Um grupo de pesquisadores britânicos está desenvolvendo um algoritmo que é capaz de predizer onde o usuário de um celular estará nas próximas 24 horas. Rodando no smartphone, o GPS do futuro¿ tem um erro de 20 metros, margem bem menor do que a de estudos no mesmo assunto, segundo o site Slate.
A base da predição são os padrões de comportamento do dono do aparelho que, conforme os pesquisadores se consolidam com o passar do tempo, tornando-se mais previsíveis.
Para a pesquisa, foram obtidos dados de smartphones de usuários e também de seus contatos mais frequentes. Uma vez cruzados pelo algoritmo, os dados indicam os padrões de deslocamento de um indivíduo quando está só e quando está acompanhado.
A partir dessa análise, o sistema dirá, por exemplo, quais as chances de um usuário ir ao cinema com amigos nas próximas 24 horas.
Pela inovação, os pesquisadores Mirco Musolesi, Manlio Domenico e Antonio Lima venceram o prêmio Nokia Mobile Data Challenge de 2012.
Comentando a tecnologia, Parmy Olson, da Forbes, vê aplicações polêmicas como o possível uso por autoridades para prever os movimentos de suspeitos de crimes, ou de qualquer outro cidadão.
Já Musolesi, um dos criadores do algoritmo, prefere pensar nas possibilidades comerciais, como anúncios personalizados de acordo com os locais que o usuário mais frequenta.
A group of British researchers are developing an algorithm that is able to predict where the user will be in a cell next 24 hours. Running on the smartphone, the future of GPS has an error of 20 meters, much smaller margin than the studies on the same subject, according to the website Slate.
The basis of prediction are the patterns of behavior of the owner of the apparatus, as the researchers are consolidated with the passage of time, becoming more predictable.
For the study, data were obtained from users of smartphones and also their most frequent contacts. Once crossed by the algorithm, the data indicate patterns shift when an individual is alone and when accompanied.
From this analysis, the system will say, for example, what are the chances of a user going to the movies with friends in the next 24 hours.
Through innovation, researchers Mirco Musolesi, Manlio Domenico and Antonio Lima won the prize Nokia Mobile Data Challenge 2012.
Commenting on the technology, Parmy Olson, Forbes, sees applications controversial as the possible use by authorities to predict the movements of criminal suspects or any other citizen.
Already Musolesi, one of the creators of the algorithm, prefers to think of the commercial possibilities, as announcements according to the sites a user frequents most.
The basis of prediction are the patterns of behavior of the owner of the apparatus, as the researchers are consolidated with the passage of time, becoming more predictable.
For the study, data were obtained from users of smartphones and also their most frequent contacts. Once crossed by the algorithm, the data indicate patterns shift when an individual is alone and when accompanied.
From this analysis, the system will say, for example, what are the chances of a user going to the movies with friends in the next 24 hours.
Through innovation, researchers Mirco Musolesi, Manlio Domenico and Antonio Lima won the prize Nokia Mobile Data Challenge 2012.
Commenting on the technology, Parmy Olson, Forbes, sees applications controversial as the possible use by authorities to predict the movements of criminal suspects or any other citizen.
Already Musolesi, one of the creators of the algorithm, prefers to think of the commercial possibilities, as announcements according to the sites a user frequents most.
Sem comentários:
Enviar um comentário