Applying Many-to-Many Alignments and Hidden Markov Models to Letter-to-Phoneme Conversion
- Sittichai Jiampojamarn, Grzegorz Kondrak, Tarek Sherif
- Computer ScienceNorth American Chapter of the Association for…
- 1 April 2007
This work presents a novel technique of training with many-to-many alignments of letters and phonemes, and applies an HMM method in conjunction with a local classification model to predict a global phoneme sequence given a word.
Joint Processing and Discriminative Training for Letter-to-Phoneme Conversion
- Sittichai Jiampojamarn, Colin Cherry, Grzegorz Kondrak
- Computer ScienceAnnual Meeting of the Association for…
- 1 June 2008
The key idea is online discriminative training, which updates parameters according to a comparison of the current system output to the desired output, allowing the model to train all of its components together.
Integrating Joint n-gram Features into a Discriminative Training Framework
- Sittichai Jiampojamarn, Colin Cherry, Grzegorz Kondrak
- Computer ScienceNorth American Chapter of the Association for…
- 2 June 2010
This work includes joint n-gram features inside a state-of-the-art discriminative sequence model for letter-to-phoneme and transliteration transduction and results indicate an improvement in overall performance.
Transliteration Generation and Mining with Limited Training Resources
- Sittichai Jiampojamarn, Kenneth Dwyer, Grzegorz Kondrak
- Computer ScienceNEWS@ACL
- 16 July 2010
DirecTL+ is presented: an online discriminative sequence prediction model based on many-to-many alignments, which is further augmented by the incorporation of joint n-gram features, which shows improvement over the results achieved by DirecTL in 2009.
DirecTL: a Language Independent Approach to Transliteration
- Sittichai Jiampojamarn, Aditya Bhargava, Qing Dou, Kenneth Dwyer, Grzegorz Kondrak
- Computer ScienceNEWS@IJCNLP
- 7 August 2009
DirecTL is an online discriminative sequence prediction model that employs a many-to-many alignment between target and source and is able to independently discover many of the language-specific regularities in the training data.
Online discriminative training for grapheme-to-phoneme conversion
- Sittichai Jiampojamarn, Grzegorz Kondrak
- Computer ScienceInterspeech
- 2009
A manyto-many alignment between graphemes and phonemes is employed, which overcomes the limitations of widely used one-to-one alignments, and the discriminative structure-prediction model incorporates input segmentation, phoneme prediction, and sequence modeling in a unified dynamic programming framework.
A Ranking Approach to Stress Prediction for Letter-to-Phoneme Conversion
- Qing Dou, S. Bergsma, Sittichai Jiampojamarn, Grzegorz Kondrak
- LinguisticsAnnual Meeting of the Association for…
- 2 August 2009
This paper represents words as sequences of substrings, and uses the substrings as features in a Support Vector Machine (SVM) ranker, which is trained to rank possible stress patterns, and advances the current state-of-the-art.
Letter-Phoneme Alignment: An Exploration
- Sittichai Jiampojamarn, Grzegorz Kondrak
- Computer ScienceAnnual Meeting of the Association for…
- 11 July 2010
This work explores several alternative alignment methods that employ phonetics, integer programming, and sets of constraints, and proposes a novel approach of refining the EM alignment by aggregation of best alignments.
Biological Named Entity Recognition Using n-grams and Classification Methods
- Sittichai Jiampojamarn, N. Cercone, V. Keselj
- Computer Science
- 2005
A biological named entity recognition system which uses classification methods and a n-gram model to annotate terms in text to retain simplicity and generalizability is proposed.
Biomedical Term Recognition Using Discriminative Training
- Sittichai Jiampojamarn, Grzegorz Kondrak, Colin Cherry
- Computer Science
We investigate the Perceptron HMM algorithm, an instance of the averaged perceptron approach, which incorporates discriminative training into the traditional Hidden Markov Model (HMM) approach. We…
...
...