Reverse-Safe Text Indexing

We introduce the notion of reverse-safe data structures. These are structures that prevent reconstruction they encode (i.e., cannot be easily reversed). A structure D is called z - when there exist at least datasets with same set answers as ones stored by . The main challenge to ensure stores many useful queries possible, constructed efficiently, and has size close original dataset it encodes. Given a text length n an integer , we propose algorithm constructs -reverse-safe ( -RSDS) O(n) decision counting pattern matching most d optimally, where maximal for any such -RSDS. construction takes O(nɷ log d) time, ɷ matrix multiplication exponent. show that, despite nɷ factor, our engineered implementation only few minutes finish million-letter texts. also plugging method in analysis applications gives insignificant or no utility loss. Furthermore, how technique can extended support under realistic adversary models. Finally, -RSDS queries, whose sublinear preliminary version this article appeared ALENEX 2020.