Reveal Help Center

Reveal Language Support

Document Purpose

The purpose of this document is to provide an understanding of how the Reveal Platform handles foreign languages, and example of translating files within review. The foreign language solution of Reveal consists of 6 main features:

  1. Extraction of Electronic Text

    • Extraction of electronic text to UTF-8 Unicode is an automated step during processing within the Reveal platform.

  2. Extraction of OCR Text

    • Reveal supports over 120 different languages during the OCR process. This process handles multiple languages within the same document on one pass. As an example, if a scanned document has Chinese, Japanese, and Korean, Reveal OCR will extract all 3 languages on a single OCR pass. Extraction of OCR text to UTF-8 Unicode is an automated step during processing within the Reveal platform.

  3. Language Identification

    • Reveal supports over 150 different languages during the language identification process. This process uses a neural network artificial intelligent n-gram approach to identify the top 3 languages, and the corresponding percentage of these languages within the entire extracted and/or OCR text of a file. Language identification is an automated step during processing within the Reveal platform.

  4. Language Translation

    • Reveal supports over 20 different languages during the language translation process. This process uses an artificial intelligent approach to translate one language within the text of a file to another target language. Language translation requires input by and Admin after loading to Review. Users can leverage the languages and their percentages from language identification to execute the translation. Please note that translation is not meant to replace the review of files by reviewers in their native language. This feature allows admins and/or review managers to search for potentially responsive material in their native language, to then prioritize the file to a reviewer that is fluent in the file’s original language.

  5. Transcription

    • Reveal supports over 30 different languages during the transcription process (extraction of text from audio/video files). This process uses an artificial intelligent approach to translate one language within the text of a file to another target language. Language transcription requires input by an Admin after loading to Review. Users can leverage the languages and their percentages from the language identification to execute the translation.

  6. Translation of User Interface

Translation

As described above, translation within Review is simply a mechanism to translate one language to another language for searching purposes. As seen below, the extracted text from a Microsoft Word file was translated from Japanese to English. As an automated step in this process, a new viewer is created automatically displaying the translated text, and the translated text is automatically added to its own index for searching purposes.

Native Rendering

The native rendering of the file maintains a high fidelity rendering of the file in the Native / HTML viewer.

Native_Rendering.png
Extracted Text

The extracted text from the file is displayed within the Extracted viewer. This is the text that is passed to translation.

Extracted_Text.png
Translation Viewer

The Translation viewer is created automatically during the translation process to toggle back and forth between the original text versus the translated version.

Translation_Viewer.png
Translation Index

An index of the translated text is created automatically during the translation process, so users can choose to search the translated version of the file. When running a search, a user can choose to have this index searched with all other indexes, or just by itself.

Translation_Index.png