Authors
Department of Computer Science, University of Technology, Baghdad, Iraq.
Abstract
Social media have been increasing obviously and widely due to the fact that it is a media
for users who express their emotions using reviews and comments on a variety of areas in life. In the
present study, a modest model has been suggested for the assessment of service departments with the
use of reviews and comments in social media pages of those departments from various governorates.
The utilization of the text mining for the sentiment classification has been used through collecting
Iraqi dialect reviews on service department pages on Facebook to be analyzed with the use of the
sentiment analysis to track the emotions from the comments and posts. Those have been classified
after that to positive, neutral or negative comment with the use of the algorithms of Naive Bayesian,
Rough Set Theory, and K-Nearest Neighbors. Out of 13 Iraqi capital (Baghdad) service departments
have been tackled, it has been found that 11% of those departments had very good assessment, 18%
from these service departments have good assessment, 21% from these service departments have
medium assessment, 24% from these service departments have acceptance assessment and 26% from
these service departments have bad assessment. The results of the evaluation showed the poor
services provided by service departments in the capital Baghdad. Experimental results were helpful
for the service departments in improving their work and programs had responded quickly and
sufficiently to the customer demands.
Keywords
)