The headlines that compose HLGD were written by news organizations. We believe their use in this dataset fall under fair use. See this article from the American Bar Association regarding fair use in the news domain.
According to the article, there are four factors to consider for fair use:
-
Has the material been transformed by adding new expression or meaning?
Yes, in our case we put the headline in a timeline (new context), and add labels to the headline (group identity).
-
The purpose and character of the use.
The Headline Grouping project (and dataset) are not for profit and intended to be distributed for research purposes.
-
Amount and substantiality of the portion taken.
The HLGD dataset only distributes the headline and the URL to the original article. The dataset does not distribute pictures, full body content of the article or other information on the source articles.
-
Effect of the use on the potential market.
The dataset consists of at most 426 headlines from a given source, and the most recent headline was published on April 20th 2020 (a year prior to publication). We expect the impact on readership for the news organizations to be minimal.
It is possible for a news organization to opt-out of being a part of the dataset. To start the procedure of being removed from the dataset, please contact [email protected].