Kuiz berkaitan panduan · Sederhana Tahap

AI dalam Kuiz Analitis Pertuturan Pusat Panggilan

Uji pengecaman pertuturan, diarisasi dan analisis sentimen dalam penyelidikan pusat panggilan, termasuk had dan skop setiap kajian yang dipetik.

Laluan panduan yang berkaitanAI Dalam Analitis Pertuturan Pusat Panggilan
soalan 1 daripada 8

What output can automatic speech recognition provide for speech analytics?

Pratonton setiap soalan8 soalan

Jawapan dan penjelasan muncul semasa anda mengambil kuiz.

  1. What output can automatic speech recognition provide for speech analytics?
    • A machine-generated transcript for downstream text analysis
    • A guaranteed speaker identity
    • A final diagnosis of the caller’s emotional state
    • A payment authorization
  2. What does speaker diarization do in a call recording?
    • Menerjemahkan panggilan ke bahasa lain
    • Segments the recording into speaker turns and assigns speaker labels
    • Mengukur betapa kuatnya panggilan itu
    • Menyulitkan rakaman
  3. In IBM’s two-stream call-center emotion-recognition study, which information streams are combined?
    • Video and keystroke timing
    • Customer purchase history and account balance
    • Acoustic and lexical information
    • Only transcript punctuation
  4. Ciri akustik manakah yang membantu mengesan keadaan emosi pelanggan melebihi perkataan yang digunakan?
    • Saiz fail rakaman
    • ID pekerja ejen
    • Pembawa telefon
    • Pitch, rentak dan kelantangan (prosodi)
  5. In a two-stream call-center emotion system, what do acoustic features contribute?
    • Evidence from vocal delivery, such as how speech sounds
    • The caller’s verified identity
    • The account’s payment history
    • A guaranteed diagnosis of the caller’s emotion
  6. What task did Vidrascu and Devillers study for customer-service phone calls?
    • Predicting credit scores from call recordings
    • Speaker segmentation and emotion-related analysis of conversational turns
    • Generating live agent sales recommendations
    • Automating legal decisions about customer accounts
  7. Why might a call-center sentiment system combine lexical and acoustic features?
    • Acoustic features replace the transcript in every task
    • Lexical features reveal the phone carrier
    • The words and the way they are spoken provide complementary signals
    • Combining them guarantees the caller’s true emotion is known
  8. What does word-error rate compare in speech recognition evaluation?
    • Speaker identity against a customer account
    • Emotion labels against call duration
    • Call volume against agent schedule
    • Substitutions, deletions and insertions against the words in a reference transcript