Transcription & Translation

Breakout 1: Translation of Texts into English

1. Download the following .txt files. Note that the link takes you to a page on your browser; use “save this page as” to download it to your computer.

   • portuguese_mz_text.txt.

   • spanish_text.txt.

   • hindi_text.txt.

   • mandarin_text.txt

STEP 1: Text Translations

Translate the text files with MS Word and Google Docs

1. Google Docs: in-doc translation (create a new google doc, cut-and-paste the text into the Google Doc, save it, and translate)

2. MS Word: in-doc translation.

3. Bonus: Chat-GPT?

STEP 2: Save

Save the translations in your Google Drive in a folder called “class_translations” with the following file names and formats:

1. MS Word: .txt file: MSWord_[original language]_to_english.txt

2. Google: new GoogleDoc: google_[original language]_to_english

STEP 3: Compare

1. Paste the translations into the Text Comparison Website. Note that you may have to adjust the spacing in the text files to make the text align.

STEP 4: Discuss

1. Are there are any differences in the way key words or concepts were translated?

2. Are there circumstances under which these could influence the interpretation of the writer’s intent? Subtleties of language like formality of tense, tone, or meaning?

3. Do any of them seem “better” or “more accurate”?

Breakout 2: TRANSCRIPTION

We will try transcribing audio recorded in different languages using different online services; note that not all languages will be available in all services.

STEP 1. Sign up for an account

Sign up for accounts with the following transcription services. You can cancel the accounts if you think you won’t need them afterwards. Additional details and links to tutorials are available on the Resources page.

• Otter.ai

• Happy Scribe

  • To explore: Google Text-to-Speech and Watson; note the “detect multiple speakers” option.

STEP 2: Transcribe

1. Download the following audio files. Click the blue ‘view raw’ text or download icon to save it to your computer.

   • portuguese-mz-audio.mp3

   • hindi-audio.mp3

   • spanish_audio1_voice.mp3

   • mandarin1_audio.mp3

2. Now transcribe these audio files with each of the different services (note that not all services will transcribe all languages).

3. Save the resulting transcriptions files as Google Docs or .txt files. Use the following naming convention: transcription_[language]_[name of service]

STEP 3: Translate

1. Translating these transcriptions to English with Google Docs

2. Save with the following file names: translation_[language]_[name of transcription service] (e.g., translation_spanish_happyscribe)

3. Paste the translations from transcription and the translation from the original .txt files (used in the translation breakout) into the Text Comparison Website. You may have to adjust the spacing in the text files to make the text align.

STEP 4: Discuss

1. Do the transcriptions appear accurate?

2. What about the translations from transcriptions? Is there a loss in quality when going from audio-to-text-to-English when compared with text-to-English

3. What would be the best approach for research:

   a. transcribing, editing the transcription, then translating, or b. transcribing, translating, then correcting?

Submission

When done please submit via Canvas:

1. the translations from part 1,
2. the transcriptions from part 2, and
3. your responses to the discussion questions