https://www.kaggle.com/, Data Science and Machine Learning platform provides large number of datasets. Are they open-source? Can we use them for our research publication?
Kaggle is a website that holds competition on regular basic and provides reward cash price for best ML model developers that can solve specific industrial problems.
But yes, you can register in koggle site as a participant and download their data for personal learning or make submission to koggle if you have solve the problem provided by the site.
I don't think we can use such data for publications as it comes under the agreement between koggle website and their client (those are providing their personal data for improved model development). But you can ask for website owner's permission if you really want to use the data sets by writing a mail to kaggle.