Test legat de ghid · Mediu Nivel

AI în chestionarul de analiză vocală a centrului de apeluri

Testați recunoașterea vorbirii, diarizarea și analiza sentimentelor în cercetarea în call-center, inclusiv limitele și domeniul de aplicare al fiecărui studiu citat.

Căi de ghidare aferenteAnaliza vorbirii AI în Call Center
Întrebare 1 de 8

What output can automatic speech recognition provide for speech analytics?

Previzualizează fiecare întrebare8 întrebări

Răspunsurile și explicațiile apar pe măsură ce susțineți testul.

  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?
    • Traduce apelul într-o altă limbă
    • Segments the recording into speaker turns and assigns speaker labels
    • Măsoară cât de tare a fost apelul
    • Criptează înregistrarea
  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. Ce caracteristică acustică ajută la detectarea stării emoționale a unui client dincolo de cuvintele folosite?
    • Dimensiunea fișierului înregistrării
    • ID-ul de angajat al agentului
    • Operatorul de telefonie
    • Ton, ritm și volum (prozodie)
  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