We introduce LAMBADA, a dataset to evaluate the capabilities of computational models for text understanding by means of a word prediction task. LAMBADA is a collection of narrative passages sharing the characteristic that human subjects are able to guess their last word if they are exposed to the whole passage, but not if they only see the last sentence preceding the target word. To succeed on LAM-BADA, computational models cannot simply rely on local context, but must be able to keep track of information in the broader discourse. We show that LAMBADA exemplifies a wide range of linguistic phenomena, and that none of several state-ofthe-art language models reaches accuracy above 1% on this novel benchmark. We thus propose LAMBADA as a challenging test set, meant to encourage the development of new models capable of genuine understanding of broad context in natural language text.

This paper presents the first unsupervised approach to lexical semantic change that makes use of contextualised word representations. We propose a novel method that exploits the BERT neural language model to obtain representations of word usages, clusters these representations into usage types, and measures change along time with three proposed metrics. We create a new evaluation dataset and show that the model representations and the detected semantic shifts are positively correlated with human judgements. Our extensive qualitative analysis demonstrates that our method captures a variety of synchronic and diachronic linguistic phenomena. We expect our work to inspire further research in this direction.

BackgroundThis article summarizes the 2012 European Renal Association—European Dialysis and Transplant Association Registry Annual Report (available at ) with a specific focus on older patients (defined as ≥65 years).MethodsData provided by 45 national or regional renal registries in 30 countries in Europe and bordering the Mediterranean Sea were used. Individual patient level data were received from 31 renal registries, whereas 14 renal registries contributed data in an aggregated form. The incidence, prevalence and survival probabilities of patients with end-stage renal disease (ESRD) receiving renal replacement therapy (RRT) and renal transplantation rates for 2012 are presented.ResultsIn 2012, the overall unadjusted incidence rate of patients with ESRD receiving RRT was 109.6 per million population (pmp) (n = 69 035), ranging from 219.9 pmp in Portugal to 24.2 pmp in Montenegro. The proportion of incident patients ≥75 years varied from 15 to 44% between countries. The overall unadjusted prevalence on 31 December 2012 was 716.7 pmp (n = 451 270), ranging from 1670.2 pmp in Portugal to 146.7 pmp in the Ukraine. The proportion of prevalent patients ≥75 years varied from 11 to 32% between countries. The overall renal transplantation rate in 2012 was 28.3 pmp (n = 15 673), with the highest rate seen in the Spanish region of Catalonia. The proportion of patients ≥65 years receiving a transplant ranged from 0 to 35%. Five-year adjusted survival for all RRT patients was 59.7% (95% confidence interval, CI: 59.3–60.0) which fell to 39.3% (95% CI: 38.7–39.9) in patients 65–74 years and 21.3% (95% CI: 20.8–21.9) in patients ≥75 years.

We propose a grounded dialogue state encoder which addresses a foundational issue on how to integrate visual grounding with dialogue system components. As a test-bed, we focus on the GuessWhat?! game, a two-player game where the goal is to identify an object in a complex visual scene by asking a sequence of yes/no questions. Our visually-grounded encoder leverages synergies between guessing and asking questions, as it is trained jointly using multitask learning. We further enrich our model via a cooperative learning regime. We show that the introduction of both the joint architecture and cooperative learning lead to accuracy improvements over the baseline system. We compare our approach to an alternative system which extends the baseline with reinforcement learning. Our in-depth analysis shows that the linguistic skills of the two models differ dramatically, despite approaching comparable performance levels. This points at the importance of analyzing the linguistic output of competing systems beyond numeric comparison solely based on task success.1

This paper introduces the PhotoBook dataset, a large-scale collection of visually-grounded, task-oriented dialogues in English designed to investigate shared dialogue history accumulating during conversation. Taking inspiration from seminal work on dialogue analysis, we propose a data-collection task formulated as a collaborative game prompting two online participants to refer to images utilising both their visual context as well as previously established referring expressions. We provide a detailed description of the task setup and a thorough analysis of the 2,500 dialogues collected. To further illustrate the novel features of the dataset, we propose a baseline model for reference resolution which uses a simple method to take into account shared information accumulated in a reference chain. Our results show that this information is particularly important to resolve later descriptions and underline the need to develop more sophisticated models of common ground in dialogue interaction. 1

We describe a process for automatically detecting decision-making sub-dialogues in transcripts of multi-party, human-human meetings. Extending our previous work on action item identification, we propose a structured approach that takes into account the different roles utterances play in the decisionmaking process. We show that this structured approach outperforms the accuracy achieved by existing decision detection systems based on flat annotations, while enabling the extraction of more fine-grained information that can be used for summarization and reporting.

In this article we use well-known machine learning methods to tackle a novel task, namely the classification of non-sentential utterances (NSUs) in dialogue. We introduce a fine-grained taxonomy of NSU classes based on corpus work, and then report on the results of several machine learning experiments. First, we present a pilot study focussed on one of the NSU classes in the taxonomy-bare wh-phrases or 'sluices'-, and explore the task of disambiguating between the different readings that sluices can convey. We then extend the approach to classify the full range of NSU classes, obtaining results of around an 87% weighted F-score. Thus our experiments show that, for the taxonomy adopted, the task of identifying the right NSU class can be successfully learned, and hence provide a very encouraging basis for the more general enterprise of fully processing NSUs.(1) a. A: Who wants Beethoven music? B: Richard and James.

Children learn the meaning of words by being exposed to perceptually rich situations (linguistic discourse, visual scenes, etc). Current computational learning models typically simulate these rich situations through impoverished symbolic approximations. In this work, we present a distributed word learning model that operates on child-directed speech paired with realistic visual scenes. The model integrates linguistic and extra-linguistic information (visual and social cues), handles referential uncertainty, and correctly learns to associate words with objects, even in cases of limited linguistic exposure.