This paper considers the problem of publishing "transaction data" for research purposes. Each transaction is an arbitrary set of items chosen from a large universe. Detailed transaction data provides an electronic image of one's life. This has two implications. One, transaction data are excellent candidates for data mining research. Two, use of transaction data would raise serious concerns over individual privacy. Therefore, before transaction data is released for data mining, it must be made anonymous so that data subjects cannot be re-identified. The challenge is that transaction data has no structure and can be extremely high dimensional. Traditional anonymization methods lose too much information on such data. To date, there has been no satisfactory privacy notion and solution proposed for anonymizing transaction data. This paper proposes one way to address this issue.

Two-dimensional (2D) lead halide perovskite with a natural “multiple quantum well” (MQW) structure has shown great potential for optoelectronic applications. Continuing advancement requires a fundamental understanding of the charge and energy flow in these 2D heterolayers, particularly at the layer edges. Here, we report the distinct conducting feature at the layer edges between the insulating bulk terrace regions in the (C4H9NH3)2PbI4 2D perovskite single crystal. The edges of the 2D exhibit an extraordinarily large carrier density of ~1021 cm−3. By using various mapping techniques, we found the layer edge electrons are not related to the surface charging effect; rather, they are associated with the local nontrivial energy states of the electronic structure at the edges. This observation of the metal-like conducting feature at the layer edge of the 2D perovskite provides a different dimension for enhancing the performance of the next-generation optoelectronics and developing innovative nanoelectronics.

Polaritons in two-dimensional materials provide extreme light confinement that is difficult to achieve with metal plasmonics. However, such tight confinement inevitably increases optical losses through various damping channels. Here we demonstrate that hyperbolic phonon polaritons in hexagonal boron nitride can overcome this fundamental trade-off. Among two observed polariton modes, featuring a symmetric and antisymmetric charge distribution, the latter exhibits lower optical losses and tighter polariton confinement. Far-field excitation and detection of this high-momenta mode become possible with our resonator design that can boost the coupling efficiency via virtual polariton modes with image charges that we dub 'image polaritons'. Using these image polaritons, we experimentally observe a record-high effective index of up to 132 and quality factors as high as 501. Further, our phenomenological theory suggests an important role of hyperbolic surface scattering in the damping process of hyperbolic phonon polaritons.

Exploration of ultrastable 2D material‐based optical devices toward all‐optical signal processing is attracting rising interest. As a Group‐VA monoelemental 2D material, antimonene is becoming a promising nonlinear optical material owing to its outstanding optoelectronic advantages with long‐term stability. Herein, all‐optical signal processing based on the high optical nonlinearity of antimonene is first demonstrated. Few‐layer antimonene is fabricated and decorated on the microfiber as an optical device. The device can be applied as an all‐optical Kerr switcher with an extinction ratio as high as ≈12 dB and wavelength conversion of modulated high‐speed signals at a frequency up to 18 GHz. The findings indicate that such a few‐layer antimonene‐based photonics device is applicable in nonlinear optics, which can be potentially developed for the applications of next‐generation high‐speed optical communication.

Neural program embeddings have shown much promise recently for a variety of program analysis tasks, including program synthesis, program repair, codecompletion, and fault localization. However, most existing program embeddings are based on syntactic features of programs, such as token sequences or abstract syntax trees. Unlike images and text, a program has well-defined semantics that can be difficult to capture by only considering its syntax (i.e. syntactically similar programs can exhibit vastly different run-time behavior), which makes syntaxbased program embeddings fundamentally limited. We propose a novel semantic program embedding that is learned from program execution traces. Our key insight is that program states expressed as sequential tuples of live variable values not only capture program semantics more precisely, but also offer a more natural fit for Recurrent Neural Networks to model. We evaluate different syntactic and semantic program embeddings on the task of classifying the types of errors that students make in their submissions to an introductory programming class and on the CodeHunt education platform. Our evaluation results show that the semantic program embeddings significantly outperform the syntactic program embeddings based on token sequences and abstract syntax trees. In addition, we augment a search-based program repair system with predictions made from our semantic embedding and demonstrate significantly improved search efficiency.

We consider the problem of publishing sensitive transaction data with privacy preservation.

A number of medically important disease-causing bacteria (collectively called Gram-negative bacteria) are noted for the extra "outer" membrane that surrounds their cell. Proteins resident in this membrane (outer membrane proteins, or OMPs) are of primary research interest for antibiotic and vaccine drug design as they are on the surface of the bacteria and so are the most accessible targets to develop new drugs against. With the development of genome sequencing technology and bioinformatics, biologists can now deduce all the proteins that are likely produced in a given bacteria and have attempted to classify where proteins are located in a bacterial cell. However such protein localization programs are currently least accurate when predicting OMPs, and so there is a current need for the development of a better OMP classifier. Data mining research suggests that the use of frequent patterns has good performance in aiding the development of accurate and efficient classification algorithms. In this paper, we present two methods to identify OMPs based on frequent subsequences and test them on all Gramnegative bacterial proteins whose localizations have been determined by biological experiments. One classifier follows an association rule approach, while the other is based on support vector machines (SVMs). We compare the proposed methods with the state-of-the-art methods in the biological domain. The results demonstrate that our methods are better both in terms of accurately identifying OMPs and providing biological insights that increase our understanding of the structures and functions of these important proteins.

Carrier transport in GaN terahertz (THz) quantum cascade laser (QCL) structures is theoretically investigated using a non-equilibrium Green's function method. Although scattering due to polar optical phonons in GaN is greatly enhanced with respect to GaAs/AlGaAs THz QCLs, the phonon-induced broadening of the laser levels is found to remain much smaller than other sources of broadening arising from impurity and electron-electron scattering. The gain is calculated self-consistently accounting for the correlation effects in level broadening. Three-well based design with resonant-phonon scheme shows a peak gain of 88/cm at 10 K, and 34/cm at 280 K, which remains above the calculated loss of a double metal waveguide. The results suggest that lasing at 6.6 THz, which is beyond the traditional GaAs THz QCLs, is possible up to 280 K.