Maswali yanayounganishwa na mwongozo · Kati Kiwango

AI katika Maswali ya Uchanganuzi wa Matamshi ya Kituo cha Simu

Jaribu utambuzi wa usemi, uwekaji kumbukumbu na uchanganuzi wa hisia katika utafiti wa kituo cha simu, ikijumuisha mipaka na upeo wa kila utafiti uliotajwa.

Njia za mwongozo zinazohusianaAI Katika Uchanganuzi wa Hotuba ya Kituo cha Simu
Swali 1 ya 8

What output can automatic speech recognition provide for speech analytics?

Hakiki kila swali8 maswali

Majibu na maelezo huonekana unapojibu maswali.

  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?
    • Inatafsiri simu kwa lugha nyingine
    • Segments the recording into speaker turns and assigns speaker labels
    • Hupima jinsi simu ilivyokuwa kubwa
    • Husimba rekodi kwa njia fiche
  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. Ni kipengele gani cha acoustic kinachosaidia kutambua hali ya kihisia ya mteja zaidi ya maneno yaliyotumiwa?
    • Saizi ya faili ya rekodi
    • Kitambulisho cha mfanyakazi wa wakala
    • Mtoa huduma wa simu
    • Lami, kasi, na sauti (prosody)
  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