Prospector, the Source English Terminology Extraction Tool, Is Updated

Logrus Global Logo

The Logrus Global company has significantly updated Prospector, the terminology extraction tool confirmed as best-in-class by an independent study.

Continuous terminology updating is very important. Terminology follows research, development, and user experience, is extensive and evolves. With new terminology extraction and glossary maintaining tools, keeping corporate subject matter terminology up to date has become easier. The typical terminology extraction use cases include both creating glossaries from scratch for a new domain, and updating corporate glossaries with new content.

Maintaining terminology is difficult, but we made it easier. You need established terminology to maintain subject matter language throughout all content processes, including authoring, review, custom NMT and more. It is now easier than ever – with quick manual verification of term candidates on the fly the junk is reduced from very low to zero.

Terminology is key for custom-trained NMT. Modern NMT engines support integration with external user glossaries much better than even a year ago. Now you can force already trained NMT engine to follow the terminology on the fly without updating the training corpora with new terms and re-training engine each time you update the glossary.

Not simply another term-extraction utility: The quality of such a tool is defined not only by how well it recognizes terms but also by the amount of junk it extracts, the more recognition and less junk is better. Prospector’s monolingual English source terminology extraction engine is based on a unique algorithm developed by a team of in-house experts and uses the world’s largest Corpus of Contemporary American (COCA) English language from Brigham Young University.

How does Prospector stack up against other utilities? There are several key differentiators:

  • High recall: Search completeness or a fraction of true terms found
  • Low noise: A fraction of false positives or items that aren’t actual terms
  • Incremental term extraction: If you have a glossary for a specific domain, Prospector will separate term candidates into two term lists in the output report: new terms (missing from the glossary) and existing terms (already in the glossary).
  • Manual term candidates’ approval in context: Prospector has a term selector and term context visualization functionality. For each new term candidate the linguist can see the paragraph or sentence where it was found in the source text and then manually exclude false positives or approve true term candidates instantly with nothing more than a browser.
  • Files or URLs: You can extract terms not only from the single file but also from the web page. Supported formats for source content include: Text files (.txt); web pages (.html, .htm); Microsoft Word documents (.docx, .doc); and localization interchange files (.xlf, .xliff, .sdlxliff)

The result is an Excel spreadsheet with several term lists: new terms; terms already existing in available glossaries; named entities; and acronyms.

An independent expert has tested Prospector and compared it against its peers, now the noise is zero. For details, see:

Various usage and deployment scenarios are available for Prospector as a cloud-based service, from simply using it in Logrus Global Language Cloud to deployment behind private corporate secure perimeter.

Prospector also offers full control of access rights for users with different roles.

About Logrus Global

We have a quarter-century of experience as a powerhouse in globalization for the world’s most demanding multinationals along with national leaders, midsize firms and startups.

The Logrus Global team has a reputation for completing seemingly impossible projects. With a broad command of technologies, we deliver outstanding language quality and superior service.

Our human experts and high-quality translators and editors use the most innovative tools. Human translation quality is our imperative.

The software developers and engineers at Logrus Global have the knowledge, experience and tools with which to solve all kinds of technical problems. That’s why our clients trust us for the delivery of expansive projects even under challenging time constraints. Our partners are as resilient, scalable and reliable as we are.

Logrus Global is certified for compliance with the ISO 17100, ISO 18587 and ISO 9001 standards.

For more information, visit

About Logrus Global Localization Cloud

Prospector is a member of the Logrus Global Localization Cloud toolset, a family of cloud and desktop tools and services for translation industry. For more information, visit