Search

The ultimate goal of dialog research is to develop systems that can be effectively used in interactive settings by real users. To this end, we introduced the Interactive Evaluation of Dialog Track at the 9th Dialog System Technology Challenge. This track consisted of two sub-tasks. The first sub-task involved building knowledge-grounded response generation models. The second sub-task aimed to extend dialog models beyond static datasets by assessing them in an interactive setting with real users. Our track challenges participants to develop strong response generation models and explore strategies that extend them to back-and-forth interactions with real users. The progression from static corpora to interactive evaluation introduces unique challenges and facilitates a more thorough assessment of open-domain dialog systems. This paper provides an overview of the track, including the methodology and results. Furthermore, it provides insights into how to best evaluate open-domain dialog models, Comment: Presented at LREC 2022 and DSTC9 Workshop at AAAI 2021

This is a report on the NSF Future Directions Workshop on Automatic Evaluation of Dialog. The workshop explored the current state of the art along with its limitations and suggested promising directions for future work in this important and very rapidly changing area of research., Comment: Report from the NSF AED Workshop (http://dialrc.org/AED/)

We present a fully automated workflow for phylogenetic reconstruction on large datasets, consisting of two novel methods, one for fast detection of cognates and one for fast Bayesian phylogenetic inference. Our results show that the methods take less than a few minutes to process language families that have so far required large amounts of time and computational power. Moreover, the cognates and the trees inferred from the method are quite close, both to gold standard cognate judgments and to expert language family trees. Given its speed and ease of application, our framework is specifically useful for the exploration of very large datasets in historical linguistics.

We describe a multi-phased Wizard-of-Oz approach to collecting human-robot dialogue in a collaborative search and navigation task. The data is being used to train an initial automated robot dialogue system to support collaborative exploration tasks. In the first phase, a wizard freely typed robot utterances to human participants. For the second phase, this data was used to design a GUI that includes buttons for the most common communications, and templates for communications with varying parameters. Comparison of the data gathered in these phases show that the GUI enabled a faster pace of dialogue while still maintaining high coverage of suitable responses, enabling more efficient targeted data collection, and improvements in natural language understanding using GUI-collected data. As a promising first step towards interactive learning, this work shows that our approach enables the collection of useful training data for navigation-based HRI tasks., Comment: Presented at AI-HRI AAAI-FSS, 2018 (arXiv:1809.06606)

We describe the adaptation and refinement of a graphical user interface designed to facilitate a Wizard-of-Oz (WoZ) approach to collecting human-robot dialogue data. The data collected will be used to develop a dialogue system for robot navigation. Building on an interface previously used in the development of dialogue systems for virtual agents and video playback, we add templates with open parameters which allow the wizard to quickly produce a wide variety of utterances. Our research demonstrates that this approach to data collection is viable as an intermediate step in developing a dialogue system for physical robots in remote locations from their users - a domain in which the human and robot need to regularly verify and update a shared understanding of the physical environment. We show that our WoZ interface and the fixed set of utterances and templates therein provide for a natural pace of dialogue with good coverage of the navigation domain., Comment: 7 pages, 2 figures, accepted for oral presentation at the Symposium on Natural Communication for Human-Robot Collaboration, AAAI Fall Symposium Series, November 9-11, 2017, https://www.aaai.org/ocs/index.php/FSS/FSS17

The Equilibrium Theory put forward by Argyle and Dean, posits that in human-human interactions, gaze and proxemic behaviors work together in establishing and maintaining a particular level of intimacy. This theory has been evaluated and used in Virtual Reality settings where people interact with Virtual Humans. In this study we disentangle the single and joint effects of proxemic and gaze behavior in this setting further, and examine how these behaviors affect the perceived personality of the agents. We simulate a social encounter with Virtual Humans in immersive Virtual Reality. Gaze and proxemic behaviors of the agents are manipulated dynamically while the participants’ gaze and proxemic responses are being measured. As could be expected, participants showed strongest gaze and proxemic responses when agents manipulated both at the same time. However, agents that only manipulated gaze elicited weaker responses compared to agents that only manipulated proxemics. Agents that exhibited more directed gaze and reduced interpersonal distance were attributed higher scores on intimacy related items than agents that exhibited averted gaze and increased interpersonal distance.

In natural communication, humans enrich their utterances with pragmatic information indicating, e.g., what is important to them or what they are not certain about. We investigate whether and how virtual humans (VH) can employ this kind of meta-communication. In an empirical study we have identified three modifying functions that humans produce and perceive in multimodal utterance, one being to create or attenuate focus. In this paper we test whether such modifying functions are also observed in speech and/or gesture of a VH, and whether this changes the perception of a VH overall. Results suggest that, although the VH’s behaviour is judged rather neutral overall, focusing is distinctively recognised, leads to better recall, and affects perceived competence. These effects are strongest if focus is created jointly by speech and gesture.

Recently, there has been an increasing interest in unsupervised parsers that optimize semantically oriented objectives, typically using reinforcement learning. Unfortunately, the learned trees often do not match actual syntax trees well. Shen et al. (2018) propose a structured attention mechanism for language modeling (PRPN), which induces better syntactic structures but relies on ad hoc heuristics. Also, their model lacks interpretability as it is not grounded in parsing actions. In our work, we propose an imitation learning approach to unsupervised parsing, where we transfer the syntactic knowledge induced by the PRPN to a Tree-LSTM model with discrete parsing actions. Its policy is then refined by Gumbel-Softmax training towards a semantically oriented objective. We evaluate our approach on the All Natural Language Inference dataset and show that it achieves a new state of the art in terms of parsing $F$-score, outperforming our base models, including the PRPN., Comment: ACL2019

State-of-the-art models of lexical semantic change detection suffer from noise stemming from vector space alignment. We have empirically tested the Temporal Referencing method for lexical semantic change and show that, by avoiding alignment, it is less affected by this noise. We show that, trained on a diachronic corpus, the skip-gram with negative sampling architecture with temporal referencing outperforms alignment models on a synthetic task as well as a manual testset. We introduce a principled way to simulate lexical semantic change and systematically control for possible biases., Comment: To appear in the 57th Annual Meeting of the Association for Computational Linguistics (ACL2019)

The cognitive mechanisms needed to account for the English past tense have long been a subject of debate in linguistics and cognitive science. Neural network models were proposed early on, but were shown to have clear flaws. Recently, however, Kirov and Cotterell (2018) showed that modern encoder-decoder (ED) models overcome many of these flaws. They also presented evidence that ED models demonstrate humanlike performance in a nonce-word task. Here, we look more closely at the behaviour of their model in this task. We find that (1) the model exhibits instability across multiple simulations in terms of its correlation with human data, and (2) even when results are aggregated across simulations (treating each simulation as an individual human participant), the fit to the human data is not strong---worse than an older rule-based model. These findings hold up through several alternative training regimes and evaluation measures. Although other neural architectures might do better, we conclude that there is still insufficient evidence to claim that neural nets are a good cognitive model for this task., Comment: Accepted at ACL 2019

Accurate entity linkers have been produced for domains and languages where annotated data (i.e., texts linked to a knowledge base) is available. However, little progress has been made for the settings where no or very limited amounts of labeled data are present (e.g., legal or most scientific domains). In this work, we show how we can learn to link mentions without having any labeled examples, only a knowledge base and a collection of unannotated texts from the corresponding domain. In order to achieve this, we frame the task as a multi-instance learning problem and rely on surface matching to create initial noisy labels. As the learning signal is weak and our surrogate labels are noisy, we introduce a noise detection component in our model: it lets the model detect and disregard examples which are likely to be noisy. Our method, jointly learning to detect noise and link entities, greatly outperforms the surface matching baseline. For a subset of entity categories, it even approaches the performance of supervised learning., Comment: ACL 2019

Building virtual humans is a task of formidable complexity. We believe that, especially when building agents that interact with biological humans in real-time over multiple sensorial channels, graphical, data flow oriented programming environments are the development tool of choice. In this paper, we describe a toolbox for the system control and block diagramming environment Simulink that supports the construction of virtual humans. Available blocks include sources for stochastic processes, utilities for coordinate transformation and messaging, as well as modules for controlling gaze and facial expressions.