1: \begin{abstract}

2: This paper introduces and studies a notion of \emph{algorithmic randomness} for subgroups of rationals.

3: Given a randomly generated additive subgroup $(G,+)$ of rationals, two main questions are addressed: first, what are

4: the model-theoretic and recursion-theoretic properties of $(G,+)$; second, what learnability properties can one extract

5: from $G$ and its subclass of finitely generated subgroups?   

6: For the first question, it is shown that the theory of $(G,+)$ coincides with that of

7: the additive group of integers and is therefore decidable; furthermore, while the word problem for $G$ with respect 

8: to any generating sequence for $G$ is not even semi-decidable,

9: one can build a generating sequence $\beta$ 

10: such that the word problem for $G$ with respect to $\beta$ is co-recursively enumerable

11: (assuming that the set of generators of $G$ is limit-recursive).

12: In regard to the second question, it is proven that 

13: there is a generating sequence $\beta$ for $G$ such that every non-trivial finitely generated subgroup

14: of $G$ is recursively enumerable and the class of all such subgroups of $G$ is behaviourally correctly learnable,

15: that is, every non-trivial finitely generated subgroup can be semantically identified in the limit

16: (again assuming that the set of generators of $G$ is limit-recursive).

17: On the other hand, the class of non-trivial finitely generated subgroups of $G$ cannot be

18: syntactically identified in the limit with respect to any generating sequence for $G$.

19: The present work thus contributes to a recent line of research studying algorithmically random

20: infinite structures and uncovers an interesting connection between the arithmetical complexity of 

21: the set of generators of a randomly generated subgroup of rationals and the learnability of its

22: finitely generated subgroups.     

23: \end{abstract}