Exploiting Multiple Sources of Information in Learning an Artificial Language: Human Data and Modeling