Description of 6 tasks and each task asks users to look for relevant datasets from the dataset repository.
Collected queries launched by users in order to finish the tasks described in "task_desc.csv". Each task has 20 queries.
Annotated task-dataset pairs. "data_path" is the relative path from the decompressed dataset directory(will be released later) to the dataset file. "score" is the average task-dataset relevance score from annotators. "data_url" is the origin link to download this dataset.
The collection can be downloaded at http://www.cse.lehigh.edu/~brian/data/ecir2020_datagov.zip.