Recent studies give more attention to the anomaly detection (AD) methods that can leverage a handful of labeled anomalies along with abundant unlabeled data. These existing anomaly-informed AD methods rely on manually predefined score target(s), e.g., prior constant or margin hyperparameter(s), to realize discrimination in anomaly scores between normal and abnormal data. However, such methods would be vulnerable to the existence of anomaly contamination in the unlabeled data, and also lack adaptation to different data scenarios.

Deep learning (DL) techniques have recently found success in anomaly detection (AD) across various fields such as finance, medical services, and cloud computing. However, most of the current research tends to view deep AD algorithms as a whole, without dissecting the contributions of individual design choices like loss functions and network architectures. This view tends to diminish the value of preliminary steps like data preprocessing, as more attention is given to newly designed loss functions, network architectures, and learning paradigms. In this paper, we aim to bridge this gap by asking two key questions: (i) Which design choices in deep AD methods are crucial for detecting anomalies? (ii) How can we automatically select the optimal design choices for a given AD dataset, instead of relying on generic, pre-existing solutions? To address these questions, we introduce ADGym, a platform specifically crafted for comprehensive evaluation and automatic selection of AD design elements in deep methods. Our extensive experiments reveal that relying solely on existing leading methods is not sufficient. In contrast, models developed using ADGym significantly surpass current state-of-the-art techniques.

Recognizing terminology entities across domains from professional texts is an important but challenging task in NLP. Most existing methods focus on recognizing generic entities, but few methods are to recognize the domain-specific entities across domains due to the very large discrepancy of entity representations between the source and target domains. To address this issue, we introduce domain vectors and context vectors to represent domain-specific semantics of entities and domain-irrelevant semantics of the context words, respectively. Based on the two types of vectors, we present a simple yet effective novel cross-domain named entity recognition approach, which aligns entity distributions between domains and separates entity distributions from context distributions for easily identifying entities. Experimental results demonstrate that the proposed approach can obtain significant improvement compared to existing cross-domain NER methods.

Current text classification methods typically encode the text merely into embedding before a naive or complicated classifier, which ignores the suggestive information contained in the label text. As a matter of fact, humans classify documents primarily based on the semantic meaning of the subcategories. We propose a novel model structure via siamese BERT and interactive double attentions named IDEA (Interactive DoublE Attentions) to capture the information exchange of text and label names. Interactive double attentions enable the model to exploit the inter-class and intra-class information from coarse to fine, which involves distinguishing among all labels and matching the semantical subclasses of ground truth labels. Our proposed method outperforms the state-of-the-art methods using label texts significantly with more stable results.

Given a long list of anomaly detection algorithms developed in the last few decades, how do they perform with regard to (i) varying levels of supervision, (ii) different types of anomalies, and (iii) noisy and corrupted data? In this work, we answer these key questions by conducting (to our best knowledge) the most comprehensive anomaly detection benchmark with 30 algorithms on 57 benchmark datasets, named ADBench. Our extensive experiments (98,436 in total) identify meaningful insights into the role of supervision and anomaly types, and unlock future directions for researchers in algorithm selection and design. With ADBench, researchers can easily conduct comprehensive and fair evaluations for newly proposed methods on the datasets (including our contributed ones from natural language and computer vision domains) against the existing baselines. To foster accessibility and reproducibility, we fully open-source ADBench and the corresponding results.

Representing the true label as one-hot vector is the common practice in training text classification models. However, the one-hot representation may not adequately reflect the relation between the instance and labels, as labels are often not completely independent and instances may relate to multiple labels in practice. The inadequate one-hot representations tend to train the model to be over-confident, which may result in arbitrary prediction and model overfitting, especially for confused datasets (datasets with very similar labels) or noisy datasets (datasets with labeling errors). While training models with label smoothing can ease this problem in some degree, it still fails to capture the realistic relation among labels. In this paper, we propose a novel Label Confusion Model (LCM) as an enhancement component to current popular text classification models. LCM can learn label confusion to capture semantic overlap among labels by calculating the similarity between instance and labels during training and generate a better label distribution to replace the original one-hot label vector, thus improving the final classification performance. Extensive experiments on five text classification benchmark datasets reveal the effectiveness of LCM for several widely used deep learning classification models. Further experiments also verify that LCM is especially helpful for confused or noisy datasets and superior to the label smoothing method.

In online social networks, users always expect to share some information for benefits (e.g., personalized services) while hiding the others for privacy. Unfortunately, the hidden information is likely to be predicted by various powerful inference attacks with the rapid advances in machine learning. Then, what is the risk that a user's private information could be disclosed? What countermeasures can be taken to fight against the privacy violation for the user? To tackle these issues, this article proposes a general Framework for Private Attribute Disclosure estimation (F-PAD) including three steps: 1) private attribute prediction; 2) disclosure model training; 3) disclosure risk estimation. Not like most prior risk estimation studies focusing on one specific attack model and private attribute, F-PAD can estimate disclosure risk for individual users in terms of disclosure probability and risk level within a high confidence given a basket of potential inference attack models; furthermore, F-PAD can adapt to various attributes (e.g., gender, age) and offer countermeasures to help users lower the risk. Extensive experiment studies on two real social network datasets, Facebook and Book-Crossing, have verified the effectiveness of F-PAD in `current city', `gender' and `age' disclosure risk estimation.