Start with the equation y=x+1. Its graph is a line, and every point on it has y exactly 1 more than x. Now change it to y>x+1. The points that work have y more than 1 above x: for x=2, any y above 3. Do that for every x and you get the whole region above the line. The line splits the plane into two halves, and the inequality picks one of them.
The line itself is the boundary. In y>x+1 the points on the line have y equal to x+1, not greater, so they do not count, and the line is drawn dashed. In y≥x+1 they do count, and the line is drawn solid.
To choose the side, test a point that is not on the line. For 2x+3y≤12, draw the solid line through (6,0) and (0,4), then try (0,0): 2(0)+3(0)=0, and 0≤12 is true, so shade the side containing the origin. If the line passes through (0,0), test a different point, such as (1,0).
Graphing the equation y=x+1 gives only the line, with no shading, because only points exactly on it work. A one-variable inequality like x>2 is usually drawn on a number line as a ray; on a coordinate plane it becomes the region to the right of the vertical line x=2.
Think of it like this. The boundary line is a fence across a field, and the inequality tells you which side of the fence is your yard. A solid fence belongs to the yard (≤ or ≥); a dashed one is just a marker you cannot stand on (< or >). Unlike a real yard, the shaded region keeps going forever in its direction.
Core idea. Graphing an inequality means drawing its boundary line (solid if the endpoint counts, dashed if not) and shading the half-plane where the inequality holds.