Practice Test 3 | Google Cloud Certified Professional Data Engineer | Dumps | Mock Test

Question 12

A system with over 4,000 cameras was installed to collect images of main roads in the city in order to detect traffic density. A model built using TensorFlow processes the images coming from the cameras and based on the number of vehicles covering the road it returns the traffic size as a number between 0 (no traffic) and 1 (heavy traffic).

This system is used by map navigation apps to provide traffic details for navigators and recommend the best route based on their current location and destination. For better data quality usage, it was advised to aggregate traffic density by the time frame of 60 seconds with road number as the dimension. This will provide the average traffic density for every minute on each road. If the Dataflow pipeline is being developed to aggregate traffic stats based on this scenario, which time window should be used?

A. Fixed-time window
B. Sliding-time window
C. Per-session window
D. Single global window

Answer

Correct Answer: B

Here we have 4000 cameras installed at different locations in the city. This means, that due to network latency, there may be a lag in receiving the data from the time scheduled to transfer the data. Some Cameras will be able to send the data on time and some require a longer time. This means we need to keep open windows for a longer time.

A sliding time window uses time intervals in the data stream to define bundles of data. However, with sliding time windowing, the windows overlap. Each window might capture five minutes worth of data, but a new window starts every ten seconds. The frequency with which sliding windows begin is called the period. Therefore, our example would have a window size of five minutes and a period of ten seconds.

Sliding-time window is the windowing function recommended for this scenario.