Although the number of our data is limited, the result
is encouraging. It explicitly indicates that people’s mental status correlates with their browsing behavior, so we can
identify people’s mental health in time if we can track
people’s browsing behavior. Taking depression as an example, the prediction precision is 88.90%, which is fairly
well. In fact, the result also demonstrates web behavior
can be used to do psychological prediction to some extent.
Furthermore, if we can predict all nine dimensions of SCL-
90, we are able to know people’s mental health in every
aspect, which is really useful to help people.