Learn Before
Concept

Automatic Metrics Assessment: String Similarity Metrics

The most commonly used are BLEU and TER.

BLEU (BiLingual Evaluation Understudy) is a precision-oriented metric, which means that it depends on the number of n-grams in the candidate translation that match with n-grams of the reference, independent of position. BLEU values range from 0 to 1 (or to 100); the higher the better.

TER (Translation Edit Rate) measures the minimum number of edits necessary to change a candidate translation so that it matches perfectly to one of the references, normalized by the average length of the references.

Only the reference that is closest is considered for the final score. The edits to be considered are insertions, deletions, substitutions of single words, and shifts (positional changes) of word sequences. TER is an edit-distance metric Equation TER=# of edits/average # of reference words, with values ranging from 0 to 100; lower values are better.

0

1

Updated 2022-08-21

Tags

Data Science