Let $R$ be a commutative ring. Let $S$ be a collection of pairs $(a,b)$ of elements of $R$. Define addition and multiplication as follows - $$ (a,b)+(a',b') = (a+a',b+b') $$ and $$(a, b) · (a', b') = (a · a' +b · b' · r, a · b' +a' · b) $$ where $r \in R$
Likewise, now I have to define a multiplication rule for 4-tuple $$(0, 1, 0, 0)^2 = r ; (0, 0, 1, 0)^2 = s ; (0, 1, 0, 0)·(0, 0, 1, 0) = (0, 0, 0, 1) = −(0, 0, 1, 0)·(0, 1, 0, 0)$$ for some $r,s \in R$. (possibly different $r,s$ than for 2-tuple case)
I don't understand how this is defined. How did we got $(0, 1, 0, 0)·(0, 0, 1, 0) = (0, 0, 0, 1)$?
As far as I can see, you mean to talk about the generalized quaternions as seen in this post. You don't deduce any of those relations: you need them all to be given so that you can multiply.
To find the general form of the product of two $4$-tuples, you just enforce linearity of multiplication on those generators with those relations.
So, $$(\alpha_1, \alpha_2,\alpha_3,\alpha_4)(\beta_1, \beta_2,\beta_3,\beta_4) \\ =(\alpha_1(1,0,0,0)+\alpha_2(0,1,0,0)+\alpha_3(0,0,1,0)+\alpha_4(0,0,0,1))(\beta_1(1,0,0,0)+\beta_2(0,1,0,0)+\beta_3(0,0,1,0)+\beta_4(0,0,0,1)) \\ =\alpha_1\beta_1 (1,0,0,0) +\alpha_1\beta_2(0,1,0,0)+\alpha_1\beta_3(0,0,1,0)+\alpha_1\beta_4(0,0,0,1) \\ + \alpha_2\beta_1(0,1,0,0)+r\alpha_2\beta_2(1,0,0,0)+\alpha_2\beta_3(0,0,0,1)+r\alpha_2\beta_4(0,0,1,0) \\ +\alpha_3\beta_1 (0,0,1,0)-\alpha_3\beta_2(0,0,0,1)+s\alpha_3\beta_3(1,0,0,0)-s\alpha_3\beta_4(0,1,0,0) \\ + \alpha_4\beta_1 (0,0,0,1) -r\alpha_4\beta_2(0,0,1,0)+s\alpha_4\beta_3(0,1,0,0)-rs\alpha_4\beta_4(1,0,0,0) \\ =(\alpha_1\beta_1+r\alpha_2\beta_2+s\alpha_3\beta_3-rs\alpha_4\beta_4)(1, 0, 0, 0) \\ +(\alpha_1\beta_2+\alpha_2\beta_1-s\alpha_3\beta_4+s\alpha_4\beta_3)(0, 1, 0, 0) \\ +(\alpha_1\beta_3+r\alpha_2\beta_4+\alpha_3\beta_1-r\alpha_4\beta_2)(0, 0, 1, 0) \\ +(\alpha_1\beta_4+\alpha_2\beta_3-\alpha_3\beta_2+\alpha_4\beta_1)(0, 0, 0, 1) $$
So there in the last step, you have the generali mutliplication rule. I proofread as best I could a couple times... hopefully if any typos are here someone will help. (At least in the normal case when $r=s=-1$ the equations hold correctly, I think.)