Experimental results show our proposed approach discovers dramatically more accurate clusters than models without sense disambiguation , and that incorporating global features, such as the document theme, is crucial.

Experimental results show our approach discovers precise relation clusters and outperforms a generative model approach and a clustering method which does not address sense disambiguation .

Without using sense disambiguation , the performance of hierarchical clustering decreases significantly, losing 17% in precision in the pairwise measure, and 15% in terms of B3.

The clusters produced by HAC (without sense disambiguation ) is coherent if all the paths in one relation take a particular sense.

For the sense disambiguation model, we set the number of topics (senses) to 50.

One sense per path (HAC): This system uses only hierarchical clustering to discover relations, skipping sense disambiguation .

2.1 Sense Disambiguation

Selectional preferences discovery (Ritter et al., 2010; Seaghdha, 2010) can help path sense disambiguation , however, we show that using global features performs better than entity type features.

And our sense disambiguation model is inspired by this work.

Our approach employs generative models for path sense disambiguation , which achieves better performance than directly applying generative models to unsupervised relation discovery.

Previous research has conflicting conclusions on whether word sense disambiguation (WSD) systems can improve information retrieval (IR) performance.

Word sense disambiguation (WSD) is the task of identifying the correct meaning of a word in context.

4.1 Word sense disambiguation system