Word alignment supports applications such as spell-checking and learning word order and morphology. Even more directly, it indicates the strength of translation consistency. With its color-coded word alignment strengths, the Greek Room specifically helps to pinpoint words that are unusual translations, or missing or spurious words. Hovering with the mouse over a word in the source or target language will reveal additional information such as the glosses of a word.

Examples from English-German and English-Hindi

• Word alignment visualization for English NRSV and German LU84NR06 (selection)
• Word alignment visualization for English ESVUS16 and Hindi IRV (selection)

Color and text-decoration codes

• Red indicates an unreliable link. This might indicate a translation problem, but could also just reflect a rare word or an unusual translation.
• Copper rose indicates a link of mixed reliability.
• Green indicates a fairly reliable link.
• Orange indicates that Greek Room has determined a phrase to be spurious (with respect to the other translation).
• Blue indicates a link bolstered by phonetic string similarity (not used if the link is already highly reliable otherwise).
• Light blue indicates that a link was added (in mouse-over hover mode).
• Purple indicates that a link was deleted (in mouse-over hover mode).
• An underscore indicates that a word is not aligned.
• A green underscore indicates that a word is not aligned and that the word is often not aligned (“no worries”).
• A double-underscore indicates that a word alignment was updated.
• A dashed-underscore indicates that a word is aligned to multiple non-contiguous words on the other side.
• The ⟡ symbol to mark spell-checking information (use mouse-over).

Abbreviations

When you hover over a word, additional information becomes available. Below are abbreviations that occur in the pop-ups.

• gloss: a fairly literal translation of the word, without considering context
• NULL: this indicates that the word often does not correspond to a word in the other language
• c: count (not counting present word)
• p: probabilities (for both directions)
• jc: joint count (of word and its aligned word, not counting present link)
• fw: function word (e.g. and and the)
• sed: smart (phonetic) edit distance; the closer the sed is to 0, the closer the two words are phonetically
• s: score
• rom: romanization (conversion to Latin alphabet)
• []: number in square brackets indicates word position starting at 0: [0] = first word, [1] = second word etc.