Home robots may come with many sophisticated built-in abilities, however there will always be a degree of customization needed for each user and environment. Ideally this should be accomplished through one-shot learning, as collecting the large number of examples needed for statistical inference is tedious. A particularly appealing approach is to simply explain to the robot, via speech, what it should be doing. In this paper we describe the ALIA cognitive architecture that is able to effectively incorporate user-supplied advice and prohibitions in this manner. The functioning of the implemented system on a small robot is illustrated by an associated video [11]. 1 INTRODUCTION A typical home robot of the future might have built-in navigation, object recognition, task planning, and dexterous manipulation. Y et, despite these sophisticated capabilities, there are still things it cannot know when it first arrives. For instance, what a particular room in the house is called, even if it can identify the general type.

The purpose of this paper is to describe and analyze the development of a knowledge-based infrastructure to support MSP decision-making processes. The paper emerged from a study to define specifications for a knowledge-based infrastructure to provide decision support for community-level MSPs in the Canadian province of Quebec. As part of the study, a process assessment was conducted to understand the needs of communities as they collect, organize, and analyze data to make decisions about their priorities. The result of this process is a portrait, which is an epidemiological profile of health and nutrition in their community. Portraits inform strategic planning and development of interventions and are used to assess the impact of interventions. Our key findings indicate ambiguities and disagreement among MSP decision-makers regarding causal relationships between actions and outcomes, and the relevant data needed for making decisions. MSP decision-makers expressed a desire for easy-to-use tools that facilitate the collection, organization, synthesis, and analysis of data, to enable decision-making in a timely manner. Findings inform conceptual modeling and ontological analysis to capture the domain knowledge and specify relationships between actions and outcomes. This modeling and analysis provide the foundation for an ontology, encoded using OWL 2 Web Ontology Language. The ontology is developed to provide semantic support for the MSP process, defining objectives, strategies, actions, indicators, and data sources. In the future, software interacting with the ontology can facilitate interactive browsing by decision-makers in the MSP in the form of concepts, instances, relationships, and axioms. Our ontology also facilitates the integration and interpretation of community data and can help in managing semantic interoperability between different knowledge sources.

Learning how to automatically solve optimization problems has the potential to provide the next big leap in optimization technology. The performance of automatically learned heuristics on routing problems has been steadily improving in recent years, but approaches based purely on machine learning are still outperformed by state-of-the-art optimization methods. To close this performance gap, we propose a novel large neighborhood search (LNS) framework for vehicle routing that integrates learned heuristics for generating new solutions. The learning mechanism is based on a deep neural network with an attention mechanism and has been especially designed to be integrated into an LNS search setting. We evaluate our approach on the capacitated vehicle routing problem (CVRP) and the split delivery vehicle routing problem (SDVRP). On CVRP instances with up to 297 customers our approach significantly outperforms an LNS that uses only handcrafted heuristics and a well-known heuristic from the literature. Furthermore, we show for the CVRP and the SDVRP that our approach surpasses the performance of existing machine learning approaches and comes close to the performance of state-of-the-art optimization approaches.

We survey a burgeoning and promising new research area that considers the online nature of many practical fair division problems. We identify wide variety of such online fair division problems, as well as discuss new mechanisms and normative properties that apply to this online setting. The online nature of such fair division problems provides both opportunities and challenges such as the possibility to develop new online mechanisms as well as the difficulty of dealing with an uncertain future. Introduction Fair division (Brams and Taylor 1996) is an important problem facing society today as increasing economical, environmental, and other pressures require us to try to do more with limited resources. Much previous work in fair division assumes the problem is offline and fixed. That is, we suppose that the agents being allocated resources, and the resources being allocated to these agents are all known and fixed. But practical reality is often quite different (Walsh 2014a; 2015). Fair division problems are often online, with either the agents, or the resources to be allocated, or both not being fixed and potentially changing over time.

What Do Y ou Mean'Why?': Resolving Sluices in Conversations Victor Petr en Bach Hansen, 1 2 Anders Søgaard 1 3 1 Department of Computer Science, University of Copenhagen, Denmark 2 Topdanmark A/S, Denmark 3 Google Research, Berlin victor.petren@di.ku.dk, soegaard@di.ku.dk Abstract In conversation, we often ask one-word questions such as'Why?' or'Who?'. Such questions are typically easy for humans to answer, but can be hard for computers, because their resolution requires retrieving both the right semantic frames and the right arguments from context. This paper introduces the novel ellipsis resolution task of resolving such one-word questions, referred to as sluices in linguistics. We present a crowd-sourced dataset containing annotations of sluices from over 4,000 dialogues collected from conversational QA datasets, as well as a series of strong baseline architectures. 1 Introduction Stand-alone wh-word questions, such as When? in Figure 1, are easy for us to understand, but in order to interpret them we need to retrieve implicit information from context. Learning to do so is an instance of sluicing, an ellipsis phenomenon, defined by Ross (1969) as'the effect of deleting everything but the preposed constituent of an embedded question, under the condition that the remainder of the question is identical to some other part of the sentence, or a preceding sentence.' In the context of conversations, one-word wh-word questions are particularly frequent (Anand and Hardt 2016; Rønning, Hardt, and Søgaard 2018), and because they are often hard to resolve, they seem to be a frequent source of error in conversational question answering (Choi et al. 2018; Reddy, Chen, and Manning 2018) and dialogue understanding (Vlachos and Clark 2014). We refer to this type of sluicing as conversational sluicing . Unlike previous work where sluice resolution is treated as predicting the span of the antecedent (Anand and Hardt 2016; Rønning, Hardt, and Søgaard 2018), we frame conversational sluice resolution as a Natural Language Generation (NLG) task, in which we seek to automatically generate the full question, given a question-answer context and a one-word question. Q 1: Where was the bombing?

This paper presents a novel incremental learning algorithm for pedestrian motion prediction, with the ability to improve the learned model over time when data is incrementally available. In this setup, trajectories are modeled as simple segments called motion primitives. Transitions between motion primitives are modeled as Gaussian Processes. When new data is available, the motion primitives learned from the new data are compared with the previous ones by measuring the inner product of the motion primitive vectors. Similar motion primitives and transitions are fused and novel motion primitives are added to capture newly observed behaviors. The proposed approach is tested and compared with other baselines in intersection scenarios where the data is incrementally available either from a single intersection or from multiple intersections with different geometries. In both cases, our method incrementally learns motion patterns and outperforms the offline learning approach in terms of prediction errors. The results also show that the model size in our algorithm grows at a much lower rate than standard incremental learning, where newly learned motion primitives and transitions are simply accumulated over time.

One of the most ambitious use cases of computer-assisted learning is to build a lifelong learning recommendation system. Unlike short-term courses, lifelong learning presents unique challenges, requiring sophisticated recommendation models that account for a wide range of factors such as background knowledge of learners or novelty of the material while effectively maintaining knowledge states of masses of learners for significantly longer periods of time (ideally, a lifetime). This work presents the foundations towards building a dynamic, scalable and transparent recommendation system for education, modelling learner's knowledge from implicit data in the form of engagement with open educational resources. We i) use a text ontology based on Wikipedia to automatically extract knowledge components of educational resources and, ii) propose a set of online Bayesian strategies inspired by the well-known areas of item response theory and knowledge tracing. Our proposal, TrueLearn, focuses on recommendations for which the learner has enough background knowledge (so they are able to understand and learn from the material), and the material has enough novelty that would help the learner improve their knowledge about the subject and keep them engaged. We further construct a large open educational video lectures dataset and test the performance of the proposed algorithms, which show clear promise towards building an effective educational recommendation system. Introduction One-on-one tutoring has shown learning gains of the order of two standard deviations (Corbett 2001). Machine learning now promises to provide such benefits of high quality personalised teaching to anyone in the world in a cost effective manner (Piech et al. 2015). Meanwhile, Open Educational Resources (OERs), defined as teaching, learning and research material available in the public domain or published under an open license (UNESCO 2019), are growing at a very fast pace.

Reinforcement Learning in domains with sparse rewards is a difficult problem, and a large part of the training process is often spent searching the state space in a more or less random fashion for any learning signals. For control problems, we often have some controller readily available which might be suboptimal but nevertheless solves the problem to some degree. This controller can be used to guide the initial exploration phase of the learning controller towards reward yielding states, reducing the time before refinement of a viable policy can be initiated. In our work, the agent is guided through an auxiliary behaviour cloning loss which is made conditional on a Q-filter, i.e. it is only applied in situations where the critic deems the guiding controller to be better than the agent. The Q-filter provides a natural way to adjust the guidance throughout the training process, allowing the agent to exceed the guiding controller in a manner that is adaptive to the task at hand and the proficiency of the guiding controller. The contribution of this paper lies in identifying shortcomings in previously proposed implementations of the Q-filter concept, and in suggesting some ways these issues can be mitigated. These modifications are tested on the OpenAI Gym Fetch environments, showing clear improvements in adaptivity and yielding increased performance in all robotic environments tested.

Generalized planning aims at computing an algorithm-like structure (generalized plan) that solves a set of multiple planning instances. In this paper we define negative examples for generalized planning as planning instances that must not be solved by a generalized plan. With this regard the paper extends the notion of validation of a generalized plan as the problem of verifying that a given generalized plan solves the set of input positives instances while it fails to solve a given input set of negative examples. This notion of plan validation allows us to define quantitative metrics to asses the generalization capacity of generalized plans. The paper also shows how to incorporate this new notion of plan validation into a compilation for plan synthesis that takes both positive and negative instances as input. Experiments show that incorporating negative examples can accelerate plan synthesis in several domains and leverage quantitative metrics to evaluate the generalization capacity of the synthesized plans.

Unstructured enterprise data such as reports, manuals and guidelines often contain tables. The traditional way of integrating data from these tables is through a two-step process of table detection/extraction and mapping the table layouts to an appropriate schema. This can be an expensive process. In this paper we show that by using semantic technologies (RDF/SPARQL and database dependencies) paired with a simple but powerful way to transform tables with non-relational layouts, it is possible to offer query answering services over these tables with minimal manual work or domain-specific mappings. Our method enables users to exploit data in tables embedded in documents with little effort, not only for simple retrieval queries, but also for structured queries that require joining multiple interrelated tables.