Decision-making usually takes five steps: identifying the problem, collecting data, extracting evidence, identifying pro and con arguments, and making decisions. Focusing on extracting evidence, this paper presents a hybrid model that combines latent Dirichlet allocation and word embeddings to obtain external knowledge from structured and unstructured data. We study the task of sentence-level argument mining, as arguments mostly require some degree of world knowledge to be identified and understood. Given a topic and a sentence, the goal is to classify whether a sentence represents an argument in regard to the topic. We use a topic model to extract topic- and sentence-specific evidence from the structured knowledge base Wikidata, building a graph based on the cosine similarity between the entity word vectors of Wikidata and the vector of the given sentence. Also, we build a second graph based on topic-specific articles found via Google to tackle the general incompleteness of structured knowledge bases. Combining these graphs, we obtain a graph-based model which, as our evaluation shows, successfully capitalizes on both structured and unstructured data.

Existing deep learning models applied to reaction prediction in organic chemistry can reach high levels of accuracy (> 90% for Natural Language Processing-based ones). With no chemical knowledge embedded than the information learnt from reaction data, the quality of the data sets plays a crucial role in the performance of the prediction models. While human curation is prohibitively expensive, the need for unaided approaches to remove chemically incorrect entries from existing data sets is essential to improve artificial intelligence models' performance in synthetic chemistry tasks. Here we propose a machine learning-based, unassisted approach to remove chemically wrong entries from chemical reaction collections. We applied this method to the collection of chemical reactions Pistachio and to an open data set, both extracted from USPTO (United States Patent Office) patents. Our results show an improved prediction quality for models trained on the cleaned and balanced data sets. For the retrosynthetic models, the round-trip accuracy metric grows by 13 percentage points and the value of the cumulative Jensen Shannon divergence decreases by 30% compared to its original record. The coverage remains high with 97%, and the value of the class-diversity is not affected by the cleaning. The proposed strategy is the first unassisted rule-free technique to address automatic noise reduction in chemical data sets.

We study the impact of pre and post processing for reducing discrimination in data-driven decision makers. We first analyze the fundamental trade-off between fairness and accuracy in a pre-processing approach, and propose a design for a pre-processing module based on a convex optimization program, which can be added before the original classifier. This leads to a fundamental lower bound on attainable discrimination, given any acceptable distortion in the outcome. Furthermore, we reformulate an existing post-processing method in terms of our accuracy and fairness measures, which allows comparing post-processing and pre-processing approaches. We show that under some mild conditions, pre-processing outperforms post-processing. Finally, we show that by appropriate choice of the discrimination measure, the optimization problem for both pre and post processing approaches will reduce to a linear program and hence can be solved efficiently.

This paper unapologetically reflects on the critical role that Black feminism can and should play in abolishing algorithmic oppression. Positioning algorithmic oppression in the broader field of feminist science and technology studies, I draw upon feminist philosophical critiques of science and technology and discuss histories and continuities of scientific oppression against historically marginalized people. Moreover, I examine the concepts of invisibility and hypervisibility in oppressive technologies a l\'a the canonical double bind. Furthermore, I discuss what it means to call for diversity as a solution to algorithmic violence, and I critique dialectics of the fairness, accountability, and transparency community. I end by inviting you to envision and imagine the struggle to abolish algorithmic oppression by abolishing oppressive systems and shifting algorithmic development practices, including engaging our communities in scientific processes, centering marginalized communities in design, and consensual data and algorithmic practices.

Counterfactual estimation using synthetic controls is one of the most successful recent methodological developments in causal inference. Despite its popularity, the current description only considers time series aligned across units and synthetic controls expressed as linear combinations of observed control units. We propose a continuous-time alternative that models the latent counterfactual path explicitly using the formalism of controlled differential equations. This model is directly applicable to the general setting of irregularly-aligned multivariate time series and may be optimized in rich function spaces - thereby substantially improving on some limitations of existing approaches.

We survey over 100 face datasets constructed between 1976 to 2019 of 145 million images of over 17 million subjects from a range of sources, demographics and conditions. Our historical survey reveals that these datasets are contextually informed, shaped by changes in political motivations, technological capability and current norms. We discuss how such influences mask specific practices (some of which may actually be harmful or otherwise problematic) and make a case for the explicit communication of such details in order to establish a more grounded understanding of the technology's function in the real world.

Diversity in patent language is growing and makes finding synonyms for conducting patent searches more and more challenging. In addition to that, most approaches for dealing with diverse patent language are based on manual search and human intuition. In this paper, a word embedding approach using statistical analysis of human labeled data to produce accurate and language independent word vectors for technical terms is introduced. This paper focuses on the explanation of the idea behind the statistical analysis and shows first qualitative results. The resulting algorithm is a development of the former EQMania UG (eqmania.com)

How to attribute responsibility for autonomous artificial intelligence (AI) systems' actions has been widely debated across the humanities and social science disciplines. This work presents two experiments ($N$=200 each) that measure people's perceptions of eight different notions of moral responsibility concerning AI and human agents in the context of bail decision-making. Using real-life adapted vignettes, our experiments show that AI agents are held causally responsible and blamed similarly to human agents for an identical task. However, there was a meaningful difference in how people perceived these agents' moral responsibility; human agents were ascribed to a higher degree of present-looking and forward-looking notions of responsibility than AI agents. We also found that people expect both AI and human decision-makers and advisors to justify their decisions regardless of their nature. We discuss policy and HCI implications of these findings, such as the need for explainable AI in high-stakes scenarios.

This report prepared by the Montreal AI Ethics Institute provides recommendations in response to the National Security Commission on Artificial Intelligence (NSCAI) Key Considerations for Responsible Development and Fielding of Artificial Intelligence document. The report centres on the idea that Responsible AI should be made the Norm rather than an Exception. It does so by utilizing the guiding principles of: (1) alleviating friction in existing workflows, (2) empowering stakeholders to get buy-in, and (3) conducting an effective translation of abstract standards into actionable engineering practices. After providing some overarching comments on the document from the NSCAI, the report dives into the primary contribution of an actionable framework to help operationalize the ideas presented in the document from the NSCAI. The framework consists of: (1) a learning, knowledge, and information exchange (LKIE), (2) the Three Ways of Responsible AI, (3) an empirically-driven risk-prioritization matrix, and (4) achieving the right level of complexity. All components reinforce each other to move from principles to practice in service of making Responsible AI the norm rather than the exception.

This paper draws upon the evolutionary concepts of technological relatedness and knowledge complexity to enhance our understanding of the long-term evolution of Artificial Intelligence (AI). We reveal corresponding patterns in the emergence of AI - globally and in the context of specific geographies of the US, Japan, South Korea, and China. We argue that AI emergence is associated with increasing related variety due to knowledge commonalities as well as increasing complexity. We use patent-based indicators for the period between 1974-2018 to analyse the evolution of AI's global technological space, to identify its technological core as well as changes to its overall relatedness and knowledge complexity. At the national level, we also measure countries' overall specialisations against AI-specific ones. At the global level, we find increasing overall relatedness and complexity of AI. However, for the technological core of AI, which has been stable over time, we find decreasing related variety and increasing complexity. This evidence points out that AI innovations related to core technologies are becoming increasingly distinct from each other. At the country level, we find that the US and Japan have been increasing the overall relatedness of their innovations. The opposite is the case for China and South Korea, which we associate with the fact that these countries are overall less technologically developed than the US and Japan. Finally, we observe a stable increasing overall complexity for all countries apart from China, which we explain by the focus of this country in technologies not strongly linked to AI.