ben.carrasco/batch-speech-utils
1
0
Fork 0
mirror of https://github.com/jambonz/batch-speech-utils.git synced 2026-01-25 02:08:27 +00:00
Code Issues Packages Projects Releases Wiki Activity

schema added for transcription wip

Browse Source
This commit is contained in:
surajshivakumar
2024-06-25 04:32:16 -04:00
parent ef43afca94
commit 4fd5f54dc5
2 changed files with 153 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files

25
lib/schema/transciption-template.js Normal file
View File

@@ -0,0 +1,25 @@
const audioRecordingTemplate = {
'parties': {
'N': 0,
'from': '',
'to': ''
},
'duration': 0,
'url': '',
'conversation': {
'as heard': {
'full transcript': '',
'confidence': '',
'transcription vendor': '',
'timestamps': []
},
'after the fact': {
'full transcript': '',
'confidence': '',
'transcription vendor': '',
'timestamps': []
}
}
};
module.exports = audioRecordingTemplate;

128
lib/schema/transcription_schema.json Normal file
View File

@@ -0,0 +1,128 @@
{
"$schema": "http://json-schema.org/draft-07/schema#",
"type": "object",
"properties": {
"parties": {
"type": "object",
"properties": {
"N": {
"type": "integer",
"description": "Number of parties"
},
"from": {
"type": "string",
"description": "Identifier for the initiating party"
},
"to": {
"type": "string",
"description": "Identifier for the receiving party"
}
},
"duration": {
"type": "integer",
"description": "Duration of recording in milliseconds"
},
"url": {
"type": "string",
"description": "Where recording is located",
"format": "uri"
},
"conversation": {
"type": "object",
"properties": {
"as heard": {
"type": "object",
"properties": {
"full transcript": {
"type": "string",
"description": "Transcript as heard during the conversation"
},
"confidence": {
"type": "string",
"description": "confidence score for transcription as heard"
},
"transcription vendor": {
"type": "string",
"description": "transcription vendor realtime"
},
"timestamps": {
"type": "array",
"items": {
"type": "object",
"properties": {
"word": {
"type": "string",
"description": "Word in the as heard transcript"
},
"startTime": {
"type": "string",
"description": "Start time of the word",
"format": "date-time"
},
"endTime": {
"type": "string",
"description": "End time of the word",
"format": "date-time"
},
"confidence": {
"type": "number",
"description": "Confidence level of the word"
}
},
"required": ["word", "startTime", "endTime", "confidence"]
}
}
},
"required": ["full transcript", "timestamps"]
},
"after the fact": {
"type": "object",
"properties": {
"full transcript": {
"type": "string",
"description": "Transcript generated after analyzing the conversation"
},
"confidence": {
"type": "string",
"description": "confidence score for transcription after the fact"
},
"transcription vendor": {
"type": "string",
"description": "transcription vendor used for after the fact processing"
},
"timestamps": {
"type": "array",
"items": {
"type": "object",
"properties": {
"word": {
"type": "string",
"description": "Word in the after the fact transcript"
},
"startTime": {
"type": "string",
"description": "Start time of the word",
"format": "date-time"
},
"endTime": {
"type": "string",
"description": "End time of the word",
"format": "date-time"
},
"confidence": {
"type": "number",
"description": "Confidence level of the word"
}
},
"required": ["word", "startTime", "endTime", "confidence"]
}
}
},
"required": ["full transcript", "timestamps"]
}
},
"required": ["as heard", "after the fact"]
}
},
"required": ["body", "parties", "duration", "url", "conversation"]
}