Semi-Supervised Learning to Identify UMLS Semantic Relations

By Yuan Luo, Ozlem Uzuner. 2014

9 citations recorded by the source corpus.

The UMLS Semantic Network is constructed by experts and requires periodic expert review to update. We propose and implement a semi-supervised approach for automatically identifying UMLS semantic relations from narrative text in PubMed. Our method analyzes biomedical narrative text to collect semantic entity pairs, and extracts multiple semantic, syntactic and orthographic features for the collected pairs. We experiment with seeded k-means clustering with various distance metrics. We create and annotate a ground truth corpus according to the top two levels of the UMLS semantic relation hierarchy. We evaluate our system on this corpus and characterize the learning curves of different clustering configuration. Using KL divergence consistently performs the best on the held-out test data. With full seeding, we obtain macro-averaged F-measures above 70% for clustering the top level UMLS relations (2-way), and above 50% for clustering the second level relations (7-way).

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

1. Zexian Zeng
2. Yuan Luo
3. Yu Deng
4. Ozlem Uzuner
5. Peter Szolovits
6. Xiaoyu Li
7. Tristan Naumann
8. Liang Yao

Citations in the corpus

• Natural Language Processing for EHR-Based Computational Phenotyping (2018, background): “Understanding these relations is important for knowledge representation and inference to augment structured knowledge bases [182, 183].”
• Bridging semantics and syntax with graph algorithms —state-of-the-art of extracting biomedical relations (2015, background): “In addition, it is useful but challenging to filter out named entity tuples (including pairs) that do not have relations explicitly stated in the text [142].”
• Bridging semantics and syntax with graph algorithms - state-of-the-art of extracting biomedical relations (2017, methodology): “In addition, it is useful but challenging to filter out un-related named entity pairs, for which no relations have been explicitly stated in the text [92].”