From 6043921067a742fc3f5697b7a86f8396d606aaf4 Mon Sep 17 00:00:00 2001
From: Dave Horton <daveh@beachdognet.com>
Date: Thu, 30 Nov 2023 10:38:33 -0500
Subject: [PATCH] aws stt: calculate transcript-level confidence based on word
 confidence scores

---
 lib/utils/transcription-utils.js | 7 ++++++-
 1 file changed, 6 insertions(+), 1 deletion(-)

diff --git a/lib/utils/transcription-utils.js b/lib/utils/transcription-utils.js
index 62d58b85..6a2e7c4f 100644
--- a/lib/utils/transcription-utils.js
+++ b/lib/utils/transcription-utils.js
@@ -369,7 +369,12 @@ const normalizeMicrosoft = (evt, channel, language) => {
 const normalizeAws = (evt, channel, language) => {
   const copy = JSON.parse(JSON.stringify(evt));
   const alternatives = evt.Transcript?.Results[0]?.Alternatives.map((alt) => {
-    return {transcript: alt.Transcript};
+    const items = alt.Items.filter((item) => item.Type === 'pronunciation' && 'Confidence' in item);
+    const confidence = items.reduce((acc, item) => acc + item.Confidence, 0) / alt.Items.length;
+    return {
+      transcript: alt.Transcript,
+      confidence
+    };
   });
   return {
     language_code: language,