Detecting book return patterns from library circulation data

In this paper, we use spectral methods to detect periodicity in the return pattern of books in the library. To analyze library circulation data, fourier transformation has been applied. From the time series of the two consecutive academic years of the loan data, we explore the daily number of loans from library which process using Fast Fourier Transform (FFT) as well as Discrete Fourier Transform (DFT) to generate frequency information about the time series of circulation data. We could detect a weekly periodicity as well semestral return pattern over a time span of two academic years excluding the holidays. The software package MATLAB is used to demonstrate our findings. It is proved that in the field of information science spectral methods show a lot of promises for analyzing all kinds of time series and other signals occurring. Surely this approach can raise more attention to researchers who want to analyze the periodicity of a continuous event from the data which records historical information.