The world’s leading publication for data science, AI, and ML professionals.

Where people are most interested in Data Visualization and Data Science

Today I extracted Google Trends data using R to find cities with most searches in 'Data Visualization' and 'Data science' as well as…

Today I extracted Google Trends data using R to find cities with most searches in ‘Data Visualization’ and ‘Data science’ as well as trending topics related to both.

Where the interests are

During last four years, we can see US west coast Cities like Seattle, SF and Washington had immense interest in data visualization. East coast city Boston top the search trend in 2015 then slipped when New York is gaining interest. Chicago newly made to top 10 this year.

There are a decent representation of Asian locations including Singapore, Seoul and cities in India, Bengaluru, Hyderabad and Chennai.

In terms of searching for ‘Data Science‘, Cambridge and Berkeley topped the list in last two years.

Most of the cities, even those just showing once on the list during four years, are in US, which can be attributed to high demand of the role.

Sunnyvale, a city in silicon valley, which houses headquarters of tech company like Linkedin and NetApp, seem to have faded interest in the subject.

Several Indian cities shows up in the list, with either subdued or emerging search volume.

Which topics are trending

In 2015 key word ‘world war II’ gain a lot of traction which could be this data viz film. In 2016 ‘virtual reality’ for Data Visualization is heating up. This year topics like visual analytics, SVG, React are gaining searches, so is Hans’ Rosling’s passing away.

In terms of data science, educational topics hold strong. Among online courses, interests have shifted from Johns Hopkins’ data science specialisation to Microsoft certified professional. Search for postgraduate education almost doubled in 2016. In 2017 AI and deep learning emerge as fastest growing topics.


Data and process

gtrendsR package gives access to Google Trends API and returns a list of dataframes including interests over time, interests by geographical regions and related topics. To take into consideration bigger countries would have more search volume, a search index ranging from 0 to 100 is used to normalize population effect.The API also breaks related topics into ‘top’, ‘rising’ and ‘breakout’.

This will mainly reflect the search interest of English speaking regions as countries which either don’t use google or use other languages in search will not be present in the data.

One issue with the package is it doesn’t take combined terms incorporating other spellings such as ‘visualisation’ (while Google Trend can do that). It mainly affects Australia. When the package updates, we can update.

Related reads

If you are interested in google Trends, you may find these two interesting:

Most frequent how-tos we search for by Xaquín G.V. and collaborators

gtrendsR on r-blogger

On the visual front, these posts are relevant:

How to create a bumpchart by sir viz-a-lot

Order columns within faceted ggplot2


This is #day56 of my #100dayprojects on data science and visual storytelling. If you like it, please share it. Full code on my github. Suggestions and feedbacks are always welcomed. Thanks for reading.


Related Articles