The unprecedented growth of information on the Internet has brought about the problem of information overload. To alleviate this problem, news recommendation aims to select news articles for users according to their personal interests. In security applications such as intelligence collection and public opinion monitoring, it is of great importance to obtain valuable information quickly from massive news resources. Different from other application settings, users in security-related scenarios tend to browse news with a domain-oriented purpose. In contrast to the existing news recommendation methods which focus on general-purpose solutions, news recommendation in security applications needs domain-oriented solutions to incorporate users' interests in a specific domain. To this end, in this paper, we propose the problem of domain-oriented news recommendation and develop a specific news recommendation model for security applications. Specifically, our proposed Domain-oriented News Recommendation (DNR) model extracts both general and specific preferences of the user, and performs matching between the user and the candidate news from the above two aspects to combine into the final result. We construct three security-related datasets using a large-scale real-world dataset and validate the effectiveness of our method.