Search Descriptions


Neural machine Translation

Statistical Machine Translation

Search Publications





Transliteration with Other Methods

A number of different machine learning methods have been applied to the transliteration task.

Transliteration With Other Methods is the main subject of 31 publications. 12 are discussed here.

Topics in Transliteration

Transliteration With FSM | Transliteration With Other Methods | Forward Transliteration | Transliteration Training Data | Integrating Transliteration


Improvements have been shown when including context information in a maximum entropy model (Goto et al., 2003; Goto et al., 2004). Other machine learning methods such as discriminative training with the perceptron algorithm have been applied to transliteration (Zelenko and Aone, 2006). Oh and Choi (2002) use a pipeline of different approaches for the various mapping stages. Lin and Chen (2002) present a gradient descent method to map phoneme sequences. Using phoneme and letter chunks may lead to better performance (Kang and Kim, 2000). Li et al. (2004) propose a method similar to the joint model for phrase-based machine translation. Such a substring mapping model has also been explored by others (Ekbal et al., 2006), and may be implemented as bi-stream HMM (Zhao et al., 2007) or adapting a monotone phrase-based translation model (Sherif and Kondrak, 2007). For different scripts for the same language, rule-based approaches typically suffice (Malik, 2006). Ensemble learning, i.e., the combination of multiple transliteration engines, has been shown to be successful (Oh and Isahara, 2007).



Related Topics

New Publications

  • Finch et al. (2015)
  • Ammar et al. (2012)
  • Hagiwara and Sekine (2012)
  • Kuo et al. (2012)
  • Bhargava and Kondrak (2011)
  • Hagiwara and Sekine (2011)
  • Ravi and Knight (2009)
  • Cherry and Suzuki (2009)
  • Li et al. (2009)
  • Li et al. (2009)
  • Durrani et al. (2010)
  • Finch and Sumita (2010)
  • Kirschenbaum and Wintner (2009)
  • Li and Yarowsky (2008)
  • Aramaki et al. (2008)
  • Kuo and Li (2008)
  • Surana and Singh (2008)
  • Liao (2008)
  • Finch and Sumita (2008)