Likes in the social network could predict personal aspects of users such as sexual orientation, religion, age or even race. A study carried out on Facebook reveals that through the ‘likes’ of users you can identify aspects such as sexual orientation, religion, age or even race. Researchers at the University of Cambridge have developed, using algorithms, a way to predict from their tastes in the social network the personality of the person. The research, published in the magazine PNAS, has had the participation of 58,000 volunteers, who have allowed the analysis of their ‘Like’ on Facebook. The likes on this button served to feed algorithms that matched the information provided to personality tests. From these algorithms results were extracted with 88 percent accuracy for determining male sexuality, 95 percent accuracy in determining race, and 85 percent for differentiating between Republicans and Democrats.
Christians and Muslims were correctly classified in 82 percent of cases and the status of relationship and substance abuse was predicted with an accuracy of 65 percent and 73 percent. As for specific tastes, ‘I like’ reveal that less than 5 percent of users click on obvious gay content, such as gay marriage, for example. These algorithms also add large amounts of data in terms of personalized tastes like music or television programs, allowing you to create a lot of personal profiles. The study will be of great help to social media companies interested in making more money through personalized marketing. Despite this, researchers have warned that people’s digital profiles are creating privacy threats.
Analytics metrics provide data about visitors to a web page, location, language, average time of a visit, etc. This information no longer surprises the digital community. But what about revealing your personal information through the “Like” button on Facebook?
Recently, researchers at the University of Cambrigde developed a measurement system that reveals the personal data of each user for every 100 clicks on the successful “Like” button, an extremely popular language in social networks. This research demonstrates that we leave a fingerprint on the web so latently that it is possible to predict clearly and automatically a succession of personal singularities such as ethnicity, sexual orientation, personality traits, political and religious beliefs, Happiness, intelligence, separation of parents, gender and age or even the use of addictive substances.
Researchers at the Cambridge Psychometrics Center, in cooperation with Microsoft Research Cambridge, conducted a “Like” analysis and demographic profiles of more than 58,000 Facebook Facebook volunteers as well as psychometric test results from the MyPersonality application. The program allowed to demonstrate, in 95% of the cases, the race of the people, the sexual identity of 85% of the users and the political ideology of 85 of every 100 users.
Likes clicks analysis is based on tacit attributes rather than on explicit preferences, for example, less than 5% of homosexual users evidenced their approval of gay marriage. This reveals that the precise predictions are invoked by deductions in the aggregation of large amounts of tastes and in popular information, such as musical tastes, television programs, etc.
The researchers highlight the great potential for personalized marketing that can mean a revolution in psychological assessment that, according to this research, could be carried out on an unprecedented scale and without expensive evaluation centers and questionnaires.
However, scientists also warn of user privacy threats and actions as “harmless” as following a musical group and expressing our support or rejection of a publication or a photograph, can reveal in a concise way traits Of the personality that, perhaps, we would not like that they were so visible.
Facebook assiduous user David Stillwell of Cambridge University confesses that he will continue to use the largest social network but also says he “could be more careful about using the privacy settings that Facebook offers.”… Read the rest