Scientific Publications in Water Research as Test Collection for Information Retrieval

We document the creation of two datasets consisting of scientific publications in the interdisciplinary field of water research, namely ThWIC-Chemistry and ThWIC-Hierarchy. Additional labels regarding the domain and particular attributes allow for measuring the topical diversity and the attributes' representation in data subsets.

Dataset components:

corpus: DOIs for the respecive publications

queries and information needs: from the chemistry domain

single relevance judgments

The ThWIC-Hierarchy dataset contains a topic hierarchy for keywords in water research, hierarchy labels for the corpus documents, and attribute labels additionally.