Corpus ID: 2762657

Normalized (pointwise) mutual information in collocation extraction

@inproceedings{Bouma2009NormalizedM,
  title={Normalized (pointwise) mutual information in collocation extraction},
  author={G. Bouma},
  year={2009}
}
In this paper, we discuss the related information theoretical association measures of mutual information and pointwise mutual information, in the context of collocation extraction. [...] Key Method We introduce normalized variants of these measures in order to make them more easily interpretable and at the same time less sensitive to occurrence frequency. We also provide a small empirical study to give more insight into the behaviour of these new measures in a collocation extraction setup.Expand
595 Citations
Clustering-based Approach to Multiword Expression Extraction and Ranking
  • 1
  • PDF
Improving Pointwise Mutual Information (PMI) by Incorporating Significant Co-occurrence
  • 14
  • PDF
Evaluating Topic Coherence Using Distributional Semantics
  • 161
  • Highly Influenced
  • PDF
Discovering multiword expressions
  • PDF
Comparing Similarity Measures for Distributional Thesauri
  • 10
  • Highly Influenced
  • PDF
...
1
2
3
4
5
...

References

SHOWING 1-10 OF 21 REFERENCES
Comparative Evaluation of Collocation Extraction Metrics
  • 73
  • PDF
An Evaluation of Methods for the Extraction of Multiword Expressions
  • 69
  • PDF
Reference Data for Czech Collocation Extraction
  • 18
  • Highly Influential
  • PDF
A Lexicographic Evaluation of German Adjective-Noun Collocations
  • 13
  • Highly Influential
  • PDF
AMachine Learning Approach to Multiword Expression Extraction
  • 77
  • Highly Influential
  • PDF
Word Association Norms, Mutual Information and Lexicography
  • 4,070
  • PDF
The Statistics of Word Cooccur-rences: Word Pairs and Collocations
  • 535
  • Highly Influential
  • PDF
Europarl: A Parallel Corpus for Statistical Machine Translation
  • 3,116
  • Highly Influential
  • PDF
Accurate Methods for the Statistics of Surprise and Coincidence
  • 2,705
  • PDF
Corpora and collocations
  • 263
  • Highly Influential
  • PDF
...
1
2
3
...