Indirect-HMM-based Hypothesis Alignment for Combining Outputs from Machine Translation Systems

by Xiaodong He , Mei Yang , Jianfeng Gao , Patrick Nguyen , Robert Moore

Citations:

18 - 2 self

Datum	Value	Source
TITLE	Indirect-HMM-based Hypothesis Alignment for Combining Outputs from Machine Translation Systems	SVM HeaderParse 0.2
AUTHOR NAME	Xiaodong He	SVM HeaderParse 0.2
AUTHOR AFFIL	Microsoft Research	SVM HeaderParse 0.2
AUTHOR ADDR	One Microsoft Way; Redmond, WA 98052 USA	SVM HeaderParse 0.2
AUTHOR NAME	Mei Yang	SVM HeaderParse 0.2
AUTHOR AFFIL	Dept. of Electrical Engineering; University of Washington	SVM HeaderParse 0.2
AUTHOR ADDR	Seattle, WA 98195, USA	SVM HeaderParse 0.2
AUTHOR NAME	Jianfeng Gao	SVM HeaderParse 0.2
AUTHOR NAME	Patrick Nguyen	SVM HeaderParse 0.2
AUTHOR NAME	Robert Moore	SVM HeaderParse 0.2
AUTHOR AFFIL	Dept. of Electrical Engineering; University of Washington	SVM HeaderParse 0.2
AUTHOR ADDR	Seattle, WA 98195, USA	SVM HeaderParse 0.2
ABSTRACT	This paper presents a new hypothesis alignment method for combining outputs of multiple machine translation (MT) systems. An indirect hidden Markov model (IHMM) is proposed to address the synonym matching and word ordering issues in hypothesis alignment. Unlike traditional HMMs whose parameters are trained via maximum likelihood estimation (MLE), the parameters of the IHMM are estimated indirectly from a variety of sources including word semantic similarity, word surface similarity, and a distance-based distortion penalty. The IHMM-based method significantly outperforms the state-of-the-art TER-based alignment model in our experiments on NIST benchmark datasets. Our combined SMT system using the	SVM HeaderParse 0.2
CITATIONS	26 found	ParsCit 1.0