Term-Based Encrypted Retrieval Privacy

A term-based encrypted retrieval privacy (TERP) data retrieval system performs data retrieval from a data repository server. The system includes a client processor included with a data requesting client and a server processor included with the data repository server. The client processor determines a vector forest that is shared with the data repository server, which includes forest vectors assigned with a respective vector ID, and generates a query including an encrypted ciphertext table that cross-references the vector IDs with a corresponding ciphertext entry. The server processor receives the query, and selects a given document from the data repository server that has assigned thereto at least one nearest neighbor vector among the forest vectors. The server processor compares a nearest neighbor vector ID of the nearest neighbor vector to the vector IDs included in the encrypted ciphertext table, and generates an encoded search result based on the encrypted ciphertext entries. The client processor decrypts the encoded search result to obtain results of the query.