Negative y-intercept means it will be added when we use it for the calculation of 'X' of unknown samples which is wrong according to me as it will enhance the result which may be wrong.
Therfore my suggestion is please try to do another xylose standard curve by DNS method by taking small concentrations like (0, 0.25, 0.5, 0.75, 1.0 mg/ml). I hope this 5 readings will give you a good output.
And 2ndly we have to minimize our pipetting error also to get a good standard curve. It is wise to do standard curve 2-3 times to have a good result. Hope it will help you.
The y-intercept should not be negative. This indicates a problem, as Abhas Kumar Maharana explained. The y-intercept should be a positive value corresponding to the background absorbance in the absence of xylose.
If the absorbance values of the standards do not lie along a straight line, but instead lie along a curve, do not try to fit the data with a straight line. Use a linear interpolation, or find a nonlinear analytical function that is a good fit.
Do not use doubling concentrations for the standards. Use evenly-spaced concentrations, as already suggested. The reason for this is that the measured value of the highest concentration in a set of doubling concentrations has a disproportionately larger effect on the linear regression than the other concentrations, so that a small error can throw off the whole curve.
The DNS assay is not recommended for detecting xylanase activity if that is what you are using your xylose std.. for https://www.researchgate.net/publication/281172083_A_Comparison_of_Polysaccharide_Substrates_and_Reducing_Sugar_Methods_for_the_Measurement_of_endo-14-b-Xylanase?ev=prf_pub
Article A Comparison of Polysaccharide Substrates and Reducing Sugar...