For a graduate class on Big Data, I wanted to provide my students a hands-on experience of the data acquisition and preparation process to build data sets for analytical modeling to help students to better understand the challenges of messy data and to learn the steps to prepare data for useful analytics.
data.world is a good tool to introduce to students for the topic of data exploration and preparation. The platform is intuitive and allows the students to download and share data easily. It also has user-friendly data summary features, which are important to understand the quality of the data.
A data wrangling assignment on an “Alcohol vs Life Expectancy” dataset was given to my two sections of Big Data class in Fall 2016 to work on individually. This exercise included querying data with SPARQL, for which a detailed document, with step-by-step instructions, was provided.
The platform is intuitive and allows the students to download and share data easily
Following the success of using data.world for the above assignment, we decided to again use data.world for my Big Data class in Spring 2017 of about 40 students. We upgraded the assignment to a half-semester mini-project where students solve a problem using data in a group activity. We chose one of the use cases from the Opportunity Project Challenge to help a US Army veteran find a best school. The students were divided into five groups and given the problem statement and a broad set of criteria to prepare the dataset and do some exploratory analyses. All the groups utilized data.world to download the data, clean them using Python or R tools, and merge them with other public data.
I am teaching two sections of Big Data class again in Fall 2017. I will switch to the “individual assignment” model for Fall 2017 and offer more than one problem statement for students to chose from and work on.
Dr. Chandrasekar Subramaniam used data.world for his graduate class "Big Data Analytics for Competitive Advantage" at UNC Charlotte.
Join the thousands of data people already using data.world to share their data, work together, and build amazing things.Join free!