Identification of Access Control Policy Sentences from Natural Language Policy Documents

摘要: Access control mechanisms are a necessary and crucial design element to any application’s security. There plethora of accepted access models in the information security realm. However, attribute-based (ABAC) has been proposed as general model that could overcome limitations dominant (i.e., role-based control) while unifying their advantages. One issue with migrating an ABAC is needs be encoded typically buried within existing natural language artifacts, hence difficult interpret. This requires processing documents extracting policies from those documents. Software requirements policy main sources declaring organizational policies, but they often huge consist lot descriptive sentences lack content. Manually these extract then using them build laborious expensive process. paper first step towards new engineering approach for by identifying contents. We take advantage multiple techniques including pointwise mutual identify evaluate our on different domains conference management, education, healthcare. Our methodology effectively identifies average recall precision 90% all datasets, which bested state-of-the-art 5%.