@inproceedings{7c2b51431b7e46dabcc4e544641cba26,
title = "Discovering relations between indirectly connected biomedical concepts",
abstract = "The complexity and scale of the knowledge in the biomedical domain has motivated research work towards mining heterogeneous data from structured and unstructured knowledge bases. Towards this direction, it is necessary to combine facts in order to formulate hypotheses or draw conclusions about the domain concepts. In this work we attempt to address this problem by using indirect knowledge connecting two concepts in a graph to identify hidden relations between them. The graph represents concepts as vertices and relations as edges, stemming from structured (ontologies) and unstructured (text) data. In this graph we attempt to mine path patterns which potentially characterize a biomedical relation. For our experimental evaluation we focus on two frequent relations, namely {"}has target{"}, and {"}may treat{"}. Our results suggest that relation discovery using indirect knowledge is possible, with an AUC that can reach up to 0.8. Finally, analysis of the results indicates that the models can successfully learn expressive path patterns for the examined relations.",
keywords = "Biomedical Concepts, Relation Discovery, Text Mining",
author = "Dirk Weissenborn and Michael Schroeder and George Tsatsaronis",
year = "2014",
doi = "10.1007/978-3-319-08590-6_11",
language = "Ingl{\'e}s",
isbn = "9783319085890",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer Verlag",
pages = "112--119",
booktitle = "Data Integration in the Life Sciences - 10th International Conference, DILS 2014, Proceedings",
note = "10th International Conference on Data Integration in the Life Sciences, DILS 2014 ; Conference date: 17-07-2014 Through 18-07-2014",
}