In our latest blog, Jack from Playgen takes us through the workings of DataPlay in this short video.
DataPlay is a data analysis and exploration game, made for people who are not experts in data analysis or data mining. It’s designed to combine the power of human intuition for pattern recognition with machine data processing, in order to help users extract meaningful stories from data in a fun and intuitive way.
“This is the landing page. What visitors see when they first visit the site. It displays engagement statistics on number of players and their discoveries as well as a preview of some of the data that can be played with after signing in. There’s also an about page, providing an outline of the project. This is the login page. Visitors can use Facebook or Google to login or create a new account.”
“Here we are at the dashboard. Once logged in, the dashboard displays discoveries that require approval, observations by other users that require validation and other recent activity on the platform. The search bar allows users to search for keywords or datasets to explore.”
“The search results page are centered around these visualisations. These are created from the datasets that correspond to the search term, accompanied by preview of their contents. Matching news articles and tweets related to the search are also know. Clicking on one of the visualizations takes you to the overview page for the dataset.”
“The overview page provides access to a large variety of visualisations from the dataset. The top section is focused exclusively on the dataset and any correlations which may exist within the data. The bottom section displays visualisations of correlations of the dataset with other datasets on the platform.”
“This is an example of a related chart, which is a chart with a single data set. Users are encouraged to approve or disapprove the chart based upon whether the data displayed is useful or not. They can also make an observation about all or specific part of the visualisation.”
“If we go back to the Overview screen, and select a chart from the second section, we see a similar page, but for a correlated chart. This chart is one that has 2 different datasets that our algorithms believe have a high-level of correlation. However, algorithms can only go so far, so users can help confirm or deny the presence of any meaningful correlation by approving, disapproving or making observations.”
“In the Navbar, we have a link to the Activity Monitor screen. This page is an overview of what’s happening in the press media, and how it corresponds to datasets in dataplay. Were there a byelection in Hackney, you would see it here. If a worrying report about greenhouse gas emissions came out, it would show up here. Want to know more about it? Clicking any of the keywords makes a search for that term, showing you relevant datasets, news articles and tweets.”
If you haven’t already, you can register at dataplay.playgen.com, and start your data adventure!
