Producing and Checking Amazon Transcribe Transcripts




Albin O. Kuhn Library & Gallery - Staff Wiki


Producing and Checking Amazon Transcribe Transcripts

Producing Transcripts

To produce transcripts, you have to have access to the Library’s AWS account. Tim Champ in campus IT can give people access.

  1. Go to https://awslogin.umbc.edu/, https://awslogin.umbc.edu/

  2. Click on UMBC - Library _ Transciption

  3. Click on LibraryTranscriptionRole to login via MyUMBC single sign on.

  4. Click on S3

  5. Upload the file or files that need transcripts into the upload folder and wait a bit.

  6. Go to the download folder, and the folder for the date when you uploaded the the file.

  7. Download the .docx

Checking Transcripts

  1. Open the Word document. Change to a better 12 point font.

  2. Go through the transcript and do the following, listening the to Podcast when necessary to help, Pay special attention to words highlighted in yellow as Amazon has identified that they may be incorrect. Note that you can easily go to when the word(s) were spoken using the timestamp at the beginning of when they began speaking.

    1. Correct punctuation. For example, “Using the pretext of a crime emergency. And fighting off attacks from his own political base over the release of the Jeffrey Epstein files. President Trump on Monday announced the federal takeover of D.C.'s police department and removal of homeless people from the district.” should be one sentence, so you should remove periods or change them to commas to make it one.

    2. The name of the producer of the I Hate Politics podcast is Sunil Dasgupta. Find and replace all misspelling of his name.

    3. Find and replace I Hate Politics to make it all upper case.

    4. Find and replace washington dc with Washington D.C--city names, county names, street names, organization names should all be upper case as well as the first first letter of the words of spelled out acronyms, eg. Parent Teacher Association, PTA.

    5. Add headings 2’s for the following sections as appropriate: Introduction, Advertisement, The Beginning of Each Interview (Interview with____), Topic Heading in H3 for significant topic changes within an interview, Summary, Conclusion, Credits.

    6. Correct the spelling of people’s names by googling them to find the correct spelling.

    7. Indicate music by typing in music when it occurs. If there are lyrics, add it after Music:

    8. Speakers given as Speaker 1, Speaker 2, Speaker 3, should be replaced with the name of the speaker using find and replace as you learn who each speaker is.

    9. Check and correct ULR and change URL’s to a link that displays the name of the website that they go to, e.g.

    UMBC at Shadygove instead of  shadygrove.umbc.edshadygrove.umbc.edu.

  3. Remove everything that Amazon Transcribe inserted at the beginning and end of the file.

  4. Add the title as a heading formatted like this:  I Hate Politics Episode 232: The Challenges of Balancing Taxation and Spending and make it link to the podcast.

  5. Save removing everything from the filename between the first and last “.” and then convert to a pdf.


Albin O. Kuhn Library & Gallery . University of Maryland, Baltimore County . 1000 Hilltop Circle . Baltimore MD 21250
(410) 455-2232. Questions and comments to: Web Services Librarian