Document-level Relation Extraction with Relation Correlations.

Document-level relation extraction faces two often overlooked challenges: long-tail problem and multi-label problem. Previous work focuses mainly on obtaining better contextual representations for entity pairs, hardly address the above challenges. In this paper, we analyze the co-occurrence correlation of relations, and introduce it into the document-level relation extraction task for the first time. We argue that the correlations can not only transfer knowledge between data-rich relations and data-scarce ones to assist in the training of long-tailed relations, but also reflect semantic distance guiding the classifier to identify semantically close relations for multi-label entity pairs. Specifically, we use relation embedding as a medium, and propose two co-occurrence prediction sub-tasks from both coarse- and fine-grained perspectives to capture relation correlations. Finally, the learned correlation-aware embeddings are used to guide the extraction of relational facts. Substantial experiments on two popular datasets (i.e., DocRED and DWIE) are conducted, and our method achieves superior results compared to baselines. Insightful analysis also demonstrates the potential of relation correlations to address the above challenges. The data and code are released at https://github.com/RidongHan/DocRE-Co-Occur.

Few-Shot Class-Incremental Learning for Medical Time Series Classification.

Continuously analyzing medical time series as new classes emerge is meaningful for health monitoring and medical decision-making. Few-shot class-incremental learning (FSCIL) explores the classification of few-shot new classes without forgetting old classes. However, little of the existing research on FSCIL focuses on medical time series classification, which is more challenging to learn due to its large intra-class variability. In this paper, we propose a framework, the Meta self-Attention Prototype Incrementer (MAPIC) to address these problems. MAPIC contains three main modules: an embedding encoder for feature extraction, a prototype enhancement module for increasing inter-class variation, and a distance-based classifier for reducing intra-class variation. To mitigate catastrophic forgetting, MAPIC adopts a parameter protection strategy in which the parameters of the embedding encoder module are frozen at incremental stages after being trained in the base stage. The prototype enhancement module is proposed to enhance the expressiveness of prototypes by perceiving inter-class relations using a self-attention mechanism. We design a composite loss function containing the sample classification loss, the prototype non-overlapping loss, and the knowledge distillation loss, which work together to reduce intra-class variations and resist catastrophic forgetting. Experimental results on three different time series datasets show that MAPIC significantly outperforms state-of-the-art approaches by 27.99%, 18.4%, and 3.95%, respectively.

Design, Fabrication and Measurement of Full-Color Reflective Electrowetting Displays.

We designed, fabricated and measured full-color, reflective electrowetting displays (EWDs). The display system is composed of three-layer cyan, magenta and yellow EWD elements fabricated with standard photolithographic techniques. The EWDs were driven successfully by the proposed control system and the measurement results show that the electro-optical performance was improved. The aperture ratio of the EWD element can be tuned from 0 to ∼80% as the applied voltage is changed from 0 to 30 V. The response time and the color gamut were measured to be ∼18 ms and ∼58% NTSC, respectively. This paper makes it possible for large numbers of reflective full-color EWDs to be fabricated directly, with advantages of saving power significantly by 85% and no eye irritation compared with LED displays.

Identification of volatile organic compounds in muscle tissues of different species based on Headspace-Gas-Chromatography Ion-Mobility spectrometry.

Species identification of unknown biological samples is crucial for forensic applications, especially in cases of explosion, disaster accidents, and body mutilation after murdering, as well as poaching, illegal trade in endangered animals, and meat food fraud. In this study, we identified 60 volatile organic compounds (VOCs) in fresh skeletal muscle tissues of seven different animal species (cattle, sheep, pigs, rabbits, rats, chickens and carp) and a human dead body by headspace-gas-chromatography ion-mobility spectrometry (HS-GC-IMS), and compared their differences by retention time, drift time and molecular weight. The results showed that these VOCs formed different gallery plot fingerprints in the skeletal muscle tissues of the human dead body and seven animal species. Principal component analysis (PCA) showed significantly different fingerprints between these species, and these fingerprints maintained good stability between the species and within the same species. Some VOCs have high species specificity, while VOCs of human fresh muscle tissues from different individual sources have little difference, demonstrating that all tested muscle tissue samples could be distinguished based on different VOCs. HS-GC-IMS has proved to be a rapid, high-throughput, highly sensitive and specific species identification method, which can be used for forensic species identification in criminal cases and disaster accidents, as well as detection in the field of food safety, such as meat fraud and adulteration.

Leveraging Long and Short-Term Information in Content-Aware Movie Recommendation via Adversarial Training.

Movie recommendation systems provide users with ranked lists of movies based on individual's preferences and constraints. Two types of models are commonly used to generate ranking results: 1) long-term models and 2) session-based models. The long-term-based models represent the interactions between users and movies that are supposed to change slowly across time, while the session-based models encode the information of users' interests and changing dynamics of movies' attributes in short terms. In this paper, we propose the LSIC model, leveraging long and short-term information for content-aware movie recommendation using adversarial training. In the adversarial process, we train a generator as an agent of reinforcement learning which recommends the next movie to a user sequentially. We also train a discriminator which attempts to distinguish the generated list of movies from the real records. The poster information of movies is integrated to further improve the performance of movie recommendation, which is specifically essential when few ratings are available. The experiments demonstrate that the proposed model has robust superiority over competitors and achieves the state-of-the-art results.