Trained classification models can unintentionally lead to biased representations and predictions, which can reinforce societal preconceptions and stereotypes. Existing debiasing methods for classification models, such as adversarial training, are often expensive to train and difficult to optimise. In this paper, we propose a method for mitigating bias in classifier training by incorporating contrastive learning, in which instances sharing the same class label are encouraged to have similar representations, while instances sharing a protected attribute are forced further apart. In such a way our method learns representations which capture the task label in focused regions, while ensuring the protected attribute has diverse spread, and thus has limited impact on prediction and thereby results in fairer models. Extensive experimental results across four tasks in NLP and computer vision show (a) that our proposed method can achieve fairer representations and realises bias reductions compared with competitive baselines; and (b) that it can do so without sacrificing main task performance; (c) that it sets a new state-of-the-art performance in one task despite reducing the bias. Finally, our method is conceptually simple and agnostic to network architectures, and incurs minimal additional compute cost.

Supervised learning is the machine learning task of learning a function that maps an input to an output based on example input-output pairs. A supervised learning algorithm analyzes the training data and produces an inferred function, which can be used later for mapping new examples. The most popular supervised learning tasks are: Regression and Classification. The result of solving the regression task is a model that can make numerical predictions. The result of solving the classification task is a model that can make classes predictions.

With the rapid development of artificial intelligence, conversational bots have became prevalent in mainstream E-commerce platforms, which can provide convenient customer service timely. To satisfy the user, the conversational bots need to understand the user's intention, detect the user's emotion, and extract the key entities from the conversational utterances. However, understanding dialogues is regarded as a very challenging task. Different from common language understanding, utterances in dialogues appear alternately from different roles and are usually organized as hierarchical structures. To facilitate the understanding of dialogues, in this paper, we propose a novel contextual dialogue encoder (i.e. DialogueBERT) based on the popular pre-trained language model BERT. Five self-supervised learning pre-training tasks are devised for learning the particularity of dialouge utterances. Four different input embeddings are integrated to catch the relationship between utterances, including turn embedding, role embedding, token embedding and position embedding. DialogueBERT was pre-trained with 70 million dialogues in real scenario, and then fine-tuned in three different downstream dialogue understanding tasks. Experimental results show that DialogueBERT achieves exciting results with 88.63% accuracy for intent recognition, 94.25% accuracy for emotion recognition and 97.04% F1 score for named entity recognition, which outperforms several strong baselines by a large margin.

Imagine having an artificial intelligence (AI) system that is capable of mimicking human language and intelligence. Given AI's capabilities, it seems simple, right? Despite recent advancements in AI (especially in the fields of natural language processing (NLP) and computer vision applications), mastering the unique complexities of human language continues to be one of AI's biggest challenges. According to IDC, worldwide revenues for the AI market are forecast to grow 16.4 percent year over year in 2021, as the market is expected to break the $500 billion mark by 2024. As companies continue to develop and deploy AI solutions to automate processes, solve complex problems and enhance customer experiences, many are realizing its shortcomings -- including the amount of data required to train machine learning (ML) algorithms and the flexibility of these algorithms in understanding human language.

Learning information-rich and generalizable representations effectively from unlabeled multivariate cardiac signals to identify abnormal heart rhythms (cardiac arrhythmias) is valuable in real-world clinical settings but often challenging due to its complex temporal dynamics. Cardiac arrhythmias can vary significantly in temporal patterns even for the same patient ($i.e.$, intra subject difference). Meanwhile, the same type of cardiac arrhythmia can show different temporal patterns among different patients due to different cardiac structures ($i.e.$, inter subject difference). In this paper, we address the challenges by proposing an Intra-inter Subject self-supervised Learning (ISL) model that is customized for multivariate cardiac signals. Our proposed ISL model integrates medical knowledge into self-supervision to effectively learn from intra-inter subject differences. In intra subject self-supervision, ISL model first extracts heartbeat-level features from each subject using a channel-wise attentional CNN-RNN encoder. Then a stationarity test module is employed to capture the temporal dependencies between heartbeats. In inter subject self-supervision, we design a set of data augmentations according to the clinical characteristics of cardiac signals and perform contrastive learning among subjects to learn distinctive representations for various types of patients. Extensive experiments on three real-world datasets were conducted. In a semi-supervised transfer learning scenario, our pre-trained ISL model leads about 10% improvement over supervised training when only 1% labeled data is available, suggesting strong generalizability and robustness of the model.

Imagine having an artificial intelligence (AI) system that is capable of mimicking human language and intelligence. Given AI's capabilities, it seems simple, right? Despite recent advancements in AI (especially in the fields of natural language processing (NLP) and computer vision applications), mastering the unique complexities of human language continues to be one of AI's biggest challenges. According to IDC, worldwide revenues for the AI market are forecast to grow 16.4 percent year over year in 2021, as the market is expected to break the $500 billion mark by 2024. As companies continue to develop and deploy AI solutions to automate processes, solve complex problems and enhance customer experiences, many are realizing its shortcomings -- including the amount of data required to train machine learning (ML) algorithms and the flexibility of these algorithms in understanding human language.

We compare two orthogonal semi-supervised learning techniques, namely tri-training and pretrained word embeddings, in the task of dependency parsing. We explore language-specific FastText and ELMo embeddings and multilingual BERT embeddings. We focus on a low resource scenario as semi-supervised learning can be expected to have the most impact here. Based on treebank size and available ELMo models, we select Hungarian, Uyghur (a zero-shot language for mBERT) and Vietnamese. Furthermore, we include English in a simulated low-resource setting. We find that pretrained word embeddings make more effective use of unlabelled data than tri-training but that the two approaches can be successfully combined.

Event extraction (EE) aims to identify structured events, including event triggers and their corresponding arguments, from unstructured text. Most of the existing works rely on a large number of labeled instances to train models, while the labeled data could be expensive to be obtained. In this work, we present a data-efficient event extraction method by formulating event extraction as a natural language generation problem. The formulation allows us to inject knowledge of label semantics, event structure, and output dependencies into the model. Given a passage and an event type, our model learns to summarize this passage into a templated sentence in a predefined structure. The template is event-type-specific, manually created, and contains event trigger and argument information. Lastly, a rule-based algorithm is used to derive the trigger and argument predictions from the generated sentence. Our method inherently enjoys the following benefits: (1) The pretraining of the generative language models help incorporate the semantics of the labels for generative EE. (2) The autoregressive generation process and our end-to-end design for extracting triggers and arguments force the model to capture the dependencies among the output triggers and their arguments. (3) The predefined templates form concrete yet flexible rules to hint the models about the valid patterns for each event type, reducing the models' burden to learn structures from the data. Empirical results show that our model achieves superior performance over strong baselines on EE tasks in the low data regime and achieves competitive results to the current state-of-the-art when more data becomes available.

Knowledge transfer from a complex high performing model to a simpler and potentially low performing one in order to enhance its performance has been of great interest over the last few years as it finds applications in important problems such as explainable artificial intelligence, model compression, robust model building and learning from small data. Known approaches to this problem (viz. Knowledge Distillation, Model compression, ProfWeight, etc.) typically transfer information directly (i.e. in a single/one hop) from the complex model to the chosen simple model through schemes that modify the target or reweight training examples on which the simple model is trained. In this paper, we propose a meta-approach where we transfer information from the complex model to the simple model by dynamically selecting and/or constructing a sequence of intermediate models of decreasing complexity that are less intricate than the original complex model. Our approach can transfer information between consecutive models in the sequence using any of the previously mentioned approaches as well as work in 1-hop fashion, thus generalizing these approaches. In the experiments on real data, we observe that we get consistent gains for different choices of models over 1-hop, which on average is more than 2\% and reaches up to 8\% in a particular case. We also empirically analyze conditions under which the multi-hop approach is likely to be beneficial over the traditional 1-hop approach, and report other interesting insights. To the best of our knowledge, this is the first work that proposes such a multi-hop approach to perform knowledge transfer given a single high performing complex model, making it in our opinion, an important methodological contribution.

Text classification tasks have improved substantially during the last years by the usage of transformers. However, the majority of researches focus on prose texts, with poetry receiving less attention, specially for Spanish language. In this paper, we propose a semi-supervised learning approach for inferring 21 psychological categories evoked by a corpus of 4572 sonnets, along with 10 affective and lexico-semantic multiclass ones. The subset of poems used for training an evaluation includes 270 sonnets. With our approach, we achieve an AUC beyond 0.7 for 76% of the psychological categories, and an AUC over 0.65 for 60% on the multiclass ones. The sonnets are modelled using transformers, through sentence embeddings, along with lexico-semantic and affective features, obtained by using external lexicons. Consequently, we see that this approach provides an AUC increase of up to 0.12, as opposed to using transformers alone.