Many applications of collaborative filtering (CF), such as news item recommendation and bookmark recommendation, are most naturally thought of as oneclass collaborative filtering (OCCF) problems. In these problems, the training data usually consist simply of binary data reflecting a user's action or inaction, such as page visitation in the case of news item recommendation or webpage bookmarking in the bookmarking scenario. Usually this kind of data are extremely sparse (a small fraction are positive examples), therefore ambiguity arises in the interpretation of the non-positive examples. Negative examples and unlabeled positive examples are mixed together and we are typically unable to distinguish them. For example, we cannot really attribute a user not bookmarking a page to a lack of interest or lack of awareness of the page. Previous research addressing this one-class problem only considered it as a classification task. In this paper, we consider the oneclass problem under the CF setting. We propose two frameworks to tackle OCCF. One is based on weighted low rank approximation; the other is based on negative example sampling. The experimental results show that our approaches significantly outperform the baselines.

No abstract

Most state-of-the-art electronic wearable sensors are powered by batteries that require regular charging and eventual replacement, which would cause environmental issues and complex management problems. Here, a device concept is reported that can break this paradigm in ambient moisture monitoring-a new class of simple sensors themselves can generate moisture-dependent voltage that can be used to determine the ambient humidity level directly. It is demonstrated that a moisture-driven electrical generator, based on the diffusive flow of water in titanium dioxide (TiO ) nanowire networks, can yield an output power density of up to 4 µW cm when exposed to a highly moist environment. This performance is two orders of magnitude better than that reported for carbon-black generators. The output voltage is strongly dependent on humidity of ambient environment. As a big breakthrough, this new type of device is successfully used as self-powered wearable human-breathing monitors and touch pads, which is not achievable by any existing moisture-induced-electricity technology. The availability of high-output self-powered electrical generators will facilitate the design and application of a wide range of new innovative flexible electronic devices.

Ongoing efforts to develop this energy have led to the creation of a variety of novel technologies based on photovoltaic, [12,13] piezoelectric, [14,15] triboelectric, [16,17] and thermoelectric principles. [18,19] Devices based on these technologies are readily incorporated into self-powered, battery-free electronic systems. This also acts to eliminate the harm to the environment caused by the use of conventional batteries. [20,21] As recyclable resource, water is not only indispensable to life but also constitutes the largest energy carrier on Earth (Figure 1a). As 71% of the Earth's surface is covered by water and 35% of solar energy incident on the Earth is absorbed by water, petawatts of energy are received in this way. [22] As the water resource is clean and sustainable, the energy contained in even a small fraction of the water on Earth can meet the current global energy demand if this energy can be efficiently harvested. Terrestrial water exists in vapor and liquid form including moisture, rain, clouds, lakes, rivers, and oceans. It also forms the basis for biological systems. Even though there is a long history of electricity generation from running water, most technologies only utilize the gravitational and kinetic energy in liquid water. [23] Such systems are inherently incompatible with the development of self-powered devices. The development of nanomaterials has enabled technology for the extraction of electrical energy from moisture, leading to The exploration of the utilization of sustainable, green energy represents one way in which it is possible to ameliorate the growing threat of the global environmental issues and the crisis in energy. Moisture, which is ubiquitous on Earth, contains a vast reservoir of low-grade energy in the form of gaseous water molecules and water droplets. It has now been found that a number of functionalized materials can generate electricity directly from their interaction with moisture. This suggests that electrical energy can be harvested from atmospheric moisture and enables the creation of a new range of self-powered devices. Herein, the basic mechanisms of moisture-induced electricity generation are discussed, the recent advances in materials (including carbon nanoparticles, graphene materials, metal oxide nanomaterials, biofibers, and polymers) for harvesting electrical energy from moisture are summarized, and some strategies for improving energy conversion efficiency and output power in these devices are provided. The potential applications of moisture electrical generators in self-powered electronics, healthcare, security, information storage, artificial intelligence, and Internet-of-things are also discussed. Some remaining challenges are also considered, together with a number of suggestions for potential new developments of this emerging technology.

We describe a rapid, simple, room-temperature technique for the production of large-scale metallic thin films with tunable plasmonic properties assembled from size-selected silver nanoplates (SNPs). We outline the properties of a series of ultrathin monolayer metallic films (8-20 nm) self-assembled on glass substrates in which the localized surface plasmon resonance can be tuned over a range from 500 to 800 nm. It is found that the resonance peaks of the films are strongly dependent on the size of the nanoplates and the refractive index of the surrounding dielectric. It is also shown that the bandwidth and the resonance peak of the plasmon resonance spectrum of the metallic films can be engineered by simply controlling aggregation of the SNP. A three-dimensional finite element method was used to investigate the plasmon resonance properties for individual SNPs in different dielectrics and plasmon coupling in SNP aggregates. A 5-17 times enhancement of scattering from these SNP films has been observed experimentally. Our experimental results, together with numerical simulations, indicate that this self-assembly method shows great promise in the production of nanoscale metallic films with enormous electric-field enhancements at visible and near-infrared wavelengths. These may be utilized in biochemical sensing, solar photovoltaic, and optical processing applications.