The difference between a translation memory and term base (and how to use them correctly)
)
Localization pros are familiar with the difference between a translation memory and a term base (glossary) as they work with these assets daily. People new to the localization and translation field might struggle to see these two translation tools' slight but essential differences.
This blog post from Lingohub sheds light on the difference, usage, and benefits of translation memory and term base features. It describes how they can support your business during the localization. Let's go!
What is a translation memory?
A translation memory is a database of previous translations characterized by the following functionalities:
- It automatically learns from every word you translate along the path to a localized product.
- A translation memory learns in real time.
- Once you save a translation, the translation memory memorizes it inevitably.
- You can use the files with previous translations, and the system will use them to provide suggestions based on this data.
- When translators work on similar texts in the future, the translation memory suggests translations based on similarity (fuzzy or exact matches).
To make this question more precise, with Lingohub, you can create virtual or file-based translation memory for your projects. The first (virtual) automatically learns from your translations in real time. In contrast, the second option implies importing the files with approved translations to teach a translation memory on your examples.
)
What is the matching in the translation memory?
Linguists often work with similar texts/phrases/segments in localization. The translation memory continuously compares the text translators working on with the data in the base. In case of a match, the system suggests the appropriate translation. There are two types of matching:
-
Fuzzy matches. Fuzzy matches are suggestions from the translation memory system where texts match approximately (in Lingohub, between 55% and 99% matches). They suit a given text in terms of meaning and spelling. Translators benefit from using fuzzy matches as a starting point to reduce the effort needed for the final translation.
-
Exact matches. As the phrase indicates, exact matches suit the given text exactly (in Lingohub, 100% matches). Exact matches are provided if the source texts of a past and future translation match completely. Thus, you can apply such exact matches without further changes.
What is a term base?
A term base or glossary is a centralized compilation of industry-specific, company/product-related, and technical terms used for terminology management and shared within translation teams.
A term base grows and changes alongside company growth and contains various data. In Lingohub, there are:
- source texts (terms),
- approved translations,
- terms that are not translated but kept in the source language (e.g., brand or product names),
- advanced information (context, usage status, grammatical gender, term type, etc.),
- case sensitive and matching rules.
)
Term base: case sensitiveness and matching explanation
To make the glossary suggestions even more accurate, we provided the ability to set up the case sensitivity and matching for each term. The "case sensitive" setting allows choosing one of the three values:
- Yes - means that the term base will recognize words that are strictly typed in the very same. For example, for "Company," the system will recognize "Company" but will ignore "company."
- Permissive - the term base is sensitive to capital letters only. For "COMPANY" Lingohub will recognize "COMPANY," "Company," or "COMpany," bit will ignore "company."
- No - the term base entries are insensitive. For "Company" Lingohub will recognize "COMPANY," "Company," "company," etc.
The matching type determines how the matching words will be recognized. There are four options:
- Fuzzy - hits are getting independent of prefix or suffix matching. For "Mutter" it also finds "Mütter."
- Prefix - is a default setting. The system recognizes the prefix matching. For "company," it will find "companies."
- Exact - find only 100% matches. In the case of "company," Lingohub will find only "company."
- Custom - the ability to specify by pipe or asterisk the matching rules. For example "bankovn|í operac|e" will retrieve "bankovních operací" and "bankovními operacemi" as well.
)
Read our help center to get more information about this topic.
How to use a translation memory?
Your personal translation memory is automatically created in line with creating a localization project on Lingohub. The system adds all saved translations without assistance. If there is an existing translation memory from previous work, you can easily import it into Lingohub.
There's no need to update a translation memory manually. Source texts and corresponding translations are paired and hosted centrally. Once you edit a translation Lingohub automatically updates the language pair in your translation memory system too. Also, you can create multiple translation memories with custom settings to provide custom suggestions for specific projects.
Benefits of a Translation Memory
Reduce the time and costs
The translation memory searches for matching strings when translators work on new texts. Using exact matches in translations can help reduce costs as it allows for the reuse of translations without any extra effort. As a result, the Per-word translation costs decrease.
At the same time, you can use the translation memory together with the prefill feature. Lingohub's prefill functionality automatically reuses matches from the translation memory for empty text segments. You can perform prefill or set up specific rules. Translators don't even have to reuse translations manually.
How can it work? Lingohub checks empty text segments in German (Austria) and automatically copies texts from German. Of course, this works for any language combination you specify. As a result, the translation costs shrink because the system performs much translation work to relieve your translators and project managers. Translation matches avoid unnecessary and redundant work because translators don't translate any text twice.
)
Increase text quality
Translation Memories are keepers of text quality and foster high-quality translations. Translators will translate matching phrases identically to ensure textual and vocal uniformity. They maintain linguistic consistency across past and future translations even if various translators work on your localization projects. You can choose segments with which status will be added to the translation memory. Select only "Approved" segments and save much more time for text checking and approving.
How to use a term base?
When you create a term base, you need to identify key terminology within your company. Using final source texts, approved translations, and carefully researched context information is essential to create a good term base to support your entire team. You can create a term base manually by adding entries or importing all the data like TBX or CSV files.
Benefits of a term base
Keeps consistency through all content
The term base can be a single source of truth, besides translators, for sales and marketing teams and external linguists. With it, you can be sure that your teams communicate with users via precise and the same terms throughout your web presence. Provide as much context information as possible - images, description, etc. to bring your team a powerful tool.
Correct usage and spelling of corporate terminology
It's in the very nature of language and translation that they're subjective. A term base guarantees the correct spelling of product or company names likely case-sensitive (e.g., Lingohub). It also informs translators about words or phrases that are not to be translated but kept in the source language. "A term base improves text quality through consistent usage and spelling of brands and corporate terms."
Accelerate translation speed
A well-kept term base eliminates the need for terminology research and reduces turnaround time. The translators do not need to search for a precise term through files. The localization process with such an approach (when the system automatically provides the correct translation) allows linguists to stay in a single interface and not switch between tabs/files/etc., saving much time.
Comparison sum up
A translation memory and term base are only a part of the CAT tools provided by Lingohub. This duet affects the localization speed and quality by automating manual processes and providing already approved translations. If we talk about differences - both solutions can be managed manually - but the translation memory has an option of virtual learning.
Going with both will reduce the localization cost and allow your translators to focus on the text adaptation without recurring translation of the same text twice, thrice, etc. If you want to try these powerful features - Lingohub will gladly help you. You can schedule a demo with our team or register an account and try Lingohub for free first 14 days!
Accelerate your translation speed, cut costs, improve linguistic consistency and quality, and support your team with Lingohub. Let's grow together!