You may have noticed a few changes to the discussions feature on the data.world dataset pages.
During the course of our research and user interviews, discussion, documentation, and commentary about data were identified as valuable context that get lost or locked away. We heard phrases like “tribal knowledge” and “answering the same questions over and over again”. We heard stories about sharing the results of a project only to find that others had already done similar analysis. Or having to redo analysis because collection methodology had changed over time — something easily accounted for if the last person who figured that out had a way to communicate it alongside the data itself.
We recently launched multi topic discussions at the dataset level, inspired by the conversations that have naturally sprung up between our users. On many of our popular open datasets, users have been chatting about data provenance, debating interpretation, finding new contributors, and, most interestingly, teaming up to create analysis and visualizations — turning raw data into stories. With such a wide range of topics, organizing discussions in relevant threads will help keep the communication flowing as a dataset gets a wider audience.
Here’s how it works:
On a dataset’s discussion tab, there is a topic list on the left hand side. Each topic has a title, and a category. Today the categories include general for discussion and todo for tracking ongoing work on a dataset.
New topics can also be marked as private, visible only to the contributors for that dataset.