The transforms are SVG transforms (for details, have a look at the standard; here are some examples). Basically, scale and translate apply the respective transformations to the coordinate system, which should work as expected in most cases. You can apply more than one transform however (e.g. first scale and then translate) and then the result might not be what you expect.
When working with the transforms, keep in mind that they transform the coordinate system. In principle, what you say is true -- if you apply a scale > 1 to an object, it will look bigger and a translate will move it to a different position relative to the other objects.
与恶龙缠斗过久,自身亦成为恶龙;凝视深渊过久,深渊将回以凝视…