Author: YuMaNzI
Date: April 2015
Paragraphs or sentences below are some writing templates I find useful during each month.
--------------------------------------------------------------------------
Obviously, these assumptions do not often hold true, and document clustering results could be terribly wrong with broken assumptions.
These ways may be helpful in some specific domains, but not general since favorable external dataset might not be always available.
They (short texts) do not provide sufficient word co-occurrence or context shared information for effective similarity measure [23], which is the basis of clustering methods. Therefore, the conventional texts clustering methods may fail to achieve satisfactory results when they are directly applied to short text tasks.