Twitter once again to make many people amazed. Research conducted by Adam Sadilek and his team at the University of Rochester in New York, United States, has found a way to predict when someone will get sick.
Only with the passing of tweets that are sent by users of twitter, it can be done. They can predict the 8 days before the risk of contracting the disease to its users, even in areas where the virus is spreading.
Sadilek and his team analyzed 4.4 million tweets a GPS location data with more than 60,000 users for more from a month in the year 2010 in New York with 90% accuracy of the results. This study uses a machine algorithm that can ignore the healthy tweet from 'tweets that are being affected by symptoms of the disease.
But unfortunately, this study is limited because not everyone is tweeting about the disease problems that were suffered. In addition, factors that can make a person stricken with the disease are not described in detail in a tweet.
For those of you who are curious about the project being developed by Adam Sadilek and his team, you can directly access from the site Corpora.io
Only with the passing of tweets that are sent by users of twitter, it can be done. They can predict the 8 days before the risk of contracting the disease to its users, even in areas where the virus is spreading.
Sadilek and his team analyzed 4.4 million tweets a GPS location data with more than 60,000 users for more from a month in the year 2010 in New York with 90% accuracy of the results. This study uses a machine algorithm that can ignore the healthy tweet from 'tweets that are being affected by symptoms of the disease.
But unfortunately, this study is limited because not everyone is tweeting about the disease problems that were suffered. In addition, factors that can make a person stricken with the disease are not described in detail in a tweet.
For those of you who are curious about the project being developed by Adam Sadilek and his team, you can directly access from the site Corpora.io
No comments:
Post a Comment