While classic NLG systems typically made use of hierarchically structured content plans that included discourse relations as central components, more recent neural approaches have mostly mapped simple, flat inputs to texts without representing discourse relations explicitly.In this paper, we investigate whether it is beneficial to include discourse relations in the input to neural data-to-text generators for texts where discourse relations play an important role.To do so, we reimplement the sentence planning and realization components of a classic NLG system, Methodius, using LSTM sequence-to-sequence (seq2seq) models.We find that although seq2seq models can learn to generate fluent and grammatical texts remarkably well with sufficiently representative Methodius training data, they cannot learn to correctly express Methodius's SIMILARITY and CONTRAST comparisons unless the corresponding RST relations are included in the inputs.Additionally, we experiment with using self-training and reverse model reranking to better handle train/test data mismatches, and find that while these methods help reduce content errors, it remains essential to include discourse relations in the input to obtain optimal performance.

Computational systems and objects are becoming increasingly closely integrated with our daily activities. Ubiquitous and pervasive computing first identified the emerging challenges of studying technology used on-the-move and in widely varied contexts. With IoT, previously sporadic experiences are interconnected across time and space in numerous and complex ways. This increasing complexity has multiplied the challenges facing those who study human experience to inform design. This paper describes the results of a study that used a chatbot or 'Ethnobot' to gather ethnographic data, and considers the opportunities and challenges in collecting this data in the absence of a human ethnographer. This study involved 13 participants gathering information about their experiences at the Royal Highland Show. We demonstrate the effectiveness of the Ethnobot in this setting, discuss the benefits and drawbacks of chatbots as a tool for ethnographic data collection, and conclude with recommendations for the design of chatbots for this purpose.

The Learning Analytics Report Card (LARC) is a pilot system which takes time-series data from a student's course-related activity in a Virtual Learning Environment and generates automatic textual summaries in real time.Students are able to generate reports as often as they like, and to choose which aspects of their behaviour are included in each report.As well as rating a student's scores against set standards, the generated texts make comparisons with the individual student's previous behaviour from the same course, and with the average scores of their student cohort.In addition, we carry out sentiment analysis on the student's forum posts, and generate a summary using quantifiers.We report some student reactions to initial trials of the system.

In this paper we present results from a user evaluation of a robot bartender system which handles state uncertainty derived from speech input by using belief tracking and generating appropriate clarification questions. We present a combination of state estimation and action selection components in which state uncertainty is tracked and exploited, and compare it to a baseline version that uses standard speech recognition confidence score thresholds instead of belief tracking. The results suggest that users are served fewer incorrect drinks when the uncertainty is retained in the state.

When humans speak they often use gram-matically incorrect sentences, which is a problem for grammar-based language pro-cessing methods, since they expect in-put that is valid for the grammar. We present two methods to transform spoken language into grammatically correct sen-tences. The first is an algorithm for au-tomatic ellipsis detection, which finds el-lipses in spoken sentences and searches in a combinatory categorial grammar for suitable words to fill the ellipses. The sec-ond method is an algorithm that computes the semantic similarity of two words us-ing WordNet, which we use to find alter-natives to words that are unknown to the grammar. In an evaluation, we show that the usage of these two methods leads to an increase of 38.64 % more parseable sen-tences on a test set of spoken sentences that were collected during a human-robot interaction experiment. 1

We address the question of whether service robots that interact with humans in public spaces must express socially appropriate behaviour. To do so, we implemented a robot bartender which is able to take drink orders from humans and serve drinks to them. By using a high-level automated planner, we explore two different robot interaction styles: in the task only setting, the robot simply fulfils its goal of asking customers for drink orders and serving them drinks; in the socially intelligent setting, the robot additionally acts in a manner socially appropriate to the bartender scenario, based on the behaviour of humans observed in natural bar interactions. The results of a user study show that the interactions with the socially intelligent robot were somewhat more efficient, but the two implemented behaviour settings had only a small influence on the subjective ratings. However, there were objective factors that influenced participant ratings: the overall duration of the interaction had a positive influence on the ratings, while the number of system order requests had a negative influence. We also found a cultural difference: German participants gave the system higher pre-test ratings than participants who interacted in English, although the post-test scores were similar.

We introduce a humanoid robot bartender that is capable of dealing with multiple customers in a dynamic, multi-party social setting. The robot system incorporates state-of-the-art components for computer vision, linguistic processing, state management, high-level reasoning, and robot control. In a user evaluation, 31 participants interacted with the bartender in a range of social situations. Most customers successfully obtained a drink from the bartender in all scenarios, and the factors that had the greatest impact on subjective satisfaction were task success and dialogue efficiency.

We describe our work in building a spoken language interface for a tutorial dialogue system. Our goal is to allow natural, unrestricted student interaction with the computer tutor, which has been shown to improve the student’s learning gain, but presents challenges for speech recognition and spoken language understanding. We discuss the choice of system components and present the results of development experiments in both acoustic and language modelling for speech recognition in this domain. Index Terms: spoken dialogue system, speech recognition, computer tutoring, adaptation