Putting dataset in CV

From the Harvard Dataset users guide:

Download Citation

You can find the citation for the dataset at the top of the dataset page in a blue box. Additionally, there is a Download Citation button that offers the option to download the citation as EndNote XML or RIS Format.

So if you go that route just upload your data and click that button.

It seems pretty presumptuous to call it the most important dataset in your field. If it's got a DOI, then you can leave off the editorial. Just give the title and the DOI. An author list might be appropriate if there is a consistent set of authors. If the dataset is continuously evolving or growing, you might leave the authors off. This may depend on the conventions of your field. Also, if there's a marker paper that describes the first use of that dataset, and you're an author on that paper, you should also put that on your CV.

We don't generally subdivide CV entries according to the usual P&T requirements, though if you're in a sub-area that's more remote within linguistics, like computational, then there might be different standards. The standard format of the entry in an academic CV would be:

2015a. Kurdish speech corpus. http://hdl.handle.net/1234.5/678910, Harvard Dataverse, V1

or if co-authored, with the added tag ("w. Y. Matras & G. Haig", or whatever is appropriate). (You can also deposit it at TROLLING).