Question answering (QA) is a fundamental means to facilitate assessment and training of narrative comprehension skills for both machines and young children, yet there is scarcity of high-quality QA datasets carefully designed to serve this purpose. In particular, existing datasets rarely distinguish fine-grained reading skills, such as the understanding of varying narrative elements. Drawing on the reading education research, we introduce FairytaleQA, a dataset focusing on narrative comprehension of kindergarten to eighth-grade students. Generated by educational experts based on an evidence-based theoretical framework, FairytaleQA consists of 10,580 explicit and implicit questions derived from 278 children-friendly stories, covering seven types of narrative elements or relations. Our dataset is valuable in two folds: First, we ran existing QA models on our dataset and confirmed that this annotation helps assess models’ fine-grained learning skills. Second, the dataset supports question generation (QG) task in the education domain. Through benchmarking with QG models, we show that the QG model trained on FairytaleQA is capable of asking high-quality and more diverse questions.

Generating educational questions of fairytales or storybooks is vital for improving children’s literacy ability. However, it is challenging to generate questions that capture the interesting aspects of a fairytale story with educational meaningfulness. In this paper, we propose a novel question generation method that first learns the question type distribution of an input story paragraph, and then summarizes salient events which can be used to generate high-cognitive-demand questions. To train the event-centric summarizer, we finetune a pre-trained transformer-based sequence-to-sequence model using silver samples composed by educational question-answer pairs. On a newly proposed educational question-answering dataset FairytaleQA, we show good performance of our method on both automatic and human evaluation metrics. Our work indicates the necessity of decomposing question type distribution learning and event-centric summary generation for educational question generation.

Automatic speech recognition (ASR) on low resource languages improves the access of linguistic minorities to technological advantages provided by artificial intelligence (AI). In this paper, we address the problem of data scarcity for the Hong Kong Cantonese language by creating a new Cantonese dataset. Our dataset, Multi-Domain Cantonese Corpus (MDCC), consists of 73.6 hours of clean read speech paired with transcripts, collected from Cantonese audiobooks from Hong Kong. It comprises philosophy, politics, education, culture, lifestyle and family domains, covering a wide range of topics. We also review all existing Cantonese datasets and analyze them according to their speech type, data source, total size and availability. We further conduct experiments with Fairseq S2T Transformer, a state-of-the-art ASR model, on the biggest existing dataset, Common Voice zh-HK, and our proposed MDCC, and the results show the effectiveness of our dataset. In addition, we create a powerful and robust Cantonese ASR model by applying multi-dataset learning on MDCC and Common Voice zh-HK.

With the rise of deep learning and intelligent vehicles, the smart assistant has become an essential in-car component to facilitate driving and provide extra functionalities. In-car smart assistants should be able to process general as well as car-related commands and perform corresponding actions, which eases driving and improves safety. However, there is a data scarcity issue for low resource languages, hindering the development of research and applications. In this paper, we introduce a new dataset, Cantonese In-car Audio-Visual Speech Recognition (CI-AVSR), for in-car command recognition in the Cantonese language with both video and audio data. It consists of 4,984 samples (8.3 hours) of 200 in-car commands recorded by 30 native Cantonese speakers. Furthermore, we augment our dataset using common in-car background noises to simulate real environments, producing a dataset 10 times larger than the collected one. We provide detailed statistics of both the clean and the augmented versions of our dataset. Moreover, we implement two multimodal baselines to demonstrate the validity of CI-AVSR. Experiment results show that leveraging the visual signal improves the overall performance of the model. Although our best model can achieve a considerable quality on the clean test set, the speech recognition quality on the noisy data is still inferior and remains an extremely challenging task for real in-car speech recognition systems. The dataset and code will be released at https://github.com/HLTCHKUST/CI-AVSR.

Code-switching is a speech phenomenon occurring when a speaker switches language during a conversation. Despite the spontaneous nature of code-switching in conversational spoken language, most existing works collect code-switching data from read speech instead of spontaneous speech. ASCEND (A Spontaneous Chinese-English Dataset) is a high-quality Mandarin Chinese-English code-switching corpus built on spontaneous multi-turn conversational dialogue sources collected in Hong Kong. We report ASCEND’s design and procedure for collecting the speech data, including annotations. ASCEND consists of 10.62 hours of clean speech, collected from 23 bilingual speakers of Chinese and English. Furthermore, we conduct baseline experiments using pre-trained wav2vec 2.0 models, achieving a best performance of 22.69% character error rate and 27.05% mixed error rate.

Text-based games can be used to develop task-oriented text agents for accomplishing tasks with high-level language instructions, which has potential applications in domains such as human-robot interaction. Given a text instruction, reinforcement learning is commonly used to train agents to complete the intended task owing to its convenience of learning policies automatically. However, because of the large space of combinatorial text actions, learning a policy network that generates an action word by word with reinforcement learning is challenging. Recent research works show that imitation learning provides an effective way of training a generation-based policy network. However, trained agents with imitation learning are hard to master a wide spectrum of task types or skills, and it is also difficult for them to generalize to new environments. In this paper, we propose a meta reinforcement learning based method to train text agents through learning-to-explore. In particular, the text agent first explores the environment to gather task-specific information and then adapts the execution policy for solving the task with this information. On the publicly available testbed ALFWorld, we conducted a comparison study with imitation learning and show the superiority of our method.

Physical common sense plays an essential role in the cognition abilities of robots for human-robot interaction. Machine learning methods have shown promising results on physical commonsense learning in natural language processing but still suffer from model generalization. In this paper, we formulate physical commonsense learning as a knowledge graph completion problem to better use the latent relationships among training samples. Compared with completing general knowledge graphs, completing a physical commonsense knowledge graph has three unique characteristics: training data are scarce, not all facts can be mined from existing texts, and the number of relationships is small. To deal with these problems, we first use a pre-training language model BERT to augment training data, and then employ constrained tucker factorization to model complex relationships by constraining types and adding negative relationships. We compare our method with existing state-of-the-art knowledge graph embedding methods and show its superior performance.

Text emotion distribution learning (EDL) aims to develop models that can predict the intensity values of a sentence across a set of emotion categories. Existing methods based on supervised learning require a large amount of well-labelled training data, which is difficult to obtain due to inconsistent perception of fine-grained emotion intensity. In this paper, we propose a meta-learning approach to learn text emotion distributions from a small sample. Specifically, we propose to learn low-rank sentence embeddings by tensor decomposition to capture their contextual semantic similarity, and use K-nearest neighbors (KNNs) of each sentence in the embedding space to generate sample clusters. We then train a meta-learner that can adapt to new data with only a few training samples on the clusters, and further fit the meta-learner on KNNs of a testing sample for EDL. In this way, we effectively augment the learning ability of a model on the small sample. To demonstrate the performance, we compare the proposed approach with state-of-the-art EDL methods on a widely used EDL dataset: SemEval 2007 Task 14 (Strapparava and Mihalcea, 2007). Results show the superiority of our method on small-sample emotion distribution learning.

We propose a novel tensor embedding method that can effectively extract lexical features for humor recognition. Specifically, we use word-word co-occurrence to encode the contextual content of documents, and then decompose the tensor to get corresponding vector representations. We show that this simple method can capture features of lexical humor effectively for continuous humor recognition. In particular, we achieve a distance of 0.887 on a global humor ranking task, comparable to the top performing systems from SemEval 2017 Task 6B (Potash et al., 2017) but without the need for any external training corpus. In addition, we further show that this approach is also beneficial for small sample humor recognition tasks through a semi-supervised label propagation procedure, which achieves about 0.7 accuracy on the 16000 One-Liners (Mihalcea and Strapparava, 2005) and Pun of the Day (Yang et al., 2015) humour classification datasets using only 10% of known labels.

This paper attempts to marry the interpretability of statistical machine learning approaches with the more robust models of joke structure and joke semantics capable of being learned by neural models. Specifically, we explore the use of semantic relatedness features based on word associations, rather than the more common Word2Vec similarity, on a binary humour identification task and identify several factors that make word associations a better fit for humour. We also explore the effects of using joke structure, in the form of humour anchors (Yang et al., 2015), for improving the performance of semantic features and show that, while an intriguing idea, humour anchors contain several pitfalls that can hurt performance.

This paper looks at the task of predicting word association strengths across three datasets; WordNet Evocation (Boyd-Graber et al., 2006), University of Southern Florida Free Association norms (Nelson et al., 2004), and Edinburgh Associative Thesaurus (Kiss et al., 1973). We achieve results of r=0.357 and p=0.379, r=0.344 and p=0.300, and r=0.292 and p=0.363, respectively. We find Word2Vec (Mikolov et al., 2013) and GloVe (Pennington et al., 2014) cosine similarities, as well as vector offsets, to be the highest performing features. Furthermore, we examine the usefulness of Gaussian embeddings (Vilnis and McCallum, 2014) for predicting word association strength, the first work to do so.

This paper explores the role of semantic relatedness features, such as word associations, in humour recognition. Specifically, we examine the task of inferring pairwise humour judgments in Twitter hashtag wars. We examine a variety of word association features derived from University of Southern Florida Free Association Norms (USF) and the Edinburgh Associative Thesaurus (EAT) and find that word association-based features outperform Word2Vec similarity, a popular semantic relatedness measure. Our system achieves an accuracy of 56.42% using a combination of unigram perplexity, bigram perplexity, EAT difference (tweet-avg), USF forward (max), EAT difference (word-avg), USF difference (word-avg), EAT forward (min), USF difference (tweet-max), and EAT backward (min).

This paper explores humour recognition for Twitter-based hashtag games. Given their popularity, frequency, and relatively formulaic nature, these games make a good target for computational humour research and can leverage Twitter likes and retweets as humour judgments. In this work, we use pair-wise relative humour judgments to examine several measures of semantic relatedness between setups and punchlines on a hashtag game corpus we collected and annotated. Results show that perplexity, Normalized Google Distance, and free-word association-based features are all useful in identifying “funnier” hashtag game responses. In fact, we provide empirical evidence that funnier punchlines tend to be more obscure, although more obscure punchlines are not necessarily rated funnier. Furthermore, the asymmetric nature of free-word association features allows us to see that while punchlines should be harder to predict given a setup, they should also be relatively easy to understand in context.