A Statistical Prediction Model of Students’ Success on Job Hunting by Log Data  

Oleh:

Hayakawa, Mao ; Mikawa, Kenta ; Ishida, Takashi ; Goto, Masayuki

Jenis: Article from Proceeding
Dalam koleksi: The 14th Asia Pacific Industrial Engineering and Management Systems Conference (APIEMS), 3-6 December 2013 Cebu, Philippines, page 1-10.
Topik: Data analysis; Weibull distribution; Mixture distribution; Parameter estimation; Predicitive model
Fulltext: 1054.pdf (563.9KB)

Isi artikel

Due to the increase of university students and recent recession, it is not an easy work for a university student to get a job in Japan. Recently, many students are using internet portal sites for job hunting, that help them to easily find suitable job offers. On the other hand, lengthening of job hunting activity became a social issue because it can be an obstacle for students to study. This problem also leads to increase the cost of companies for recruitment activities. Therefore, site managers are looking for a solution to the lengthening problem of job hunting. It is necessary for administrators of internet portal sites for job hunting to previously support students whose job hunting activity seem to be prolonged. If administrators can find students who seem to face on difficulties to get a job at an early stage, they can treat such students previously. Therefore, this study proposes a statistical model to predict when job hunting is expected to be finished for each user. Because the period of the hunting may depend on the students’ personality and the access pattern in a web site for job hunting, we show a statistical model estimated by using user’s attributes and their access log data on a database. However a simple stochastic model cannot approximate the given empirical distribution. Therefore, a tree model with the mixed Weibull distribution is introduced to predict the pattern of finish of job hunting. Through a simulation experiment by using an actual data set,the effectiveness of the proposedpredictive model is clarified.

Opini Anda

Klik untuk menuliskan opini Anda tentang koleksi ini!

Kembali