Frequently asked questions

TMXCentral was created by translators for translators to make public texts available in a format usable by computer-assisted translation tools.

We have created an automatic alignment algorithm for federal laws and other texts to create TMX translation memories, which can be used in various computer-assisted translation software.

Questions and Answers

Do I need to create an account?

You cannot make purchases as a guest. Account creation is quick and will allow you to access your previous purchases from anywhere.

We do not redistribute your data and will only communicate with you when necessary.

What happens if a product I own is updated?

The purchased product comes with a design date. Any updates after this date will be considered the creation of a new product.

If you wish to obtain the updated product, you will need to purchase it.

Why was TMXCentral created?

TMXCentral was created by former translators who would have loved to have such a repository of quality translations when starting out! Building a good pool of quality texts is challenging, and TMXCentral aims to be the solution to this problem.

Ultimately, our goal is to provide an extensive corpus of public texts organized by theme to facilitate the work of translation companies, both large and small.

How does it work?

You choose the texts you need on the site, add them to your cart, and gain access to the files after payment confirmation.

You can then import these files into your work tools.

How are prices calculated?

File size and organization size determine the price of various products: the larger the text, the more server space and bandwidth it requires; the more employees an organization has, the more it can distribute our products to a large number of translators.

Texts from a departmental corpus are priced lower than those of laws, due to their lower accuracy rate. Laws display an accuracy rate higher than 99%, while corpus texts display a rate of about 95%, given the more chaotic nature of web pages.

We believe our pricing model is fair and allows both freelancers and large organizations to benefit from our products at rates that make sense in their respective contexts.

There's a problem with a file!

TMXCentral's algorithm is designed to intelligently compare and align the provided texts. However, errors may occur.

Laws display an accuracy rate higher than 99%, while corpus texts display a rate of about 95%, given the more chaotic nature of web pages.

If you find an error in one of the files, please report it using the help request form and we will apply the fix or issue a refund if necessary.

Why are public data for sale?

TMXCentral does not sell the data, which are indeed public and available to everyone, but rather the alignment algorithm, online storage, and access from anywhere to the most up-to-date versions of the various aligned texts.

Departmental Corpus: what is it?

TMXCentral crawls public government websites (currently only federal government ones) and extracts the content offered to the public. It then uses its alignment algorithm to create bilingual English-French TMX files (eventually multilingual).

This is a distinct product from aligned laws, consisting of texts targeting public clients, such as press releases, fact sheets, forms, etc. A departmental corpus will generally use more common language and focus on current issues.

What do TMXCentral's TMX files look like?

TMXCentral places all the content of sections and their subsections in the TMX as well as certain metadata, such as the title and web location of the law.

TMX file example

However, sometimes the structure of a section in English is different from that of the French version. For example:

Divergence in laws

In such cases, our algorithm will combine excess subsections at the lowest possible level to match the structure of the law in the other language.

Corrected divergence in laws

Here is a TMX available for download: Family Homes on Reserves and Matrimonial Interests or Rights Act


Regarding departmental corpus texts, extraction is more complex, as much content is excluded to avoid misalignment issues. Thus, some content present on a web page might not be found in our TMX files for this reason.

Which parts of the laws are aligned?

Currently, only the sections and their subsections are aligned. Terms and their definitions placed at the beginning of the acts have been omitted for now as they will eventually be offered in glossaries.

Why are some laws not present?

As TMXCentral uses an automatic alignment algorithm, some laws may generate alignment errors due to special formatting or other particularities. Here is the list of federal laws that are currently not aligned by the system:

  • Bankruptcy and Insolvency Act
  • Agreements and Conventions
  • Employment Insurance Act
  • Export and Import of Rough Diamonds Act
  • Financial Administration Act
  • Global Minimum Tax Act
  • Halifax Relief Commission Pension Continuation Act
  • Appropriation Acts