1: \begin{abstract}

2: How to achieve the tradeoff between privacy and utility is one of fundamental problems

3: in private data analysis.

4: In this paper, we give a rigourous differential privacy analysis of networks in the appearance of covariates via

5: a generalized $\beta$-model, which has an $n$-dimensional degree parameter $\beta$

6: and a $p$-dimensional homophily parameter $\gamma$.

7: Under $(k_n, \epsilon_n)$-edge differential privacy, we use the popular Laplace mechanism to release the network statistics.

8: The method of moments is used to estimate the unknown model parameters.

9: We establish the conditions guaranteeing consistency of the differentially private estimators $\widehat{\beta}$ and $\widehat{\gamma}$

10: as the number of nodes $n$ goes to infinity, which reveal an

11: interesting tradeoff between a privacy parameter and model parameters.

12: The consistency is shown by applying a two-stage Newton's method to obtain the upper bound of the error between $(\widehat{\beta},\widehat{\gamma})$ and its true value $(\beta, \gamma)$

13: in terms of the $\ell_\infty$ distance,

14: which has a convergence rate of rough order $1/n^{1/2}$ for $\widehat{\beta}$ and $1/n$ for $\widehat{\gamma}$, respectively.

15: Further, we derive the asymptotic normalities of $\widehat{\beta}$ and $\widehat{\gamma}$, whose asymptotic variances are the same as those of

16: the non-private estimators under some conditions.

17: Our paper sheds light on

18: how to explore asymptotic theory under differential privacy in a principled manner;

19: these principled methods should be applicable to a class

20: of network models with covariates beyond the generalized $\beta$-model.

21: Numerical studies and a real data analysis demonstrate our theoretical findings.

22: 

23: 

24: \end{abstract}

25: