Transcript - Docs

The data dictionary and definitions for Transcripts

Fields

Parameter	Description
transcript_item_id	The Aiera-unique identifier for the transcript
event_type *	The type of transcript
transcript	The unadjusted transcription for this segment
timestamp	The start timestamp for this transcription
start_ms	The millisecond offset for the start of this transcription relative to audio
duration_ms	The millisecond duration for this transcription segment
speaker_id	The Aiera-unique identifier of the speaker
speaker_type	Machine-generated ("dia" or "id") or published ("final") speaker type
speaker_name	Name of the speaker
speaker_title	Title of the speaker
transcript_section	Name of the section to which the transcript segment belongs ("presentation" or "q_and_a")
annotated	The tag-annotated transcription (deprecated)
audio_url	The audio excerpt URL for this transcription
is_improved	Whether this transcript has been improved by larger secondary ASR
is_edited	Whether this transcript has been human-edited/reviewed
modified	The last modified date for this transcription
linguistics *	JSON object composed of linguistic-extractions, like summaries, etc.

Possibles

event_type

Option	Description
transcript	A live/raw transcription
official_transcript	A human-corrected transcription

linguistics

{
  "summaries": [
    {
      "title": SUMMARY TITLE,
      "summary": [
        SUMMARY PARAGRAPHS
      ],
      "model": (zeroshot|bullet),
      "type": SUMMARY TYPE (presentation|q_and_a|everything),
      "audio_clip": AUDIO CLIP URL,
      "video_clip": VIDEO CLIP URL,
      "priority": SUMMARY PRIORITY ORDERING,
      "created": SUMMARY CREATION DATE,
      "modified": LAST SUMMARY MODIFICATION DATE
    }, ...
  ],
  "topics": [
    {
      "topic_id": AIERA-UNIQUE TOPIC ID,
      "topic": TOPIC,
      "priority": PRIORITY ORDERING
    }, ...
  ],
  "auto_tags": [
    TAG, ...
  ],
  "sentiment": {
    "average": AVERAGE SENTIMENT SCORE (-1 to 1),
    "median": MEDIAN SENTIMENT SCORE (-1 to 1),
    "minimum": MINIMUM DETECTED SEGMENT SCORE,
    "maximum": MAXIMUM DETECTED SEGMENT SCORE
  }
}