Models of translation equivalence among words

By I. Dan Melamed. International Conference on Computational Logic, 2000

316 citations recorded by the source corpus.

View Paper: https://www.semanticscholar.org/paper/38224f0aa39e4d4b9a0060e0fe3941f9e6d1bee1

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

Citations in the corpus

A Syntax-based Statistical Translation Model (2001, background): “, 1996), word alignment of a translation corpus (Melamed, 2000), multilingual document retrieval (Franz et al.”
Squibs and Discussions: Measuring Word Alignment Quality for Statistical Machine Translation (2007, methodology): “Other examples include work relying on combining the Competitive Linking algorithm (Melamed, 2000) with discriminative scoring functions (Cherry and Lin, 2003; Taskar, Lacoste-Julien, and Klein, 2005).”
The Web as a Parallel Corpus (2003, methodology): “Using the enhanced, weighted lexicon as a Dirichlet prior (containing 77,699 entries and a total count of 85,332), a word-to-word translation model (Melamed 2000) was trained on a verse-aligned Bible (15,548 verses, averaging 25.5 English words, 23.4 French words after tokenization).”
Combining Word-Level and Character-Level Models for Machine Translation Between Closely-Related Languages (2012, methodology): “Finally, using S ( m, b ) , we induced a weighted bi-partite graph, and we performed a greedy approximation to the maximum weighted bipartite matching in that graph using competitive linking (Melamed, 2000), to produce the ﬁnal list of cognate pairs.”
A Systematic Comparison of Various Statistical Alignment Models (2003, methodology): “In Yarowsky, Ngai, and Wicentowski (2001), word alignment is used to transfer text analysis tools such as morphologic analyzers or part-of-speech taggers from a language, such as English, for which many tools already exist to languages for which such resources are scarce.”
A Discriminative Matching Approach to Word Alignment (2005, background): “at word alignment, aligning, say, each English word with the French word (or null) with the highest Dice value (see (Melamed, 2000)), simply as a matching-free heuristic model.”
SimAlign: High Quality Word Alignments without Parallel Training Data using Static and Contextualized Embeddings (2020, methodology): “Similar methods have been applied to co-occurrences (Melamed, 2000) (“competitive linking”), Dice coefficients (Och and Ney, 2003) and attention matrices (Garg et al.”