VeriDark DarkReddit+ Authorship Identification Dataset

Due to ethical concerns regarding the potential misuse of our benchmark, we require the following additional information for granting permission to use our datasets:

1. The name of the person requesting access, together with their affiliations, job title and an e-mail address. If the person holds an institutional e-mail address, we strongly recommend using it instead of a personal e-mail address.
2. The intended usage for the dataset.
3. An acknowledgement that the dataset will be strictly used in an ethical manner. Non-ethical uses of the dataset include, but are not limited to:
   • using the datasets for the task of Language Modeling or similar generative algorithms.
   • building algorithms that could aid criminals to evade law enforcement organizations.
   • building algorithms that have the aim of unmasking undercover law enforcement agents.
   • building algorithms that could interfere with the activity of law enforcement agencies.
   • building algorithms that could lead to violating any article of the United Nations Universal Declaration of Human Rights.
   • building algorithms with the purpose of exposing the identity of reporters, individuals in the political realms, leakers, whistleblowers, dissidents, or other persons who are seeking to express an opinion about what they perceive is a particular injustice in the world, without regard to what that injustice may be.
   • building algorithms that can help entities discriminate, or exacerbate bias against other persons on the basis of race, color, religion, gender, gender expression, age, national origin, familiar status, ancestry, culture, disability, political views, sexual orientation, marital status, military status, social status, or who have other protected characteristics.

We strongly encourage the inclusion of an ethical statement and discussion in any work based on this dataset.

We do not encourage the distribution of the dataset in its current form to any other parties without our consent.

DISCLAIMER: Any personal information provided when requesting access to the dataset will be used just for deciding whether access to the dataset should be granted or not. We will not disclose your personal data.