Fox News national correspondent Bryan Llenas has the latest as Pennsylvania senate nominees make last plea to voters ahead of Tuesday's midterm elections. The balance of power in Washington, D.C., may not be settled when Election Day comes to an end – or the day after, or the day after that, depending on how things go in Pennsylvania. The Senate race between Democrat John Fetterman and Republican Dr. Mehmet Oz remained a toss-up in the final Fox News Power Rankings before Election Day, and litigation over absentee ballots could draw the process out well beyond November 8. The state's supreme court ruled that undated and incorrectly dated absentee ballots had to be set aside, uncounted, and federal litigation related to this remains ongoing. During a press conference on Monday, Pennsylvania Acting Secretary of State Leigh M. Chapman confirmed that voters worried that they already made mistakes may not be able to fix them.

Multi-label classification is becoming increasingly ubiquitous, but not much attention has been paid to interpretability. In this paper, we develop a multi-label classifier that can be represented as a concise set of simple "if-then" rules, and thus, it offers better interpretability compared to black-box models. Notably, our method is able to find a small set of relevant patterns that lead to accurate multi-label classification, while existing rule-based classifiers are myopic and wasteful in searching rules,requiring a large number of rules to achieve high accuracy. In particular, we formulate the problem of choosing multi-label rules to maximize a target function, which considers not only discrimination ability with respect to labels, but also diversity. Accounting for diversity helps to avoid redundancy, and thus, to control the number of rules in the solution set. To tackle the said maximization problem we propose a 2-approximation algorithm, which relies on a novel technique to sample high-quality rules. In addition to our theoretical analysis, we provide a thorough experimental evaluation, which indicates that our approach offers a trade-off between predictive performance and interpretability that is unmatched in previous work.

The pattern of pedestrian crashes varies greatly depending on lighting circumstances, emphasizing the need of examining pedestrian crashes in various lighting conditions. Using Louisiana pedestrian fatal and injury crash data (2010-2019), this study applied Association Rules Mining (ARM) to identify the hidden pattern of crash risk factors according to three different lighting conditions (daylight, dark-with-streetlight, and dark-no-streetlight). Based on the generated rules, the results show that daylight pedestrian crashes are associated with children (less than 15 years), senior pedestrians (greater than 64 years), older drivers (>64 years), and other driving behaviors such as failure to yield, inattentive/distracted, illness/fatigue/asleep. Additionally, young drivers (15-24 years) are involved in severe pedestrian crashes in daylight conditions. This study also found pedestrian alcohol/drug involvement as the most frequent item in the dark-with-streetlight condition. This crash type is particularly associated with pedestrian action (crossing intersection/midblock), driver age (55-64 years), speed limit (30-35 mph), and specific area type (business with mixed residential area). Fatal pedestrian crashes are found to be associated with roadways with high-speed limits (>50 mph) during the dark without streetlight condition. Some other risk factors linked with high-speed limit related crashes are pedestrians walking with/against the traffic, presence of pedestrian dark clothing, pedestrian alcohol/drug involvement. The research findings are expected to provide an improved understanding of the underlying relationships between pedestrian crash risk factors and specific lighting conditions. Highway safety experts can utilize these findings to conduct a decision-making process for selecting effective countermeasures to reduce pedestrian crashes strategically.

With 84.75 million Filipinos online, the ability for models to process online text is crucial for developing Filipino NLP applications. To this end, spelling correction is a crucial preprocessing step for downstream processing. However, the lack of data prevents the use of language models for this task. In this paper, we propose an N-Gram + Damerau Levenshtein distance model with automatic rule extraction. We train the model on 300 samples, and show that despite limited training data, it achieves good performance and outperforms other deep learning approaches in terms of accuracy and edit distance. Moreover, the model (1) requires little compute power, (2) trains in little time, thus allowing for retraining, and (3) is easily interpretable, allowing for direct troubleshooting, highlighting the success of traditional approaches over more complex deep learning models in settings where data is unavailable.

Rule-based classification models described in the language of logic directly predict boolean values, rather than modeling a probability and translating it into a prediction as done in statistical models. The vast majority of existing uncertainty quantification approaches rely on models providing continuous output not available to rule-based models. In this work, we propose an uncertainty quantification framework in the form of a meta-model that takes any binary classifier with binary output as a black box and estimates the prediction accuracy of that base model at a given input along with a level of confidence on that estimation. The confidence is based on how well that input region is explored and is designed to work in any OOD scenario. We demonstrate the usefulness of this uncertainty model by building an abstaining classifier powered by it and observing its performance in various scenarios.

AI image generators, which create fantastical sights at the intersection of dreams and reality, bubble up on every corner of the web. Their entertainment value is demonstrated by an ever-expanding treasure trove of whimsical and random images serving as indirect portals to the brains of human designers. A simple text prompt yields a nearly instantaneous image, satisfying our primitive brains, which are hardwired for instant gratification. Although seemingly nascent, the field of AI-generated art can be traced back as far as the 1960s with early attempts using symbolic rule-based approaches to make technical images. Yilun Du, a Ph.D. student in the Department of Electrical Engineering and Computer Science and affiliate of MIT's Computer Science and Artificial Intelligence Laboratory (CSAIL), recently developed a new method that makes models like DALL-E 2 more creative and have better scene understanding. Here, Du describes how these models work, whether this technical infrastructure can be applied to other domains, and how we draw the line between AI and human creativity.

Interaction is a fundamental part of using any computer system but it is still an issue for people with special needs. In order to improve this situation, this paper describes a new device-interaction model based on adaptation rules for user models. The aim is the adaptation at the interaction level, taking into account the interaction device features in order to improve the usability through the user experience in the education sector. In the evaluation process, several students from a special education center have participated. These students have either a physical or sensory disability or autism. The results are promising enough to consider that this model will be able to help students with disabilities to interact with a computer system which will inevitably provide tremendous benefits to their academic and personal development.

Regulations govern many aspects of citizens' daily lives. Governments and businesses routinely automate these in the form of coded rules (e.g., to check a citizen's eligibility for specific benefits). However, the path to automation is long and challenging. To address this, recent global initiatives for digital government, proposing to simultaneously express policy in natural language for human consumption as well as computationally amenable rules or code, are gathering broad public-sector interest. We introduce the problem of semi-automatically building decision models from eligibility policies for social services, and present an initial emerging approach to shorten the route from policy documents to executable, interpretable and standardised decision models using AI, NLP and Knowledge Graphs. Despite the many open domain challenges, in this position paper we explore the enormous potential of AI to assist government agencies and policy experts in scaling the production of both human-readable and machine executable policy rules, while improving transparency, interpretability, traceability and accountability of the decision making.

Multi-hop knowledge graph (KG) reasoning has been widely studied in recent years to provide interpretable predictions on missing links with evidential paths. Most previous works use reinforcement learning (RL) based methods that learn to navigate the path towards the target entity. However, these methods suffer from slow and poor convergence, and they may fail to infer a certain path when there is a missing edge along the path. Here we present SQUIRE, the first Sequence-to-sequence based multi-hop reasoning framework, which utilizes an encoder-decoder Transformer structure to translate the query to a path. Our framework brings about two benefits: (1) It can learn and predict in an end-to-end fashion, which gives better and faster convergence; (2) Our Transformer model does not rely on existing edges to generate the path, and has the flexibility to complete missing edges along the path, especially in sparse KGs. Experiments on standard and sparse KGs show that our approach yields significant improvement over prior methods, while converging 4x-7x faster.

Abstractive summarization models often generate inconsistent summaries containing factual errors or hallucinated content. Recent works focus on correcting factual errors in generated summaries via post-editing. Such correction models are trained using adversarial non-factual summaries constructed using heuristic rules for injecting errors. However, generating non-factual summaries using heuristics often does not generalize well to actual model errors. In this work, we propose to generate hard, representative synthetic examples of non-factual summaries through infilling language models. With this data, we train a more robust fact-correction model to post-edit the summaries to improve factual consistency. Through quantitative and qualitative experiments on two popular summarization datasets -- CNN/DM and XSum -- we show that our approach vastly outperforms prior methods in correcting erroneous summaries. Our model -- FactEdit -- improves factuality scores by over ~11 points on CNN/DM and over ~31 points on XSum on average across multiple summarization models, producing more factual summaries while maintaining competitive summarization quality.