Teller Report

Now you can see non-English news...

Data protection: Personalization yes, but please not with my data!


Many users find personalized recommendations on the net acceptable. If they are not based on their data. Political advertising is particularly unpopular, according to a study.

The consumer is sometimes a very contradictory creature. Of course he wants to avoid plastic somehow, but not if he is standing in the supermarket and buying a six-pack of water bottles. And of course he wants good meat, but not if it is two euros more expensive.

This contradiction can also be observed on the Internet. This is shown once again by a representative survey commissioned by the Max Planck Institute for Human Development and the University of Bristol, which is available in advance from ZEIT ONLINE (Kozyreva, Herzog et al., 2020). A team of scientists has 1,065 people in Germany on their attitudes asked about data protection, personalized messages and automatically generated recommendations on the Internet. The results show: Many people do not seem to be aware of what data companies collect about them and what they use them for.

For example, most of the respondents indicated that they considered it acceptable to be automatically recommended to restaurants or shops, events, films or music based on their personal data and behavior. To a large extent, they did not think it was okay based on information such as relationship status, sexual orientation or household income. Most of the respondents also did not want their own behavior on the Internet to be documented: They refused to store or continue to use information such as the purchase history, viewed videos, location or content of e-mails as well as online messages.

Just not for political purposes!

"Because of the non-transparent design of the services, people obviously do not make a connection between the use of services and the use of data," says scientist Anastasia Kozyreva, who does research at the Max Planck Institute for Educational Research. Many seem to be unclear which personal data companies use to show suitable recommendations. That the location is queried, for example, so that Google can display suitable restaurants nearby. That the purchase history could influence which films Amazon recommends. Or that Facebook could refer to events based on the videos you watched.

The surveyed users clearly reject certain personalizations. According to the survey, they do not want their data to be used for political purposes. The majority of respondents were against personalized political advertising, tailor-made social media feeds and personalization in online newspapers. This shows that people realized that independent political opinion formation is important for the functioning of a society, Kozyreva said.

Once again, the study shows that the desire for more data protection does not necessarily lead to a change in behavior among users. According to the survey, the topic causes 82 percent of the respondents concerns. Nevertheless, one in five stated that they had not changed any settings on websites or social networks in the past year, nor had they used digital tools to protect data - such as an incognito mode in the browser or data-efficient search engines such as DuckDuckGo.

In science, this behavior is also referred to as the data protection paradox: people always state in surveys that they do not want their data to be used for business purposes or recommendations. However, the benefits of data-intensive use outweigh the risks in everyday life (e.g. Telematics and Informatics: Barth, Jong, 2017 ).

In addition: "Even if it is really important to someone to protect their own data, it is incredibly difficult," says study author Kozyreva. It is not transparent which data the companies evaluate for which purposes. It therefore wants regulatory measures from the Federal Government and the European Union: Users should be able to understand more easily what data is collected and what it is used for. "We cannot leave that to the companies," Kozyreva said.

The study authors also asked what terms users know. A large majority stated that they more or less knew what words like artificial intelligence or personalized advertising meant (yes, it is actually so vaguely worded in the survey). The terms computer algorithms and machine learning appeared to be less well known. What unfortunately does not emerge from the survey: what people actually understand by these words.

Source: zeit

You may like

Life/Entertain 2020-02-13T14:43:45.248Z
News/Politics 2019-09-03T22:47:01.477Z
News/Politics 2020-02-16T14:39:07.121Z
Tech/Game 2019-09-16T11:29:09.375Z

Trends 24h


tech 2020/03/28    

© Communities 2019 - Privacy