Enhanced NLP Artificial Intelligence – XTM Cloud 12.7
XTM Cloud 12.7 delivers a catalogue of enhancements in NLP Artificial Intelligence.
Import TM, create glossaries all in one step
Terminologists and Project Managers are now able to run bilingual term extraction from files such as TMX or XLIFF. Terms are extracted during the TM import without the need for additional alignment.
Free up your time with automation driven by NLP AI
Inter-language Vector Space tells you how likely a translation is to be accurate. That may sound simple but the process underpinning it is hugely complex, drawing on an exhaustive data search across the entire internet as well as XTM’s vast library of bilingual dictionaries. The result is a mathematically precise calculation that takes simple tasks away from linguists, so human intelligence can be applied in the areas where it makes the greatest impact. XTM Cloud 12.7 refines the algorithms behind this technology still further, freeing up more of your people’s precious time with automation you can trust.
Robust NLP AI automation for language variants
We’ve streamlined the Aligner and Auto-inlines to work effectively in scenarios where the source language is the same as the target language. This adds major value when localizing between variations on the same language, such as UK and American English, European and Canadian French or European and Brazilian Portuguese.
Enhanced handling of stop words
The detection of stop words such as pronouns, prepositions and function words has been improved for the Aligner, Auto-inlines and Auto-bilingual term extraction. Stop words need more elaborate matching mechanisms, applying non-standard translation rules and therefore should not be extracted from terminology listings. XTM Cloud 12.7 addresses this issue directly; we have added brand-new stop words dictionaries.
Improved compound noun splitting
We’ve upgraded the mechanism of noun-splitting for languages including German, Dutch, Danish and Afrikaans and added new linguistic resources. This mechanism fully supports the Aligner, Auto-inlines and Auto-bilingual terminology extraction.
Enhanced support for inline elements in the Aligner
XTM Cloud 12.7 enhances the management of inline elements in the source and target files in the Aligner. Wherever possible, the inline elements in the target segment are matched with the corresponding inlines in the source segment. The benefits of this upgrade become clear when importing the results of the Aligner into the Translation Memory.
Punctuation in NLP AI – the devil is in the detail
XTM Cloud 12.7 has developed a punctuation recognition mechanism used by the Aligner, Auto-inlines and Bilingual Terminology Extractor to handle punctuation characters in all scripts included in the Unicode standard. The new mechanism captures the correspondence between punctuation characters across various scripts, e.g. the Latin full stop “.” and the Japanese full stop “‘。” or the Latin question mark “?” and the Greek question mark “;”.
Clients have repeatedly told us that these small details make a great deal of difference to them, and it’s our pleasure to deliver an upgrade that drills down into their needs as accurately as possible.
Detailed NLP AI improvements:
- The Aligner now has support for the apostrophe character in the starting position in the sentence.
- Improved error handling in the Aligner – any error in the aligning process is now reported in the Aligner results file with a detailed description.
Read the release notes to learn more about the enhancements in XTM Cloud 12.7.
XTM Cloud 12.7 speaks directly to the needs of all users, honing in on the fine details that upgrade a valuable service into an indispensable professional partnership. It makes a very good product even better, and represents a sound investment in your future success.
