VIDEO DESCRIPTION:

Play, download and edit the free video [VIS19 Preview] Examining the Visualization Practices of Data Scientists on Kaggle (poster).

Poster

Authors: Md Naimul Hoque , Darius Coelho , Klaus Mueller

Abstract: Kaggle, founded in 2010, has become the leading platform for data scientists to collaboratively explore and build data-based models, participate in competitions, and communicate with each other mostly through interactive notebooks and forum discussions. The growing community has a large userbase with 2.9M current users across 194 countries. This large community regularly produces a large number of solutions (Kernels) for problems and datasets posted on the website. These solutions tend to be of high quality as they are often commissioned by companies like Google, LinkedIn through competitions where users can win prizes for building the best data-based models. Visualization, an integral part of data science, is employed in a large portion of these kernels to either explore data or present results. In this project, we examine the content of these kernels to understand the visualization practices among Kaggle data scientists. Our work reveals insights about the libraries used, the most popular visual representations and the types of color palettes used by these data scientists.

Download, play and edit free videos and free audios from [VIS19 Preview] Examining the Visualization Practices of Data Scientists on Kaggle (poster) using RedcoolMedia.net web apps