In this statement the coefficient of X2 means the difference in the intercept for dummy variable's choices. In fact, if you want to investigate the influence of the dummy on the whole process which is under consideration, it'll be much better to use another model (so-called Gujarati approach), i.e. log Y = a + b*x2 + c*x3 + (d*x3)*x2, and you can recognize if your dummy means something not for intercept only, but for the slope too.

Vladislav Shchekoldin: Thank you, sir.

when log(y) is the dependent variable in a model, the coefficient on a dummy variable. Then coefficient will multiplied by 100 and interpreted as the percentage difference in y, holding all other factors fixed.(Jeffrey M. Wooldridge)

Abdullah: Thank you.

As my knowledge that the dummy variables used to help the researcher to interpret the influence of variables ,and there is some variables we can not measure directly in the model ,also I can advise that to use the model of panel data to compare between fixed effect and random effect GLS.

It's difficult for a dummy variable - whose values (e.g. 0/1) are typically assigned by the investigator, to be endogenous - because endogenous variables have a random component.