Big data collection is constantly in the news in regards to privacy. Nielsen’s chief data and research officer, Mainak Mazumdar, explains in his article, Artificial Intelligence is the Link Between Big Data and Persons-Level Measurement, that uncalibrated data is inherently biased and underrepresents minority audiences.
To help combat data collection issues, Mazumdar explains that artificial intelligence (AI) may be the key to cracking the big data code. Through application of AI, big data collection can broaden measurement capabilities while preserving quality and representativeness.
Collecting accurate data with AI is done through recognizing patterns allowing data to be processed where AI removes faulty data items, identifies the person that the data is being collected on behalf of, and accurately reports demographic characteristics of the persons and households.
For example, Nielsen conducted an analysis in the U.S. earlier this year that compared set-top box data. The set-top box is a device that allows a digital signal to be received, decoded and displayed on a television. The set-top box data collected was calibrated with Nielsen panel data. It found that this data did not represent the population accurately. Using AI, the population using the set-top box was able to more accurately portray the television viewing habits of U.S. citizens.
Big data makes incredible amounts of information available to us, and we need to be aware of biases that occur in the data. AI is a powerful way to address bias in the big data, and we are excited to see what Nielsen will do with this technology.