Incorporating knowledge bases (KB) into end-to-end task-oriented dialogue systems is challenging, since it requires to properly represent the entity of KB, which is associated with its KB context and dialogue context. The existing works represent the
entity with only perceiving a part of its KB context, which can lead to the less effective representation due to the information loss, and adversely favor KB reasoning and response generation. To tackle this issue, we explore to fully contextualize the entity representation by dynamically perceiving all the relevant entities and dialogue history. To achieve this, we propose a COntext-aware Memory Enhanced Transformer framework (COMET), which treats the KB as a sequence and leverages a novel Memory Mask to enforce the entity to only focus on its relevant entities and dialogue history, while avoiding the distraction from the irrelevant entities. Through extensive experiments, we show that our COMET framework can achieve superior performance over the state of the arts.
Recent development in NLP shows a strong trend towards refining pre-trained models with a domain-specific dataset. This is especially the case for response generation where emotion plays an important role. However, existing empathetic datasets remain
small, delaying research efforts in this area, for example, the development of emotion-aware chatbots. One main technical challenge has been the cost of manually annotating dialogues with the right emotion labels. In this paper, we describe a large-scale silver dataset consisting of 1M dialogues annotated with 32 fine-grained emotions, eight empathetic response intents, and the Neutral category. To achieve this goal, we have developed a novel data curation pipeline starting with a small seed of manually annotated data and eventually scaling it to a satisfactory size. We compare its quality against a state-of-the-art gold dataset using both offline experiments and visual validation methods. The resultant procedure can be used to create similar datasets in the same domain as well as in other domains.
Document-level relation extraction aims to identify relations between entities in a whole document. Prior efforts to capture long-range dependencies have relied heavily on implicitly powerful representations learned through (graph) neural networks, w
hich makes the model less transparent. To tackle this challenge, in this paper, we propose LogiRE, a novel probabilistic model for document-level relation extraction by learning logic rules. LogiRE treats logic rules as latent variables and consists of two modules: a rule generator and a relation extractor. The rule generator is to generate logic rules potentially contributing to final predictions, and the relation extractor outputs final predictions based on the generated logic rules. Those two modules can be efficiently optimized with the expectation-maximization (EM) algorithm. By introducing logic rules into neural networks, LogiRE can explicitly capture long-range dependencies as well as enjoy better interpretation. Empirical results show that significantly outperforms several strong baselines in terms of relation performance and logical consistency. Our code is available at https://github.com/rudongyu/LogiRE.
Codifying commonsense knowledge in machines is a longstanding goal of artificial intelligence. Recently, much progress toward this goal has been made with automatic knowledge base (KB) construction techniques. However, such techniques focus primarily
on the acquisition of positive (true) KB statements, even though negative (false) statements are often also important for discriminative reasoning over commonsense KBs. As a first step toward the latter, this paper proposes NegatER, a framework that ranks potential negatives in commonsense KBs using a contextual language model (LM). Importantly, as most KBs do not contain negatives, NegatER relies only on the positive knowledge in the LM and does not require ground-truth negative examples. Experiments demonstrate that, compared to multiple contrastive data augmentation approaches, NegatER yields negatives that are more grammatical, coherent, and informative---leading to statistically significant accuracy improvements in a challenging KB completion task and confirming that the positive knowledge in LMs can be re-purposed'' to generate negative knowledge.
We tackle the problem of self-training networks for NLU in low-resource environment---few labeled data and lots of unlabeled data. The effectiveness of self-training is a result of increasing the amount of training data while training. Yet it becomes
less effective in low-resource settings due to unreliable labels predicted by the teacher model on unlabeled data. Rules of grammar, which describe the grammatical structure of data, have been used in NLU for better explainability. We propose to use rules of grammar in self-training as a more reliable pseudo-labeling mechanism, especially when there are few labeled data. We design an effective algorithm that constructs and expands rules of grammar without human involvement. Then we integrate the constructed rules as a pseudo-labeling mechanism into self-training. There are two possible scenarios regarding data distribution: it is unknown or known in prior to training. We empirically demonstrate that our approach substantially outperforms the state-of-the-art methods in three benchmark datasets for both scenarios.
Aspect-based sentiment analysis (ABSA) typically focuses on extracting aspects and predicting their sentiments on individual sentences such as customer reviews. Recently, another kind of opinion sharing platform, namely question answering (QA) forum,
has received increasing popularity, which accumulates a large number of user opinions towards various aspects. This motivates us to investigate the task of ABSA on QA forums (ABSA-QA), aiming to jointly detect the discussed aspects and their sentiment polarities for a given QA pair. Unlike review sentences, a QA pair is composed of two parallel sentences, which requires interaction modeling to align the aspect mentioned in the question and the associated opinion clues in the answer. To this end, we propose a model with a specific design of cross-sentence aspect-opinion interaction modeling to address this task. The proposed method is evaluated on three real-world datasets and the results show that our model outperforms several strong baselines adopted from related state-of-the-art models.
Modern approaches to Constituency Parsing are mono-lingual supervised approaches which require large amount of labelled data to be trained on, thus limiting their utility to only a handful of high-resource languages. To address this issue of data-spa
rsity for low-resource languages we propose Universal Recurrent Neural Network Grammars (UniRNNG) which is a multi-lingual variant of the popular Recurrent Neural Network Grammars (RNNG) model for constituency parsing. UniRNNG involves Cross-lingual Transfer Learning for Constituency Parsing task. The architecture of UniRNNG is inspired by Principle and Parameter theory proposed by Noam Chomsky. UniRNNG utilises the linguistic typology knowledge available as feature-values within WALS database, to generalize over multiple languages. Once trained on sufficiently diverse polyglot corpus UniRNNG can be applied to any natural language thus making it Language-agnostic constituency parser. Experiments reveal that our proposed UniRNNG outperform state-of-the-art baseline approaches for most of the target languages, for which these are tested.
We introduce a method for generating error-correction rules for grammar pattern errors in a given annotated learner corpus. In our approach, annotated edits in the learner corpus are converted into edit rules for correcting common writing errors. The
method involves automatic extraction of grammar patterns, and automatic alignment of the erroneous patterns and correct patterns. At run-time, grammar patterns are extracted from the grammatically correct sentences, and correction rules are retrieved by aligning the extracted grammar patterns with the erroneous patterns. Using the proposed method, we generate 1,499 high-quality correction rules related to 232 headwords. The method can be used to assist ESL students in avoiding grammatical errors, and aid teachers in correcting students' essays. Additionally, the method can be used in the compilation of collocation error dictionaries and the construction of grammar error correction systems.
Linguistic typology is an area of linguistics concerned with analysis of and comparison between natural languages of the world based on their certain linguistic features. For that purpose, historically, the area has relied on manual extraction of lin
guistic feature values from textural descriptions of languages. This makes it a laborious and time expensive task and is also bound by human brain capacity. In this study, we present a deep learning system for the task of automatic extraction of linguistic features from textual descriptions of natural languages. First, textual descriptions are manually annotated with special structures called semantic frames. Those annotations are learned by a recurrent neural network, which is then used to annotate un-annotated text. Finally, the annotations are converted to linguistic feature values using a separate rule based module. Word embeddings, learned from general purpose text, are used as a major source of knowledge by the recurrent neural network. We compare the proposed deep learning system to a previously reported machine learning based system for the same task, and the deep learning system wins in terms of F1 scores with a fair margin. Such a system is expected to be a useful contribution for the automatic curation of typological databases, which otherwise are manually developed.
This paper explains the design of a heterogeneous system that ranked eighth in competition in SemEval2021 Task 8. We analyze ablation experiments and demonstrate how the system components, namely tokenizer, unit identifier, modifier classifier, and l
anguage model, affect the overall score. We compare our results to similar experiments from the literature and introduce a grouping algorithm developed in the post-evaluation phase that increased our system's overall score, hypothetically elevating our competition rank from eight to six.
