Which clustering technique would be best for each of the following scenarios? Justify your answers.
a) A company wishes to perform clustering on a set of documents based on term similarities in order to analyze them better. Since one document might fit well into two different clusters, while another might not fit well into any, the company would like each document to belong to zero or more clusters.
b) As a networking activity, a teacher wishes to assign her students to groups based on data she collected on the students’ interests, so that the students in each group will have similar interests to discuss.
c) A new social networking site is trying to figure out the best algorithm to design their friend recommendation feature around, based on similar interests, similar friends, and so on.
Already registered? Login
Not Account? Sign up
Enter your email address to reset your password
Back to Login? Click here