The importance of rationales, or natural language explanations, lies in their capacity to bridge the gap between machine predictions and human understanding, by providing human-readable insights into why a text classifier makes specific decisions. This paper presents a novel multi-task rationalisation approach tailored to enhancing the explainability of multi-label text classifiers to identify indicators of forced labour. Our framework integrates a rationale extraction task with the classification objective and allows the inclusion of human explanations during training. We conduct extensive experiments using transformer-based models on a dataset consisting of 2,800 news articles, each annotated with labels and human-generated explanations. Our findings reveal a statistically significant difference between the best-performing architecture leveraging human rationales during training and variants using only labels. Specifically, the supervised model demonstrates a 10% improvement in predictive performance measured by the weighted F1 score, a 15% increase in the agreement between human and machine-generated rationales, and a 4% improvement in the generated rationales’ comprehensiveness. These results hold promising implications for addressing complex human rights issues with greater transparency and accountability using advanced NLP techniques.

As in other NLP tasks, Automatic Short Answer Grading (ASAG) systems have evolved from using rule-based and interpretable machine learning models to utilizing deep learning architectures to boost accuracy. Since proper feedback is critical to student assessment, explainability will be crucial for deploying ASAG in real-world applications. This paper proposes a framework to generate explainable outcomes for assessing question-answer pairs of a Data Mining course in a binary manner. Our framework utilizes a fine-tuned Transformer-based classifier and an explainability module using SHAP or Integrated Gradients to generate language explanations for each prediction. We assess the outcome of our framework by calculating accuracy-based metrics for classification performance. Furthermore, we evaluate the quality of the explanations by measuring their agreement with human-annotated justifications using Intersection-Over-Union at a token level to derive a plausibility score. Despite the relatively limited sample, results show that our framework derives explanations that are, to some degree, aligned with domain-expert judgment. Furthermore, both explainability methods perform similarly in their agreement with human-annotated explanations. A natural progression of our work is to analyze the use of our explainable ASAG framework on a larger sample to determine the feasibility of implementing a pilot study in a real-world setting.

Forced labour is the most common type of modern slavery, and it is increasingly gaining the attention of the research and social community. Recent studies suggest that artificial intelligence (AI) holds immense potential for augmenting anti-slavery action. However, AI tools need to be developed transparently in cooperation with different stakeholders. Such tools are contingent on the availability and access to domain-specific data, which are scarce due to the near-invisible nature of forced labour. To the best of our knowledge, this paper presents the first openly accessible English corpus annotated for multi-class and multi-label forced labour detection. The corpus consists of 989 news articles retrieved from specialised data sources and annotated according to risk indicators defined by the International Labour Organization (ILO). Each news article was annotated for two aspects: (1) indicators of forced labour as classification labels and (2) snippets of the text that justify labelling decisions. We hope that our data set can help promote research on explainability for multi-class and multi-label text classification. In this work, we explain our process for collecting the data underpinning the proposed corpus, describe our annotation guidelines and present some statistical analysis of its content. Finally, we summarise the results of baseline experiments based on different variants of the Bidirectional Encoder Representation from Transformer (BERT) model.