Support and Documentation

Amazon Transcribe

Amazon Transcribe is an automatic speech recognition (ASR) service provided by Amazon Web Services (AWS) that allows developers to add speech-to-text capability to their applications. With the Amazon Transcribe plugin, you can easily generate a transcript of an audio or video file that you uploaded to Brightspot. Editors can then perform a text search to find the file and use it in their own content.

Configuring Amazon Transcribe

Administrators perform this task.

Procedure. To configure Amazon Transcribe:
  1. From the Navigation Menu, select Sites & Settings.

  2. In the Sites widget, select Global. The Edit Global widget appears.

  3. Under Main, expand AWS Transcribe.

  4. Turn on Enable Transcribe Service.

  5. In the SQS Queue Name field, enter your Amazon SQS queue name.

  6. In the Region field, enter your AWS region, or leave the field blank to use the region in your Beam configuration.

  7. Under Transcribe Settings, click |mi-search|. A content picker appears.

  8. Select an existing transcription setting, or create a new setting by doing the following:

    1. Under Create, click New Transcribe Settings. A New Transcribe Settings content edit page appears.

    2. In the Name field, type a name for the setting.

    3. From the Language Code list, select one of the available languages, or select Default to use US English.

    4. To accommodate more than one voice in an audio file, turn on Enable Multiple Speakers.

    5. If you want to share this setting with your other sites, in the Sites widget, do the following:

      1. From the Owner list, select None.

      2. From the Access list, select All Others.

    6. Click Save.

    7. Click Back to Select Transcribe Settings. You return to the content picker.

    8. In the content picker, click the transcription setting you created. You return to the Edit Global widget.

  9. Click Save.

Transcribe is configured. When editors upload audio and video files as described in Submitting a file to Amazon Transcribe, Brightspot sends the file for transcription.

Submitting a file to Amazon Transcribe

Editors perform this task.

You can use Amazon's Transcribe service to generate transcripts of files you upload to Brightspot. You can submit any file that contains an audio track, such as podcasts, MP3s, and videos.

Procedure. To submit a file to Amazon Transcribe:
  1. Create a new audio, podcast, video, or other item that has an audio track.

  2. From the Provider list, select AWS Transcribe. A form appears.

  3. Click |mi-add_circle_outline|, and do one of the following:

    • Under File, select New Upload, and navigate to a file on your laptop.

    • Under File, select New URL, and paste the URL of a file.

  4. Toward the right of the widget, select |mi-more_horiz| > Advanced.

  5. Expand Transcribe, and select one of the transcription settings. (If you need transcription settings different from those listed, contact your Brightspot administrator.)

  6. Complete the remaining fields in the content edit form.

  7. Click Publish.

Brightspot sends the file to Amazon Transcribe for transcription, and receives the text a few minutes later. When editors search for keywords in the search panel, Brightspot includes in the search results those items whose transcripts include the keywords. For example, if you have a transcription of a podcast about the US Senate, editors searching for the term senate see the podcast in the search results.

Viewing and adding transcripts

Editors perform this task.

You can view a transcript of an audio or video file. You can attach your own transcript to a file; this feature is useful if you want to annotate the original transcription.

Procedure. To view and add a transcript:
  1. Search for and open the item in the content edit page.

  2. Expand AWS Transcribe. A list of transcripts appears.

  3. Click |mi-expand_more| to view the entire transcript.

  4. To add a transcript, do the following:

    1. Click |mi-add_circle_outline|. A form appears.

    2. In the Text field, type or paste the transcript.

    3. Click Publish.