policy makers. Such a prediction usually needs to follow the evolution of technologies and extrapolate them analytically. Today, technological progress manifests itself via mass media through countless channels. You can access easily gigantic amount of tech news online. But the same abundance of information could obscure the important trends and knowledge patterns. So anyone who wants to discover meaningful information in that media data “soup” has to refine and extract it properly.
Statistical data analysis is the sole way to extract information from online media. Data analysis starts with finding the right data stream; in the extremely complex online information flow you have to know where to look and monitor. Then you will extract raw data in the form of unstructured numbers and texts so data and text mining will play a major role in this step. After crunching the analytic trends you can make predictions.
In order to conduct my analysis I use a newsfeed data API collecting real-time news articles of 150.000 per day from 100.000 sources worldwide. By especially focusing on science & technology news I revealed the major keywords which are the top hot topics circulated in the mass media.
As you see material and electronic technologies cover the mainstream news. The lack of IT news might be related to the API classification approach but in my experience it is a very accurate picture of the current media.
The relation patterns between technologies give insights about their organization which could show their correlations and dynamics. This is represented by the keyword connection graphs.
Another way to monitor the information trends is Google’s data sources. Under the project of Google Books millions of books are being scanned giving invaluable book contents from even 1800s. By utilizing the suitable tools you can mine that text corpus. I chose several keywords defining 20th century technology and looked how they evolve in the book literature.
A more efficient analysis of 21th century trends can be done using the Google search hits. I visualized the Google Trends data for several modern technologies. The interesting finding is that in 2012-2013 there is a shift in the technological conjuncture.
This a very coarse draft of what data analysis is. Making real-life predictions would need deeper techniques. If you want to know much more about that work you can contact me.
?brahim Mutlay
Grafen Co. / www.grafen.com.tr / [email protected]
Dark by Design ZeroTrust Principal Executioner.
7 年Interesting about the API as that correlates to other business needs like Cybersecurity where we see Zero Trust Networking leverage much of the same AI and principles.
CEO at Nanocore Tech. Vice President Research and Development CAYAKI Canadian Charred Wood Experts Founding President Asia Pacific Design Center. China
8 年Very good texts and images Ibrahim. Congratulation.
Senior Member at AMAC-Alumni FER
9 年Ibrahim, excellent post. There are millions of scientific and engineering documents in digital form. However, not all documents and books are original and true. Text data mining algorithms may identify (and score) the veracity (truthfulness) of scientific documents and books. That is the goal of KDD-Cup 2016 conference and competition.
Founder of Scimagine | Helping to safely store, organize and visualize experimental data on Advanced Materials | Nano
9 年It is serious analysis and might be very useful in understanding of trends in technology.
Great post Ibrahim Pasha !