TMSR

Researchers conduct systematic reviews to summarize the scientific literature in a research area. However, the size of the literature grows exponentially, making it increasingly difficult to do so. Text Mining Systematic Reviews (TMSR) overcomes this limitation by extracting knowledge from literature in a transparent, unbiased, and scalable way. This Open eScience project addresses three remaining challenges. First, it develops COSMAS (Centralized Online Scientific Manuscript Accessing Service): a cloud-hosted service that helps researchers access full-text scientific publications they are legally entitled to via their institution, thereby bringing TMSR within reach of all researchers. Second, it implements a pre-processing pipeline within COSMAS to homogenize text- and metadata for TMSR. Third, it connects COSMAS to TMSR methods currently being developed at Tilburg University that extract a causal diagram from a body of literature. This causal diagram can be used to derive testable hypotheses and guide future research.