Kvíz spojený s průvodcem · Střední úroveň

Kvíz AI v Call Center Speech Analytics

Otestujte rozpoznávání řeči, diarizaci a analýzu sentimentu ve výzkumu call centra, včetně limitů a rozsahu každé citované studie.

Související vodicí cestyAI v Call Center Analýza řeči
Otázka 1 z 8

What output can automatic speech recognition provide for speech analytics?

Náhled každé otázky8 otázky

Odpovědi a vysvětlení se zobrazují, jak budete v kvízu.

  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?
    • Přeloží hovor do jiného jazyka
    • Segments the recording into speaker turns and assigns speaker labels
    • Měří, jak hlasitý byl hovor
    • Zašifruje záznam
  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. Která akustická funkce pomáhá odhalit emocionální stav zákazníka nad rámec použitých slov?
    • Velikost souboru nahrávky
    • ID zaměstnance agenta
    • Telefonní operátor
    • Výška, tempo a hlasitost (prozódie)
  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