We present a new resource for the NLP community, namely a large (3.5M+ sentence) knowledge base of *generic statements*, e.g., "Trees remove carbon dioxide from the atmosphere", collected from multiple corpora. This is the first large resource to contain *naturally occurring* generic sentences, as opposed to extracted or crowdsourced triples, and thus is rich in high-quality, general, semantically complete statements. All GenericsKB sentences are annotated with their topical term, surrounding context (sentences), and a (learned) confidence. We also release GenericsKB-Best (1M+ sentences), containing the best-quality generics in GenericsKB augmented with selected, synthesized generics from WordNet and ConceptNet. In tests on two existing datasets requiring multihop reasoning (OBQA and QASC), we find using GenericsKB can result in higher scores and better explanations than using a much larger corpus. This demonstrates that GenericsKB can be a useful resource for NLP applications, as well as providing data for linguistic studies of generics and their semantics. GenericsKB is available at https://allenai.org/data/genericskb.

The term AI has been around since the 1950s. In short, it depicts our struggle to build machines that can challenge what made humans the dominant lifeform on the planet: our intelligence. However, defining "intelligence" has turned out to be rather tricky, because what we perceive as intelligent changes over time. Early AIs were rule-based computer programs that could solve somewhat complex problems. Instead of hardcoding every decision the software was supposed to make, the program was divided into a knowledge base and an inference engine. Developers would fill out the knowledge base with facts, and the inference engine would then query those facts to arrive at results.

The term AI has been around since the 1950s. In short, it depicts our struggle to build machines that can challenge what made humans the dominant lifeform on the planet: our intelligence. However, defining "intelligence" has turned out to be rather tricky, because what we perceive as intelligent changes over time. Early AIs were rule-based computer programs that could solve somewhat complex problems. Instead of hardcoding every decision the software was supposed to make, the program was divided into a knowledge base and an inference engine. Developers would fill out the knowledge base with facts, and the inference engine would then query those facts to arrive at results.

Existential rules are a positive fragment of first-order logic that generalizes function-free Horn rules by allowing existentially quantified variables in rule heads. This family of languages has recently attracted significant interest in the context of ontology-mediated query answering. Forward chaining, also known as the chase, is a fundamental tool for computing universal models of knowledge bases, which consist of existential rules and facts. Several chase variants have been defined, which differ on the way they handle redundancies. A set of existential rules is bounded if it ensures the existence of a bound on the depth of the chase, independently from any set of facts. Deciding if a set of rules is bounded is an undecidable problem for all chase variants. Nevertheless, when computing universal models, knowing that a set of rules is bounded for some chase variant does not help much in practice if the bound remains unknown or even very large. Hence, we investigate the decidability of the k-boundedness problem, which asks whether the depth of the chase for a given set of rules is bounded by an integer k. We identify a general property which, when satisfied by a chase variant, leads to the decidability of k-boundedness. We then show that the main chase variants satisfy this property, namely the oblivious, semi-oblivious (aka Skolem), and restricted chase, as well as their breadth-first versions. This paper is under consideration for publication in Theory and Practice of Logic Programming.

Knowledge present in a domain is well expressed as relationships between corresponding concepts. For example, in zoology, animal species form complex hierarchies; in genomics, the different (parts of) molecules are organized in groups and subgroups based on their functions; plants, molecules, and astronomical objects all form complex taxonomies. Nevertheless, when applying supervised machine learning (ML) in such domains, we commonly reduce the complex and rich knowledge to a fixed set of labels, and induce a model shows good generalization performance with respect to these labels. The main reason for such a reductionist approach is the difficulty in eliciting the domain knowledge from the experts. Developing a label structure with sufficient fidelity and providing comprehensive multi-label annotation can be exceedingly labor-intensive in many real-world applications. In this paper, we provide a method for efficient hierarchical knowledge elicitation (HKE) from experts working with high-dimensional data such as images or videos. Our method is based on psychometric testing and active deep metric learning. The developed models embed the high-dimensional data in a metric space where distances are semantically meaningful, and the data can be organized in a hierarchical structure. We provide empirical evidence with a series of experiments on a synthetically generated dataset of simple shapes, and Cifar 10 and Fashion-MNIST benchmarks that our method is indeed successful in uncovering hierarchical structures.

Most algorithms for representation learning and link prediction in relational data have been designed for static data. However, the data they are applied to usually evolves with time, such as friend graphs in social networks or user interactions with items in recommender systems. This is also the case for knowledge bases, which contain facts such as (US, has president, B. Obama, [2009-2017]) that are valid only at certain points in time. For the problem of link prediction under temporal constraints, i.e., answering queries such as (US, has president, ?, 2012), we propose a solution inspired by the canonical decomposition of tensors of order 4. We introduce new regularization schemes and present an extension of ComplEx (Trouillon et al., 2016) that achieves state-of-the-art performance. Additionally, we propose a new dataset for knowledge base completion constructed from Wikidata, larger than previous benchmarks by an order of magnitude, as a new reference for evaluating temporal and non-temporal link prediction methods.

Most of todays work on knowledge graph completion is concerned with sub-symbolic approaches that focus on the concept of embedding a given graph in a low dimensional vector space. Against this trend, we propose an approach called AnyBURL that is rooted in the symbolic space. Its core algorithm is based on sampling paths, which are generalized into Horn rules. Previously published results show that the prediction quality of AnyBURL is on the same level as current state of the art with the additional benefit of offering an explanation for the predicted fact. In this paper, we are concerned with two extensions of AnyBURL. Firstly, we change AnyBURLs interpretation of rules from $\Theta$-subsumption into $\Theta$-subsumption under Object Identity. Secondly, we introduce reinforcement learning to better guide the sampling process. We found out that reinforcement learning helps finding more valuable rules earlier in the search process. We measure the impact of both extensions and compare the resulting approach with current state of the art approaches. Our results show that AnyBURL outperforms most sub-symbolic methods.

We study the problem of generating inferential texts of events for a variety of commonsense like \textit{if-else} relations. Existing approaches typically use limited evidence from training examples and learn for each relation individually. In this work, we use multiple knowledge sources as fuels for the model. Existing commonsense knowledge bases like ConceptNet are dominated by taxonomic knowledge (e.g., \textit{isA} and \textit{relatedTo} relations), having a limited number of inferential knowledge. We use not only structured commonsense knowledge bases, but also natural language snippets from search-engine results. These sources are incorporated into a generative base model via key-value memory network. In addition, we introduce a meta-learning based multi-task learning algorithm. For each targeted commonsense relation, we regard the learning of examples from other relations as the meta-training process, and the evaluation on examples from the targeted relation as the meta-test process. We conduct experiments on Event2Mind and ATOMIC datasets. Results show that both the integration of multiple knowledge sources and the use of the meta-learning algorithm improve the performance.

Many areas of research are characterised by the deluge of large-scale highly-dimensional time-series data. However, using the data available for prediction and decision making is hampered by the current lag in our ability to uncover and quantify true interactions that explain the outcomes.We are interested in areas such as intensive care medicine, which are characterised by i) continuous monitoring of multivariate variables and non-uniform sampling of data streams, ii) the outcomes are generally governed by interactions between a small set of rare events, iii) these interactions are not necessarily definable by specific values (or value ranges) of a given group of variables, but rather, by the deviations of these values from the normal state recorded over time, iv) the need to explain the predictions made by the model. Here, while numerous data mining models have been formulated for outcome prediction, they are unable to explain their predictions. We present a model for uncovering interactions with the highest likelihood of generating the outcomes seen from highly-dimensional time series data. Interactions among variables are represented by a relational graph structure, which relies on qualitative abstractions to overcome non-uniform sampling and to capture the semantics of the interactions corresponding to the changes and deviations from normality of variables of interest over time. Using the assumption that similar templates of small interactions are responsible for the outcomes (as prevalent in the medical domains), we reformulate the discovery task to retrieve the most-likely templates from the data.

The paper presents a framework to reconstruct explanations for multiple choices science questions through explanation-centred corpora. Building upon the notion of unification in science, the framework ranks explanatory facts with respect to question and candidate answer by leveraging a combination of two different scores: (a) A Relevance Score (RS) that represents the extent to which a given fact is specific to the question; (b) A Unification Score (US) that takes into account the explanatory power of a fact, determined according to its frequency in explanations for similar questions. An extensive evaluation of the framework is performed on the Worldtree corpus, adopting IR weighting schemes for its implementation. The following findings are presented: (1) The proposed approach achieves competitive results when compared to state-of-the-art Transformers, yet possessing the property of being scalable to large explanatory knowledge bases; (2) The combined model significantly outperforms IR baselines ( 7.8/8.4 MAP), confirming the complementary aspects of relevance and unification score; (3) The constructed explanations can support downstream models for answer prediction, improving the accuracy of BERT for multiple choices QA on both ARC easy ( 6.92%) and challenge ( 15.69%) questions.