Protecting privacy of individuals is critical for forensic genetics. In a kinship/identity testing, related DNA profiles between user's query and the DNA database need to be extracted. However, unrelated profiles cannot be revealed to each other. The challenge is today's DNA database usually contains millions of DNA profiles, which is too big to perform privacy-preserving query with current cryptosystem directly. In this thesis, we propose a scalable system to support privacy-preserving query in DNA Database. A two-phase strategy is designed: the first is a Short Tandem Repeat index tree for quick fetching candidate profiles from disk. It groups loci of DNA profiles by matching probability, so as to reduce I/O cost required to find a particular profile. The second is an Elliptic Curve Cryptosystem based privacy-preserving matching engine, which performs match between candidates and user's sample. In particular, a privacy-preserving DNA profile matching algorithm is designed, which achieves O(n) computing time and communication cost. Experimental results show that our system performs well at query latency, query hit rate, and communication cost. For a database of one billion profiles, it takes 80 seconds to return results to the user.

<p>Forensic DNA databases are expanding in both use and range. In particular, the U.K. and U.S. are developing new techniques and policies in regards to their forensic DNA databases with the hope of increasing the role of forensic DNA databases in criminal investigations. Despite the goal of reducing crime, there are ethical concerns that arise with the ways in which these forensic DNA databases are being developed. This paper outlines the technical aspects of forensic DNA databases and then describes different conceptions of identity, using race as an example of a constructed identity that is relevant in the use of forensic DNA databases. Then it explains how forensic DNA databases construct a unique identity with the goal of ascribing this to people and groups. This ascribed identity is problematic, and different problems that are related to identity are discussed. Despite the benefits of forensic DNA databases, these problems are ethically relevant and as such, a series of policy recommendations are made with the aim of balancing the harms and benefits of forensic DNA databases.</p>

Forensic DNA databases are expanding in both use and range. In particular, the U.K. and U.S. are developing new techniques and policies in regards to their forensic DNA databases with the hope of increasing the role of forensic DNA databases in criminal investigations. Despite the goal of reducing crime, there are ethical concerns that arise with the ways in which these forensic DNA databases are being developed. This paper outlines the technical aspects of forensic DNA databases and then describes different conceptions of identity, using race as an example of a constructed identity that is relevant in the use of forensic DNA databases. Then it explains how forensic DNA databases construct a unique identity with the goal of ascribing this to people and groups. This ascribed identity is problematic, and different problems that are related to identity are discussed. Despite the benefits of forensic DNA databases, these problems are ethically relevant and as such, a series of policy recommendations are made with the aim of balancing the harms and benefits of forensic DNA databases.