Learning Mixtures of Submodular Shells with Application to Document Summarization
We introduce a method to learn a mixture of submodular "shells" in a large-margin setting. A submodular shell is an abstract submodular function that can be instantiated with a ground set and a set of parameters to produce a submodular function. A mixture of such shells can then also be so instantiated to produce a more complex submodular function. What our algorithm learns are the mixture weights over such shells. We provide a risk bound guarantee when learning in a large-margin structured-prediction setting using a projected subgradient method when only approximate submodular optimization is possible (such as with submodular function maximization). We apply this method to the problem of multi-document summarization and produce the best results reported so far on the widely used NIST DUC-05 through DUC-07 document summarization corpora.
Oct-16-2012
- Country:
- Europe > United Kingdom
- Scotland (0.14)
- North America > United States
- California (0.14)
- Colorado (0.14)
- New York (0.14)
- Europe > United Kingdom
- Genre:
- Research Report (0.50)
- Technology:
- Information Technology
- Artificial Intelligence
- Machine Learning
- Inductive Learning (0.67)
- Statistical Learning (0.46)
- Supervised Learning (0.49)
- Natural Language (1.00)
- Representation & Reasoning
- Optimization (0.68)
- Uncertainty (0.47)
- Machine Learning
- Data Science > Data Mining (0.93)
- Artificial Intelligence
- Information Technology