Reproducible experiments on word and sentence similarity measures for the biomedical domain

This dataset introduces a set of reproducibility resources with the aim of allowing the exact replication of the experiments introduced by our main paper, which is a reproducible experimental survey on biomedical sentence similarity with the following aims: (1) to elucidate the state of the art of the problem; (2) to solve some reproducibility problems preventing the evaluation of most of current methods; (3) to evaluate several unexplored sentence similarity methods; (4) to evaluate for the first time an unexplored benchmark, called Corpus-Transcriptional-Regulation (CTR); (5) to carry out a study on the impact of the pre-processing stages and Named Entity Recognition (NER) tools on the performance of the sentence similarity methods; and finally, (6) to bridge the lack of software and data reproducibility resources for methods and experiments in this line of research.

This dataset sets a self-contained reproducibility platform which contains the Java source code and binaries of our main benchmark program, as well as a Docker image which allows the exact replication of our experiments in any software platform supported by Docker, such as all Linux-based operating systems, Windows or MacOS. Our benchmark program is distributed with the UMLS SNOMED-CT and MeSH ontologies by courtesy of the US National Library of Medicine (NLM), as well as all needed software components with the aim of making the setup process easier.