History

At present, some instrument companies such as Horiba have developed their own software to slightly process the determined EEM data in Aqualog. This software is developed based on commercial software Origin and have pretreatment functions of blank subtraction, inner filter effect correction, and elimination of Raman and Rayleigh scatters, which simplify the data processing of EEMs. The commercial software Solo developed by Eigenvector Research Inc. (EVRI) can realize PARAFAC analysis of EEM. The software above has a good interface, but users need to purchase them and some academic user might not afford those fee. Wade Sheldon at University of Georgia, USA, developed a software package, namely FLToolbox, using MATLAB. It can eliminate Raman and Rayleigh scatters and show EEM data in the way of contour plot and 3-D shaded surface plot (Zepp et al., 2004). Murphy et al (2010) at the University of New South Wales designed and developed a MATLAB code, namely FDOMcorrect.m, to propose a standard protocol of pretreatment of EEM data. It includes instrument spectra correction, background (blank) subtraction, inner filter (self-absorption) effect correction, Raman normalization, and quinoline sulfate calibration (Murphy et al., 2010). Those procedures finally would make the EEM dataset from various instrument accurate and comparable. In regard to the development of PARAFAC packages, Stedmon and Bro at the University of Copenhagen, Denmark firstly launched DOMFlour package based on N-way toolbox of MATLAB. And that software become the mainstream way for PARAFAC analysis of fluorescent organic matter (Stedmon and Bro, 2008). They developed another software, namely, drEEM, including functions of EEM data import, EEM's display, and EEM correction, in collaboration with Murphy (Murphy et al., 2013). The drEEM consider more user experience. Fellman et al. (2009) wrote Matlab code to use the known PARAFAC model to fit the new EEM dataset. In order to quantify the similarity between the user's PARAFAC model and the PARAFAC models in the literatures, Parr et al. (2014) wrote comPARAFAC in R language. And He et al. (2015) also wrote a MATLAB code comPARAFAC, both two codes can provides a tool for the user to compare two PARAFAC models.

It is important to note that there are still some problems for the present non-commercialized software and code (Figure 2): (1) the user-friendly software such as FLtoolbox does not have function of PARAFAC analysis, (2) the software with powerful functions such as drEEM improved from DOMflour does not have friendly interface and it takes people who are not familiar with MATLAB language a certain amount of time to learn; (3) some functions can only be realized by code such as comPARAFAC and PARAFAC model fitting, which is unfavorable for method promotion. Sometimes, the code fails in working and the results can't be exported. The users, who are not familiar with MATLAB language, can't identify codes' incompatible and wrong expressions.

According to the above problems, our software absorbs the advantages the software or code above, makes up the lack of theory, and improve the user experience. The software is a dedicated to the pretreatment and PARAFAC analysis of EEM dataset. The beneficiaries covers researchers in both industry and academia.

Figure 2 Non-commercialized software and coder for data handling of EEM dataset