To make machines better understand sentiments, research needs to move from polarity identification to understanding the reasons that underlie the expression of sentiment. Categorizing the goals or needs of humans is one way to explain the expression of sentiment in text. Humans are good at understanding situations described in natural language and can easily connect them to the character's psychological needs using commonsense knowledge. We present a novel method to extract, rank, filter and select multi-hop relation paths from a commonsense knowledge resource to interpret the expression of sentiment in terms of their underlying human needs. We efficiently integrate the acquired knowledge paths in a neural model that interfaces context representations with knowledge using a gated attention mechanism. We assess the model's performance on a recently published dataset for categorizing human needs. Selectively integrating knowledge paths boosts performance and establishes a new state-of-the-art. Our model offers interpretability through the learned attention map over commonsense knowledge paths. Human evaluation highlights the relevance of the encoded knowledge.

Social Commonsense Reasoning requires understanding of text, knowledge about social events and their pragmatic implications, as well as commonsense reasoning skills. In this work we propose a novel multi-head knowledge attention model that encodes semistructured commonsense inference rules and learns to incorporate them in a transformerbased reasoning cell. We assess the model's performance on two tasks that require different reasoning skills: Abductive Natural Language Inference and Counterfactual Invariance Prediction as a new task. We show that our proposed model improves performance over strong state-of-the-art models (i.e., RoBERTa) across both reasoning tasks. Notably we are, to the best of our knowledge, the first to demonstrate that a model that learns to perform counterfactual reasoning helps predicting the best explanation in an abductive reasoning task. We validate the robustness of the model's reasoning capabilities by perturbing the knowledge and provide qualitative analysis on the model's knowledge incorporation capabilities.

In this paper, we address the problem of effectively self-training neural networks in a lowresource setting. Self-training is frequently used to automatically increase the amount of training data. However, in a low-resource scenario, it is less effective due to unreliable annotations created using self-labeling of unlabeled data. We propose to combine self-training with noise handling on the self-labeled data. Directly estimating noise on the combined clean training set and self-labeled data can lead to corruption of the clean data and hence, performs worse. Thus, we propose the Clean and Noisy Label Neural Network which trains on clean and noisy self-labeled data simultaneously by explicitly modelling clean and noisy labels separately. In our experiments on Chunking and NER, this approach performs more robustly than the baselines. Complementary to this explicit approach, noise can also be handled implicitly with the help of an auxiliary learning task. To such a complementary approach, our method is more beneficial than other baseline methods and together provides the best performance overall.

Most information we consume as a society is obtained over the Web. News-often from questionable sources-are spread online, as are election campaigns; calls for (collective) action spread with unforeseen speed and intensity. All such actions have argumentation at their core, and the conveyed content is often strategically selected or rhetorically framed. The responsibility of critical analysis of arguments is thus tacitly transferred to the content consumer who is often not prepared for the task, nor aware of the responsibility. The ExpLAIN project aims at making the structure and reasoning of arguments explicit-not only for humans, but for Robust Argumentation Machines that are endowed with language understanding capacity. Our vision is a system that is able to deeply analyze argumentative text: that identifies arguments and counter-arguments, and reveals their internal structure, conveyed content and reasoning. A particular challenge for such a system is to uncover implicit knowledge which many arguments rely on. This requires human background knowledge and reasoning capacity, in order to explicate the complete reasoning of an argument. This article presents ongoing research of the ExpLAIN project that aims to make the vision of such a system a tangible aim. We introduce the problems and challenges we need to address, and present the progress we achieved until now by applying advanced natural language and knowledge processing methods. Our approach puts particular focus on leveraging available sources of structured and unstructured background knowledge, the automatic extension of such knowledge, the uncovering of implicit content, and reasoning techniques suitable for informal, everyday argumentation.

Social Commonsense Reasoning requires understanding of text, knowledge about social events and their pragmatic implications, as well as commonsense reasoning skills. In this work we propose a novel multi-head knowledge attention model that encodes semistructured commonsense inference rules and learns to incorporate them in a transformerbased reasoning cell. We assess the model's performance on two tasks that require different reasoning skills: Abductive Natural Language Inference and Counterfactual Invariance Prediction as a new task. We show that our proposed model improves performance over strong state-of-the-art models (i.e., RoBERTa) across both reasoning tasks. Notably we are, to the best of our knowledge, the first to demonstrate that a model that learns to perform counterfactual reasoning helps predicting the best explanation in an abductive reasoning task. We validate the robustness of the model's reasoning capabilities by perturbing the knowledge and provide qualitative analysis on the model's knowledge incorporation capabilities.