IBM's open source team has updated the Data Asset eXchange (DAX) with free data sets and resources, including weather data and a huge repository of images of fashion items.
IBM launched the DAX last year, hoping to create a hub where developers and data scientists can exchange free data sets, distributed under open data licenses.
Be free, my children
In the past six months, IBM's Center for Open Source Data and AI Technologies added numerous new data-related assets to the exchange.
These include the DAX Weather Project, based off of the NOAA Weather - JFK Airport data set; a Fashion-MNIST Project exploring potential uses for image data of clothing articles; and the Groningen Meaning Bank Project, which can be used to build and train a machine learning model to recognize word entities.
The group added seven new 'Watson Studio notebooks' which show how users can extract, clean, analyze, and model the data using IBM’s Watson Studio commercial platform - one of the reasons the company can afford hosting the data sets for free. Watson Studio starts as a free service, but will inevitably try to upsell you on its ecosystem and tools.
The center also released three 'Watson Studio projects,' which package multiple notebooks together.
IBM is giving away notebooks based on Project Debater from IBM Research, along with Wikipedia Oriented Relatedness and Wikipedia Category Stance, to allow users to explore how to work with text extracted from Wikipedia.
Another project will help users train in the laudable art of water management, while two other, perhaps less admirable, focus on oil reservoir simulations and oil well data.
The updates arrive at a difficult time for IBM, whose profits have been hit by the ongoing COVID-19 pandemic.
Last week it emerged the company was laying off a significant number of employees, including thousands across its Watson, AI, and cloud divisions. IBM currently has no plans to reduce its upcoming shareholder dividend of five percent.