Gender bias may emerge from an unequal representation of agency and power, for example, by portraying women frequently as passive and powerless ("She accepted her future'') and men as proactive and powerful ("He chose his future''). When language models learn from respective texts, they may reproduce or even amplify the bias. An effective way to mitigate bias is to generate counterfactual sentences with opposite agency and power to the training. Recent work targeted agency-specific verbs from a lexicon to this end. We argue that this is insufficient, due to the interaction of agency and power and their dependence on context. In this paper, we thus develop a new rewriting model that identifies verbs with the desired agency and power in the context of the given sentence. The verbs' probability is then boosted to encourage the model to rewrite both connotations jointly. According to automatic metrics, our model effectively controls for power while being competitive in agency to the state of the art. In our evaluation, human annotators favored its counterfactuals in terms of both connotations, also deeming its meaning preservation better.

Word embedding models reflect bias towards genders, ethnicities, and other social groups present in the underlying training data. Metrics such as ECT, RNSB, and WEAT quantify bias in these models based on predefined word lists representing social groups and bias-conveying concepts. How suitable these lists actually are to reveal bias - let alone the bias metrics in general - remains unclear, though. In this paper, we study how to assess the quality of bias metrics for word embedding models. In particular, we present a generic method, Bias Silhouette Analysis (BSA), that quantifies the accuracy and robustness of such a metric and of the word lists used. Given a biased and an unbiased reference embedding model, BSA applies the metric systematically for several subsets of the lists to the models. The variance and rate of convergence of the bias values of each model then entail the robustness of the word lists, whereas the distance between the models' values gives indications of the general accuracy of the metric with the word lists. We demonstrate the behavior of BSA on two standard embedding models for the three mentioned metrics with several word lists from existing research.

Overcoming a range of challenges that traditional therapy faces, VRET yields great potential for the treatment of phobias such as acrophobia, the fear of heights. We investigate this potential and present playful user-generated treatment (PUT), a novel game-based approach for VRET. Based on a requirement analysis consisting of a literature review and semi-structured interviews with professional therapists, we designed and implemented the PUT concept as a two-step VR game design. To validate our approach, we conducted two studies. (1) In a study with 31 non-acrophobic subjects, we investigated the effect of content creation on player experience, motivation and height perception, and (2) in an online survey, we collected feedback from professional therapists. Both studies reveal that the PUT approach is well applicable. In particular, the analysis of the user study shows that the design phase leads to increased interest and enjoyment without notably influencing affective measures during the exposure session. Our work can help guiding researchers and practitioners at the intersection of game design and exposure therapy.

Attention mechanisms have seen some success for natural language processing downstream tasks in recent years and generated new state-of-the-art results. A thorough evaluation of the attention mechanism for the task of Argumentation Mining is missing. With this paper, we report a comparative evaluation of attention layers in combination with a bidirectional long short-term memory network, which is the current state-of-the-art approach for the unit segmentation task. We also compare sentence-level contextualized word embeddings to pre-generated ones. Our findings suggest that for this task, the additional attention layer does not improve the performance. In most cases, contextualized embeddings do also not show an improvement on the score achieved by pre-defined embeddings.