Event Temporal Relation Extraction (ETRE) is paramount but challenging. Within a discourse, event pairs are situated at different distances or the so-called proximity bands. The temporal ordering communicated about event pairs where at more remote (i.e., ``long'') or less remote (i.e., ``short'') proximity bands are encoded differently. SOTA models have tended to perform well on events situated at either short or long proximity bands, but not both. Nonetheless, real-world, natural texts contain all types of temporal event-pairs. In this paper, we present MulCo: Distilling Multi-Scale Knowledge via Contrastive Learning, a knowledge co-distillation approach that shares knowledge across multiple event pair proximity bands to improve performance on all types of temporal datasets. Our experimental results show that MulCo successfully integrates linguistic cues pertaining to temporal reasoning across both short and long proximity bands and achieves new state-of-the-art results on several ETRE benchmark datasets.
Contextualized word embeddings can lead to state-of-the-art performances in natural language understanding. Recently, a pre-trained deep contextualized text encoder such as BERT has shown its potential in improving natural language tasks including abstractive summarization. Existing approaches in dialogue summarization focus on incorporating a large language model into summarization task trained on large-scale corpora consisting of news articles rather than dialogues of multiple speakers. In this paper, we introduce self-supervised methods to compensate shortcomings to train a dialogue summarization model. Our principle is to detect incoherent information flows using pretext dialogue text to enhance BERT's ability to contextualize the dialogue text representations. We build and fine-tune an abstractive dialogue summarization model on a shared encoder-decoder architecture using the enhanced BERT. We empirically evaluate our abstractive dialogue summarizer with the SAMSum corpus, a recently introduced dataset with abstractive dialogue summaries. All of our methods have contributed improvements to abstractive summary measured in ROUGE scores. Through an extensive ablation study, we also present a sensitivity analysis to critical model hyperparameters, probabilities of switching utterances and masking interlocutors.
Dialogue bots have been widely applied in customer service scenarios to provide timely and user-friendly experience. These bots must classify the appropriate domain of a dialogue, understand the intent of users, and generate proper responses. Existing dialogue pre-training models are designed only for several dialogue tasks and ignore weakly-supervised expert knowledge in customer service dialogues. In this paper, we propose a novel unified knowledge prompt pre-training framework, UFA (\textbf{U}nified Model \textbf{F}or \textbf{A}ll Tasks), for customer service dialogues. We formulate all the tasks of customer service dialogues as a unified text-to-text generation task and introduce a knowledge-driven prompt strategy to jointly learn from a mixture of distinct dialogue tasks. We pre-train UFA on a large-scale Chinese customer service corpus collected from practical scenarios and get significant improvements on both natural language understanding (NLU) and natural language generation (NLG) benchmarks.
Large Transformer models achieved the state-of-the-art status for Natural Language Understanding tasks and are increasingly becoming the baseline model architecture for modeling source code. Transformers are usually pre-trained on large unsupervised corpora, learning token representations and transformations relevant to modeling generally available text, and are then fine-tuned on a particular downstream task of interest. While fine-tuning is a tried-and-true method for adapting a model to a new domain -- for example, question-answering on a given topic -- generalization remains an on-going challenge. In this paper, we explore and evaluate transformer model fine-tuning for personalization. In the context of generating unit tests for Java methods, we evaluate learning to personalize to a specific software project using several personalization techniques. We consider three key approaches: (i) custom fine-tuning, which allows all the model parameters to be tuned; (ii) lightweight fine-tuning, which freezes most of the model's parameters, allowing tuning of the token embeddings and softmax layer only or the final layer alone; (iii) prefix tuning, which keeps model parameters frozen, but optimizes a small project-specific prefix vector. Each of these techniques offers a trade-off in total compute cost and predictive performance, which we evaluate by code and task-specific metrics, training time, and total computational operations. We compare these fine-tuning strategies for code generation and discuss the potential generalization and cost benefits of each in various deployment scenarios.
For companies with customer service, mapping intents inside their conversational data is crucial in building applications based on natural language understanding (NLU). Nevertheless, there is no established automated technique to gather the intents from noisy online chats or voice transcripts. Simple clustering approaches are not suited to intent-sparse dialogues. To solve this intent-landscape task, we propose an unsupervised pipeline that extracts the intents and the taxonomy of intents from real-world dialogues. Our pipeline mines intent-span candidates with an extractive Question-Answering Electra model and leverages sentence embeddings to apply a low-level density clustering followed by a top-level hierarchical clustering. Our results demonstrate the generalization ability of an ELECTRA large model fine-tuned on the SQuAD2 dataset to understand dialogues. With the right prompting question, this model achieves a rate of linguistic validation on intent spans beyond 85%. We furthermore reconstructed the intent schemes of five domains from the MultiDoGo dataset with an average recall of 94.3%.
Large language models (LLMs) have achieved state-of-the-art performance on a series of natural language understanding tasks. However, these LLMs might rely on dataset bias and artifacts as shortcuts for prediction. This has significantly affected their generalizability and adversarial robustness. In this paper, we provide a review of recent developments that address the shortcut learning and robustness challenge of LLMs. We first introduce the concepts of shortcut learning of language models. We then introduce methods to identify shortcut learning behavior in language models, characterize the reasons for shortcut learning, as well as introduce mitigation solutions. Finally, we discuss key research challenges and potential research directions in order to advance the field of LLMs.
A cornerstone in AI research has been the creation and adoption of standardized training and test datasets to earmark the progress of state-of-the-art models. A particularly successful example is the GLUE dataset for training and evaluating Natural Language Understanding (NLU) models for English. The large body of research around self-supervised BERT-based language models revolved around performance improvements on NLU tasks in GLUE. To evaluate language models in other languages, several language-specific GLUE datasets were created. The area of speech language understanding (SLU) has followed a similar trajectory. The success of large self-supervised models such as wav2vec2 enable creation of speech models with relatively easy to access unlabelled data. These models can then be evaluated on SLU tasks, such as the SUPERB benchmark. In this work, we extend this to Indic languages by releasing the IndicSUPERB benchmark. Specifically, we make the following three contributions. (i) We collect Kathbath containing 1,684 hours of labelled speech data across 12 Indian languages from 1,218 contributors located in 203 districts in India. (ii) Using Kathbath, we create benchmarks across 6 speech tasks: Automatic Speech Recognition, Speaker Verification, Speaker Identification (mono/multi), Language Identification, Query By Example, and Keyword Spotting for 12 languages. (iii) On the released benchmarks, we train and evaluate different self-supervised models alongside a commonly used baseline FBANK. We show that language-specific fine-tuned models are more accurate than baseline on most of the tasks, including a large gap of 76\% for the Language Identification task. However, for speaker identification, self-supervised models trained on large datasets demonstrate an advantage. We hope IndicSUPERB contributes to the progress of developing speech language understanding models for Indian languages.
Pre-trained Language Models (PLMs) have achieved remarkable performance gains across numerous downstream tasks in natural language understanding. Various Chinese PLMs have been successively proposed for learning better Chinese language representation. However, most current models use Chinese characters as inputs and are not able to encode semantic information contained in Chinese words. While recent pre-trained models incorporate both words and characters simultaneously, they usually suffer from deficient semantic interactions and fail to capture the semantic relation between words and characters. To address the above issues, we propose a simple yet effective PLM CLOWER, which adopts the Contrastive Learning Over Word and charactER representations. In particular, CLOWER implicitly encodes the coarse-grained information (i.e., words) into the fine-grained representations (i.e., characters) through contrastive learning on multi-grained information. CLOWER is of great value in realistic scenarios since it can be easily incorporated into any existing fine-grained based PLMs without modifying the production pipelines.Extensive experiments conducted on a range of downstream tasks demonstrate the superior performance of CLOWER over several state-of-the-art baselines.
Intent detection is a key part of any Natural Language Understanding (NLU) system of a conversational assistant. Detecting the correct intent is essential yet difficult for email conversations where multiple directives and intents are present. In such settings, conversation context can become a key disambiguating factor for detecting the user's request from the assistant. One prominent way of incorporating context is modeling past conversation history like task-oriented dialogue models. However, the nature of email conversations (long form) restricts direct usage of the latest advances in task-oriented dialogue models. So in this paper, we provide an effective transfer learning framework (EMToD) that allows the latest development in dialogue models to be adapted for long-form conversations. We show that the proposed EMToD framework improves intent detection performance over pre-trained language models by 45% and over pre-trained dialogue models by 30% for task-oriented email conversations. Additionally, the modular nature of the proposed framework allows plug-and-play for any future developments in both pre-trained language and task-oriented dialogue models.
Generalized text representations are the foundation of many natural language understanding tasks. To fully utilize the different corpus, it is inevitable that models need to understand the relevance among them. However, many methods ignore the relevance and adopt a single-channel model (a coarse paradigm) directly for all tasks, which lacks enough rationality and interpretation. In addition, some existing works learn downstream tasks by stitches skill block(a fine paradigm), which might cause irrationalresults due to its redundancy and noise. Inthis work, we first analyze the task correlation through three different perspectives, i.e., data property, manual design, and model-based relevance, based on which the similar tasks are grouped together. Then, we propose a hierarchical framework with a coarse-to-fine paradigm, with the bottom level shared to all the tasks, the mid-level divided to different groups, and the top-level assigned to each of the tasks. This allows our model to learn basic language properties from all tasks, boost performance on relevant tasks, and reduce the negative impact from irrelevant tasks. Our experiments on 13 benchmark datasets across five natural language understanding tasks demonstrate the superiority of our method.
Natural language understanding (NLU) is the task of semantic decoding of human languages by machines. NLU models rely heavily on large training data to ensure good performance. However, substantial languages and domains have very few data resources and domain experts. It is necessary to overcome the data scarcity challenge, when very few or even zero training samples are available. In this thesis, we focus on developing cross-lingual and cross-domain methods to tackle the low-resource issues. First, we propose to improve the model's cross-lingual ability by focusing on the task-related keywords, enhancing the model's robustness and regularizing the representations. We find that the representations for low-resource languages can be easily and greatly improved by focusing on just the keywords. Second, we present Order-Reduced Modeling methods for the cross-lingual adaptation, and find that modeling partial word orders instead of the whole sequence can improve the robustness of the model against word order differences between languages and task knowledge transfer to low-resource languages. Third, we propose to leverage different levels of domain-related corpora and additional masking of data in the pre-training for the cross-domain adaptation, and discover that more challenging pre-training can better address the domain discrepancy issue in the task knowledge transfer. Finally, we introduce a coarse-to-fine framework, Coach, and a cross-lingual and cross-domain parsing framework, X2Parser. Coach decomposes the representation learning process into a coarse-grained and a fine-grained feature learning, and X2Parser simplifies the hierarchical task structures into flattened ones. We observe that simplifying task structures makes the representation learning more effective for low-resource languages and domains.
Weak supervision has been applied to various Natural Language Understanding tasks in recent years. Due to technical challenges with scaling weak supervision to work on long-form documents, spanning up to hundreds of pages, applications in the document understanding space have been limited. At Lexion, we built a weak supervision-based system tailored for long-form (10-200 pages long) PDF documents. We use this platform for building dozens of language understanding models and have applied it successfully to various domains, from commercial agreements to corporate formation documents. In this paper, we demonstrate the effectiveness of supervised learning with weak supervision in a situation with limited time, workforce, and training data. We built 8 high quality machine learning models in the span of one week, with the help of a small team of just 3 annotators working with a dataset of under 300 documents. We share some details about our overall architecture, how we utilize weak supervision, and what results we are able to achieve. We also include the dataset for researchers who would like to experiment with alternate approaches or refine ours. Furthermore, we shed some light on the additional complexities that arise when working with poorly scanned long-form documents in PDF format, and some of the techniques that help us achieve state-of-the-art performance on such data.
Benchmark datasets play an important role in evaluating Natural Language Understanding (NLU) models. However, shortcuts -- unwanted biases in the benchmark datasets -- can damage the effectiveness of benchmark datasets in revealing models' real capabilities. Since shortcuts vary in coverage, productivity, and semantic meaning, it is challenging for NLU experts to systematically understand and avoid them when creating benchmark datasets. In this paper, we develop a visual analytics system, ShortcutLens, to help NLU experts explore shortcuts in NLU benchmark datasets. The system allows users to conduct multi-level exploration of shortcuts. Specifically, Statistics View helps users grasp the statistics such as coverage and productivity of shortcuts in the benchmark dataset. Template View employs hierarchical and interpretable templates to summarize different types of shortcuts. Instance View allows users to check the corresponding instances covered by the shortcuts. We conduct case studies and expert interviews to evaluate the effectiveness and usability of the system. The results demonstrate that ShortcutLens supports users in gaining a better understanding of benchmark dataset issues through shortcuts, inspiring them to create challenging and pertinent benchmark datasets.
Extracting spatial relations from texts is a fundamental task for natural language understanding and previous studies only regard it as a classification task, ignoring those spatial relations with null roles due to their poor information. To address the above issue, we first view spatial relation extraction as a generation task and propose a novel hybrid model HMCGR for this task. HMCGR contains a generation and a classification model, while the former can generate those null-role relations and the latter can extract those non-null-role relations to complement each other. Moreover, a reflexivity evaluation mechanism is applied to further improve the accuracy based on the reflexivity principle of spatial relation. Experimental results on SpaceEval show that HMCGR outperforms the SOTA baselines significantly.
Despite recent advances in natural language understanding and generation, and decades of research on the development of conversational bots, building automated agents that can carry on rich open-ended conversations with humans "in the wild" remains a formidable challenge. In this work we develop a real-time, open-ended dialogue system that uses reinforcement learning (RL) to power a bot's conversational skill at scale. Our work pairs the succinct embedding of the conversation state generated using SOTA (supervised) language models with RL techniques that are particularly suited to a dynamic action space that changes as the conversation progresses. Trained using crowd-sourced data, our novel system is able to substantially exceeds the (strong) baseline supervised model with respect to several metrics of interest in a live experiment with real users of the Google Assistant.
Transformer-based pretrained language models (LMs) are ubiquitous across natural language understanding, but cannot be applied to long sequences such as stories, scientific articles and long documents, due to their quadratic complexity. While a myriad of efficient transformer variants have been proposed, they are typically based on custom implementations that require expensive pretraining from scratch. In this work, we propose SLED: SLiding-Encoder and Decoder, a simple approach for processing long sequences that re-uses and leverages battle-tested short-text pretrained LMs. Specifically, we partition the input into overlapping chunks, encode each with a short-text LM encoder and use the pretrained decoder to fuse information across chunks (fusion-in-decoder). We illustrate through controlled experiments that SLED offers a viable strategy for long text understanding and evaluate our approach on SCROLLS, a benchmark with seven datasets across a wide range of language understanding tasks. We find that SLED is competitive with specialized models that are up to 50x larger and require a dedicated and expensive pretraining step.
Robotic task instructions often involve a referred object that the robot must locate (ground) within the environment. While task intent understanding is an essential part of natural language understanding, less effort is made to resolve ambiguity that may arise while grounding the task. Existing works use vision-based task grounding and ambiguity detection, suitable for a fixed view and a static robot. However, the problem magnifies for a mobile robot, where the ideal view is not known beforehand. Moreover, a single view may not be sufficient to locate all the object instances in the given area, which leads to inaccurate ambiguity detection. Human intervention is helpful only if the robot can convey the kind of ambiguity it is facing. In this article, we present DoRO (Disambiguation of Referred Object), a system that can help an embodied agent to disambiguate the referred object by raising a suitable query whenever required. Given an area where the intended object is, DoRO finds all the instances of the object by aggregating observations from multiple views while exploring & scanning the area. It then raises a suitable query using the information from the grounded object instances. Experiments conducted with the AI2Thor simulator show that DoRO not only detects the ambiguity more accurately but also raises verbose queries with more accurate information from the visual-language grounding.
Supervised deep learning-based approaches have been applied to task-oriented dialog and have proven to be effective for limited domain and language applications when a sufficient number of training examples are available. In practice, these approaches suffer from the drawbacks of domain-driven design and under-resourced languages. Domain and language models are supposed to grow and change as the problem space evolves. On one hand, research on transfer learning has demonstrated the cross-lingual ability of multilingual Transformers-based models to learn semantically rich representations. On the other, in addition to the above approaches, meta-learning have enabled the development of task and language learning algorithms capable of far generalization. Through this context, this article proposes to investigate the cross-lingual transferability of using synergistically few-shot learning with prototypical neural networks and multilingual Transformers-based models. Experiments in natural language understanding tasks on MultiATIS++ corpus shows that our approach substantially improves the observed transfer learning performances between the low and the high resource languages. More generally our approach confirms that the meaningful latent space learned in a given language can be can be generalized to unseen and under-resourced ones using meta-learning.
Modern language models leverage increasingly large numbers of parameters to achieve performance on natural language understanding tasks. Ensembling these models in specific configurations for downstream tasks show even further performance improvements. In this paper, we perform an analysis of bagging language models and compare single language models to bagged ensembles that are roughly equivalent in terms of final model size. We explore an array of model bagging configurations for natural language understanding tasks with final ensemble sizes ranging from 300M parameters to 1.5B parameters and determine that our ensembling methods are at best roughly equivalent to single LM baselines. We note other positive effects of bagging and pruning in specific scenarios according to findings in our experiments such as variance reduction and minor performance improvements.
Sparsely Mixture of Experts (MoE) has received great interest due to its promising scaling capability with affordable computational overhead. MoE converts dense layers into sparse experts, and utilizes a gated routing network to make experts conditionally activated. However, as the number of experts grows, MoE with outrageous parameters suffers from overfitting and sparse data allocation. Such problems are especially severe on tasks with limited data, thus hindering the progress for MoE models to improve performance by scaling up. In this work, we propose Mixture of Expert Clusters - a general approach to enable expert layers to learn more diverse and appropriate knowledge by imposing variance-based constraints on the routing stage. We further propose a cluster-level expert dropout strategy specifically designed for the expert cluster structure. Our experiments reveal that MoEC could improve performance on machine translation and natural language understanding tasks, and raise the performance upper bound for scaling up experts under limited data. We also verify that MoEC plays a positive role in mitigating overfitting and sparse data allocation.