我正在尝试使用dialogflow来构建自动电话系统（IVR）。以下是对话流程的示例结果。 DialogFlow吐出两个置信分数是两个置信分数，言语识别信念和意图检测信心。
speechRecognitionConfidence的最低阈值是多少？即是否有一个值我们可以肯定地说转录是不准确的？是否有超出最小阈值的范围，结果可能是不准确的（在这种情况下，我想与用户确认他们的输入是否有效）。还有什么用于计算这个分数？
{
"responseId": "a282e405-bc99-4e32-8746-76c79713b30f",
"queryResult": {
"queryText": "Reserve something",
"speechRecognitionConfidence": 0.91684866,
"action": "input.bookFlights",
"parameters": {
"geo-country1": "",
"date": "",
"geo-country": "",
"flight_type": ""
},
"fulfillmentText": "Please specify a date",
"fulfillmentMessages": [
{
"text": {
"text": [
"Please specify a date"
]
}
}
],
"outputContexts": [
{
"name": "projects/tripplanner-94399/agent/sessions/73621660-efb6-4b44-a86f-c6088ed4cf6d/contexts/b8f55a17-3e62-4feb-9c0d-1ded08d98122_id_dialog_context",
"lifespanCount": 2,
"parameters": {
"date.original": "",
"geo-country1.original": "",
"geo-country": "",
"flight_type.original": "",
"geo-country1": "",
"date": "",
"flight_type": "",
"geo-country.original": ""
}
},
{
"name": "projects/tripplanner-94399/agent/sessions/73621660-efb6-4b44-a86f-c6088ed4cf6d/contexts/blookflight_dialog_params_date",
"lifespanCount": 1,
"parameters": {
"date.original": "",
"geo-country1.original": "",
"geo-country": "",
"flight_type.original": "",
"geo-country1": "",
"date": "",
"flight_type": "",
"geo-country.original": ""
}
},
{
"name": "projects/tripplanner-94399/agent/sessions/73621660-efb6-4b44-a86f-c6088ed4cf6d/contexts/blookflight_dialog_context",
"lifespanCount": 2,
"parameters": {
"flight_type.original": "",
"geo-country1": "",
"date": "",
"flight_type": "",
"geo-country.original": "",
"date.original": "",
"geo-country1.original": "",
"geo-country": ""
}
}
],
"intent": {
"name": "projects/tripplanner-94399/agent/intents/b8f55a17-3e62-4feb-9c0d-1ded08d98122",
"displayName": "BlookFlight"
},
"intentDetectionConfidence": 0.48,
"diagnosticInfo": {},
"languageCode": "en"
}
}