次へ: Example of Translation Sentences 上へ: Pattern-Based English-Japanese Machine Translation 戻る: Baseline System (Moses)

Experimental Results

We classified the output Japanese translation sentences into four types (The A-rank～ The D-rank). We used the likelihood(β) as a classifier. The four types are shown .

**表 VI:** Four Types for the Sum of Japanese Word Tri-gram Scores
$\scalebox{1.0}[1.0]{ \begin{tabular}{\vert l\vert l\vert} \hline A-rank & -1000... ...＜ β $\le$\ -2000.0\\ \hline D-rank & -3000.0 $\le$\ β\\ \hline \end{tabular}}$ β: Sum of Japanese word tri-gram scores with based 10 logarithm.

Subsections

平成25年9月17日