A Simple, Fast, and Effective Reparameterization of IBM Model 2
Research Showcase @ Carnegie Mellon University (Carnegie Mellon University)2018pp. 644–648
Citations Over TimeTop 1% of 2018 papers
Abstract
We present a simple log-linear reparameterization of IBM Model 2 that overcomes problems arising from Model 1’s strong assumptions and Model 2’s overparameterization. Efficient inference, likelihood evaluation, and parameter estimation algorithms are provided. Training the model is consistently ten times faster than Model 4. On three large-scale translation tasks, systems built using our alignment model outperform IBM Model 4. An open-source implementation of the alignment model described in this paper is available from http://github.com/clab/fast align .
Related Papers
- → (2021)49,957 cited
- Parallel Data, Tools and Interfaces in OPUS(2012)