According to this valuable clue I found a handy and useful video on YouTube. Now I recommend it for who are looking for a way to calculate a concentration from peak area values easily.
Yes. It is the linearity curve and is expressed as y = mx + b (a straight line!). y is the peak area. x is the concentration expressed as mg/mL , ug/mL or ng/mL. b is the y-intercept. m is the slope of the curve or response factor. All these items are determined by experiment!
To add one additional concept, in many cases it is useful to include an "internal standard" which is a compound that is spiked into the sample at a known (usually fixed) concentration. It is usually spiked into the sample at the beginning of the sample preparation process, and its purpose is to compensate for various sources of variability in the process, such as variable efficiency of extraction of a sample from a matrix, variations in instrument response, and so forth.
The internal standard should be chemically and chromatographically similar to the target compound. This works best in a mass spectrometer is the detection system for the chromatograph because one can then use isotopically labeled versions of your target compound as an internal standard. This allows the internal standard to give nearly (but not quite) perfect tracking of the target compound.
Here is how the internal standard is used in the data analysis. The peak area of the target compound is divided by the peak area of the internal standard. This number is then compared to a calibration curve, usually based on a linear equation, to obtain the concentration.
The calibration curve is prepared by preparing a set of samples containing several different concentrations of target analyte. The calibration samples all contain the same concentration of internal standard. These are then run on the instrument and the peak area ratio is calculated for each calibration sample. The concentration of the target compound is plotted on the x-axis and the measured response ratio is plotted on the y-axis. (In principle, you could swap the axis, but most people use the conventions just noted.)
Then, you run your unknown samples ("unknown" in the sense of unknown concentration of target compound, not in the sense of the chemical identity being unknown). You take the peak area ratio of an individual unknown sample, compare it to the y-axis of the calibration curve, and then drop down vertically from the calibration curve to read the unknown concentration on the x-axis.