Anonymized Keystrokes Dataset

Includes two datasets for this task, one for English-French (En-Fr) and another for English-German (En-De). For each dataset, the action sequences for full documents are provided, along with an editor identifier. The dataset contains document-level post-editing action sequences, including edit operations from keystrokes, mouse actions, and waiting times.

Source: Translator2Vec: Understanding and Representing Human Post-Editors