PREDICTING RELATIVE JOB PLACEMENT POTENTIALS AND MINING SKILL SETS BY ANALYZING ONLINE JOB ADS

Job ads contain quite valuable information. In particular, a large number of job ads in
the web analyzed together offers the opportunity to reach significant statistics. In this study,
online job ads relevant to engineering were analyzed using data mining techniques. In
particular, classification and association analysis techniques were used. We retrieved 17.347
job ads from kariyer.net automatically with the help of a program we developed. We classify
the job ads for five engineering departments: computer engineering, electrical engineering,
industrial engineering, civil engineering, and mechanical engineering. We determined the
total number of personnel advertised for each discipline. Using the total university placement
quotas as the approximate number of graduates in Turkey for these departments, we
calculated the relative job placement potential of each department. We used association
analysis methods to examine the relationships of these departments in ads. In addition, we
investigated most wanted technical skills for computer engineering.
This study contains valuable results for students who plan to select an engineering
major in a university. In addition, it will be beneficial for computer engineering students who
want to specialize on some subfields. Moreover, it will also be helpful for academics for
curriculum development. It is the first study in Turkey by examining this many online job
ads.