Figure 3From: Exploring supervised and unsupervised methods to detect topics in biomedical textTopic Clustering as a function of number of topics (similarity of threshold τ = 0.5.).Back to article page