The presentation discusses various techniques for encrypting data in databases, including deterministic encryption, searchable encryption, and homomorphic encryption.
- Deterministic encryption allows for searches on equality while keeping data encrypted
- Searchable encryption allows for searching for keywords in encrypted documents by encrypting the keywords and storing them in a database
- Homomorphic encryption allows for performing operations on encrypted data in a way that is equivalent to performing the operations before encryption
- Each technique has its limitations and trade-offs
- Maintaining an index of keyword frequency can improve the security of searchable encryption
The presenter explains that searchable encryption can be used to search for keywords in encrypted documents by encrypting the keywords and storing them in a database. However, there is a risk of frequency analysis, where common words can be identified and used to determine the contents of the documents. To address this, an index of keyword frequency can be maintained to prevent repeatable encryption. The presenter notes that there are currently no commercial solutions for this technique, so it must be implemented manually.