Also, the sensors in this developed electronic nose system
can detect special volatile compounds for indica rice, such as
ammonia and aromatic compounds which can produce
specific electronic signals by multisensor. In this study,
electronic nose mostly predicted conventional and hybrid
rice correctly (accuracy 75 %), demonstrating that general
indica rice has odor and conventional and hybrid rice hold
specific odor. It has been reported that various techniques
have been developed to evaluate and detect aroma of
aromatic rice (Dong et al. 2001; Pinson 1994), but general
indica rice had not been done. Electronic nose would be
expected to be a rapid technique for odor or aroma rice.
The differences between rice types in electronic signals
cannot be easily detected by our naked eyes or noses, but
they can be differentiated by electronic tongue or nose; in
this study, we proved that the combination of electronic
tongue and nose with the help of LLE-based models
(accuracy 80 or 85 %) can improved the effectiveness of
discrimination for the ordinary rice types.