*14:47* [Job][New]
PhD student, *SnT, University of Luxembourg, Luxembourg*
The Interdisciplinary Centre for Security, Reliability and Trust (SnT, www.securityandtrust.lu) at University of Luxembourg is looking for a PhD candidate in privacy-preserving recommender systems. The PhD topic is related to investigate the privacy issues in recommender systems and propose efficient and secure mechanisms to achieve the maximum privacy. The candidate is expected to design new privacy-preserving recommender systems for both horizontally and vertically partitioned dataset, prove their privacy properties, and study their asymptotical performances. The student will work closely with the team members of the APSIA group, led by Prof. Peter Y. A. Ryan. Moreover, the student will be encouraged to collaborate with researchers from the group of Prof. Jiuyong Li at University of South Australia (UniSA), Australia.

For informal inquiries please contact: Dr. Qiang Tang *qiang.tang (at) uni.lu*

To formally apply for this position: http://emea3.mrted.ly/9lwj

*07:43* [Job][New]
Professor (Open Rank), *Worcester Polytechnic Institute, MA, USA, below Canada*
Worcester Polytechnic Institute (WPI) invites applications for a faculty position in the Department of Electrical & Computer Engineering at all ranks, commensurate with qualifications.Required qualifications for the position include; an earned Ph.D. in Electrical & Computer Engineering, or a closely related field. Areas of particular interest include, but are not limited to: security engineering, hardware and embedded systems security, and mobile and cyber-physical systems security.

The successful candidate will be expected to establish and maintain a high quality, self-sustaining research program. WPI offers ample opportunity for collaboration with current department faculty as well as appropriate cross-campus, interdisciplinary research groups in various topics in security. In addition to excellence in teaching and research, candidates should look forward to engaging undergraduate and graduate students in a classroom and projects intensive environment, and expanding our graduate research program.

Qualified applicants should submit a detailed curriculum vitae, a brief statement of specific teaching and research objectives, and four letters of recommendation at least one of which addresses teaching experience or potential, via https://careers.wpi.edu/. Review of applications will begin on November 1, 2013 and will continue until the position is filled.

*04:17* [Pub][ePrint]
Outsourced Symmetric Private Information Retrieval, by Stanislaw Jarecki and Charanjit Jutla and Hugo Krawczyk and Marcel Rosu and Michael Steiner
In the setting of searchable symmetric encryption (SSE), a data owner D outsources a database (or document/file collection) to a remote server E in encrypted form such that D can later search the collection at E while hiding information about the database and queries from E. Leakage to E is to be confined to well-defined forms of data-access and query patterns while preventing disclosure of explicit data and query plaintext values. Recently, Cash et al presented a protocol, OXT, which can run arbitrary Boolean queries in the SSE setting and which is remarkably efficient even for very large databases.In this paper we investigate a richer setting in which the data owner

D outsources its data to a server E but D is now interested to allow clients (third parties) to search the database such that clients learn the information D authorizes them to learn but nothing else while E still does not learn about the data or queried values as in the basic SSE setting. Furthermore, motivated by a wide range of applications, we extend this model and requirements to a setting where, similarly to private information retrieval, the client\'s queried values need to be hidden also from the data owner D even though the latter still needs to authorize the query. Finally, we consider the scenario in which authorization can be enforced by the data owner D without D learning the policy, a setting that arises in court-issued search warrants.

We extend the OXT protocol of Cash et al to support arbitrary Boolean queries in all of the above models while withstanding adversarial

non-colluding servers (D and E) and arbitrarily malicious clients,

and while preserving the remarkable performance of the protocol.

*22:17* [Pub][ePrint]
Constructing Confidential Channels from Authenticated Channels---Public-Key Encryption Revisited, by Sandro Coretti and Ueli Maurer and Björn Tackmann
The security of public-key encryption (PKE), a widely-used cryptographic primitive, has received much attention in the cryptographic literature. Many security notions for PKE have been proposed, including several versions of CPA-security, CCA-security, and non-malleability. These security notions are usually defined in terms of a certain game that an efficient adversary cannot win with non-negligible probability or advantage.If a PKE scheme is used in a larger protocol, then the security of this protocol is proved by showing a reduction of breaking a certain security property of the PKE scheme to breaking the security of the protocol. A major problem is that each protocol requires in principle its own tailor-made security reduction. Moreover, which security notion of the PKE should be used in a given context is a priori not evident; the employed games model the use of the scheme abstractly through oracle access to its algorithms, and the sufficiency for specific applications is neither explicitly stated nor proven.

In this paper we propose a new approach to investigating the application of PKE, following the constructive cryptography paradigm of Maurer and Renner (ICS~2011). The basic use of PKE is to enable confidential communication from a sender A to a receiver B, assuming A is in possession of B\'s public key. One can distinguish two relevant cases: The (non-confidential) communication channel from A to B can be authenticated (e.g., because messages are signed) or non-authenticated. The application of PKE is shown to provide the construction of a secure channel from A to B from two (assumed) authenticated channels, one in each direction, or, alternatively, if the channel from A to B is completely insecure, the construction of a confidential channel without authenticity. Composition then means that the assumed channels can either be physically realized or can themselves be constructed cryptographically, and also that the resulting channels can directly be used in any applications that require such a channel. The composition theorem shows that several construction steps can be composed, which guarantees the soundness of this approach and eliminates the need for separate reduction proofs.

We also revisit several popular game-based security notions (and variants thereof) and give them a constructive semantics by demonstrating which type of construction is achieved by a PKE scheme satisfying which notion. In particular, the necessary and sufficient security notions for the above two constructions to work are CPA-security and a variant of CCA-security, respectively.

*19:17* [Pub][ePrint]
Limits of Extractability Assumptions with Distributional Auxiliary Input, by Elette Boyle and Rafael Pass
Extractability, or \"knowledge,\" assumptions (such as the \"knowledge-of-exponent\" assumption) have recently gained popularity in the cryptographic community--leading to the study of primitives such as extractable one-way functions, extractable hash functions, succinct non- interactive arguments of knowledge (SNARKs), and extractable obfuscation, and spurring the development of a wide spectrum of new applications relying on these primitives. For most of these applications, it is required that the extractability assumption holds even in the presence of attackers receiving some auxiliary information that is sampled from some fixed efficiently computable distribution Z.We show that, assuming the existence of collision-resistant hash functions, there exists a pair of efficient distributions Z, Z\'; such that either:

o extractable one-way functions w.r.t. Z do not exist, or

o extractability obfuscations for Turing machines w.r.t. Z do not exist.

A corollary of this result shows that assuming existence of fully homomorphic encryption with decryption in NC1, there exist efficient distributions Z, Z\' such that either

o extractability obfuscations for NC1 w.r.t. Z do not exist, or

o SNARKs for NP w.r.t. Z\' do not exist.

To achieve our results, we develop a \"succinct punctured program\" technique, mirroring the powerful \"punctured program\" technique of Sahai and Waters (ePrint\'13), and present several other applications of this new technique.