Gids-gekoppelde quiz · Middelmatig Niveau
AI in programmatische advertentiequiz
Test uw kennis van realtime biedveilingen, biedschaduwing en gekalibreerde voorspellingen, en hoe detectie van advertentiefraude en cookie-veranderingen de advertentietechnologie vormgeven.
Vraag 1 van 8
Hoe lang duurt een typische realtime biedveiling ongeveer?
Bekijk elke vraag8 vragen
Antwoorden en uitleg verschijnen terwijl je de quiz doet.
- Hoe lang duurt een typische realtime biedveiling ongeveer?
- Ongeveer 10 seconden
- About 100 milliseconds
- Ongeveer 5 minuten
- About 1 second per bidder, one after another
- Welke organisatie onderhoudt het OpenRTB-formaat dat vaak wordt gebruikt voor biedingsverzoeken?
- IAB Tech Lab
- W3C
- ICANN
- Media Rating Council
- Wat doet bodarcering?
- Hides the advertiser's identity from the publisher
- Blocks bids from suspected fraudulent buyers
- Estimates the lowest price likely to still win a first-price auction
- Raises the publisher's price floor after each sale
- Waarom moeten voorspellingen van de klikfrequentie bij een bieder goed gekalibreerd zijn en niet alleen goed gerangschikt?
- Because regulators require calibration certificates
- Because ranking is impossible with sparse features
- Because calibration makes models run faster
- Because the predicted probability is multiplied directly into the bid price, so overestimates cause overbidding
- Wat kondigde Google in 2024 aan over cookies van derden in Chrome?
- It had already fully removed them
- It would replace them with Apple's App Tracking Transparency
- It would not force their removal, reversing its earlier phase-out plan
- It would block all contextual targeting
- What did the Methbot operation, exposed in 2016, do?
- Stole credit card numbers from ad exchanges
- Faked video ad views on counterfeit versions of premium sites
- Manipulated search engine rankings
- Hijacked publisher ads.txt files to redirect payments
- What is the purpose of ads.txt and sellers.json?
- To store user cookies securely
- To list the ad sizes a site supports
- To speed up bid responses
- To let buyers verify who is authorized to sell a site's inventory
- Why is training a bid-shading model described as a censored-data problem?
- Because the clearing price is only observed when you win
- Because regulators censor auction data
- Because publishers delete losing bids
- Because user identities are removed from logs