Current deep learning models trained to generate radiology reports from chest radiographs are capable of producing clinically accurate, clear, and actionable text that can advance patient care. However, such systems all succumb to the same problem: making hallucinated references to non-existent prior reports. Such hallucinations occur because these models are trained on datasets of real-world patient reports that inherently refer to priors. To this end, we propose two methods to remove references to priors in radiology reports: (1) a GPT-3-based few-shot approach to rewrite medical reports without references to priors; and (2) a BioBERT-based token classification approach to directly remove words referring to priors. We use the aforementioned approaches to modify MIMIC-CXR, a publicly available dataset of chest X-rays and their associated free-text radiology reports; we then retrain CXR-RePaiR, a radiology report generation system, on the adapted MIMIC-CXR dataset. We find that our re-trained model--which we call CXR-ReDonE--outperforms previous report generation methods on clinical metrics, achieving an average BERTScore of 0.2351 (2.57% absolute improvement). We expect our approach to be broadly valuable in enabling current radiology report generation systems to be more directly integrated into clinical pipelines.
The widespread diffusion of medical and political claims in the wake of COVID-19 has led to a voluminous rise in misinformation and fake news. The current vogue is to employ manual fact-checkers to efficiently classify and verify such data to combat this avalanche of claim-ridden misinformation. However, the rate of information dissemination is such that it vastly outpaces the fact-checkers' strength. Therefore, to aid manual fact-checkers in eliminating the superfluous content, it becomes imperative to automatically identify and extract the snippets of claim-worthy (mis)information present in a post. In this work, we introduce the novel task of Claim Span Identification (CSI). We propose CURT, a large-scale Twitter corpus with token-level claim spans on more than 7.5k tweets. Furthermore, along with the standard token classification baselines, we benchmark our dataset with DABERTa, an adapter-based variation of RoBERTa. The experimental results attest that DABERTa outperforms the baseline systems across several evaluation metrics, improving by about 1.5 points. We also report detailed error analysis to validate the model's performance along with the ablation studies. Lastly, we release our comprehensive span annotation guidelines for public use.
Mislabeled examples are a common issue in real-world data, particularly for tasks like token classification where many labels must be chosen on a fine-grained basis. Here we consider the task of finding sentences that contain label errors in token classification datasets. We study 11 different straightforward methods that score tokens/sentences based on the predicted class probabilities output by a (any) token classification model (trained via any procedure). In precision-recall evaluations based on real-world label errors in entity recognition data from CoNLL-2003, we identify a simple and effective method that consistently detects those sentences containing label errors when applied with different token classification models.
E-commerce query understanding is the process of inferring the shopping intent of customers by extracting semantic meaning from their search queries. The recent progress of pre-trained masked language models (MLM) in natural language processing is extremely attractive for developing effective query understanding models. Specifically, MLM learns contextual text embedding via recovering the masked tokens in the sentences. Such a pre-training process relies on the sufficient contextual information. It is, however, less effective for search queries, which are usually short text. When applying masking to short search queries, most contextual information is lost and the intent of the search queries may be changed. To mitigate the above issues for MLM pre-training on search queries, we propose a novel pre-training task specifically designed for short text, called Extended Token Classification (ETC). Instead of masking the input text, our approach extends the input by inserting tokens via a generator network, and trains a discriminator to identify which tokens are inserted in the extended input. We conduct experiments in an E-commerce store to demonstrate the effectiveness of ETC.
Abbreviations are unavoidable yet critical parts of the medical text. Using abbreviations, especially in clinical patient notes, can save time and space, protect sensitive information, and help avoid repetitions. However, most abbreviations might have multiple senses, and the lack of a standardized mapping system makes disambiguating abbreviations a difficult and time-consuming task. The main objective of this study is to examine the feasibility of token classification methods for medical abbreviation disambiguation. Specifically, we explore the capability of token classification methods to deal with multiple unique abbreviations in a single text. We use two public datasets to compare and contrast the performance of several transformer models pre-trained on different scientific and medical corpora. Our proposed token classification approach outperforms the more commonly used text classification models for the abbreviation disambiguation task. In particular, the SciBERT model shows a strong performance for both token and text classification tasks over the two considered datasets. Furthermore, we find that abbreviation disambiguation performance for the text classification models becomes comparable to that of token classification only when postprocessing is applied to their predictions, which involves filtering possible labels for an abbreviation based on the training data.
Imbalanced classification problems are extremely common in natural language processing and are solved using a variety of resampling and filtering techniques, which often involve making decisions on how to select training data or decide which test examples should be labeled by the model. We examine the tradeoffs in model performance involved in choices of training sample and filter training and test data in heavily imbalanced token classification task and examine the relationship between the magnitude of these tradeoffs and the base rate of the phenomenon of interest. In experiments on sequence tagging to detect rare phenomena in English and Arabic texts, we find that different methods of selecting training data bring tradeoffs in effectiveness and efficiency. We also see that in highly imbalanced cases, filtering test data using first-pass retrieval models is as important for model performance as selecting training data. The base rate of a rare positive class has a clear effect on the magnitude of the changes in performance caused by the selection of training or test data. As the base rate increases, the differences brought about by those choices decreases.
Annotating text data for event information extraction systems is hard, expensive, and error-prone. We investigate the feasibility of integrating coarse-grained data (document or sentence labels), which is far more feasible to obtain, instead of annotating more documents. We utilize a multi-task model with two auxiliary tasks, document and sentence binary classification, in addition to the main task of token classification. We perform a series of experiments with varying data regimes for the aforementioned integration. Results show that while introducing extra coarse-grained data offers greater improvement and robustness, a gain is still possible with only the addition of negative documents that have no information on any event.
Building real-world complex Named Entity Recognition (NER) systems is a challenging task. This is due to the complexity and ambiguity of named entities that appear in various contexts such as short input sentences, emerging entities, and complex entities. Besides, real-world queries are mostly malformed, as they can be code-mixed or multilingual, among other scenarios. In this paper, we introduce our submitted system to the Multilingual Complex Named Entity Recognition (MultiCoNER) shared task. We approach the complex NER for multilingual and code-mixed queries, by relying on the contextualized representation provided by the multilingual Transformer XLM-RoBERTa. In addition to the CRF-based token classification layer, we incorporate a span classification loss to recognize named entities spans. Furthermore, we use a self-training mechanism to generate weakly-annotated data from a large unlabeled dataset. Our proposed system is ranked 6th and 8th in the multilingual and code-mixed MultiCoNER's tracks respectively.
The lack of label data is one of the significant bottlenecks for Chinese Spelling Check (CSC). Existing researches use the method of automatic generation by exploiting unlabeled data to expand the supervised corpus. However, there is a big gap between the real input scenario and automatic generated corpus. Thus, we develop a competitive general speller ECSpell which adopts the Error Consistent masking strategy to create data for pretraining. This error consistency masking strategy is used to specify the error types of automatically generated sentences which is consistent with real scene. The experimental result indicates our model outperforms previous state-of-the-art models on the general benchmark. Moreover, spellers often work within a particular domain in real life. Due to lots of uncommon domain terms, experiments on our built domain specific datasets show that general models perform terribly. Inspired by the common practice of input methods, we propose to add an alterable user dictionary to handle the zero-shot domain adaption problem. Specifically, we attach a User Dictionary guided inference module (UD) to a general token classification based speller. Our experiments demonstrate that ECSpell$^{UD}$, namely ECSpell combined with UD, surpasses all the other baselines largely, even approaching the performance on the general benchmark.
Toxicity is pervasive in social media and poses a major threat to the health of online communities. The recent introduction of pre-trained language models, which have achieved state-of-the-art results in many NLP tasks, has transformed the way in which we approach natural language processing. However, the inherent nature of pre-training means that they are unlikely to capture task-specific statistical information or learn domain-specific knowledge. Additionally, most implementations of these models typically do not employ conditional random fields, a method for simultaneous token classification. We show that these modifications can improve model performance on the Toxic Spans Detection task at SemEval-2021 to achieve a score within 4 percentage points of the top performing team.
Temporal expressions in text play a significant role in language understanding and correctly identifying them is fundamental to various retrieval and natural language processing systems. Previous works have slowly shifted from rule-based to neural architectures, capable of tagging expressions with higher accuracy. However, neural models can not yet distinguish between different expression types at the same level as their rule-based counterparts. In this work, we aim to identify the most suitable transformer architecture for joint temporal tagging and type classification, as well as, investigating the effect of semi-supervised training on the performance of these systems. Based on our study of token classification variants and encoder-decoder architectures, we present a transformer encoder-decoder model using the RoBERTa language model as our best performing system. By supplementing training resources with weakly labeled data from rule-based systems, our model surpasses previous works in temporal tagging and type classification, especially on rare classes. Our code and pre-trained experiments are available at: https://github.com/satya77/Transformer_Temporal_Tagger
In this work, we propose a novel problem formulation for de-identification of unstructured clinical text. We formulate the de-identification problem as a sequence to sequence learning problem instead of a token classification problem. Our approach is inspired by the recent state-of -the-art performance of sequence to sequence learning models for named entity recognition. Early experimentation of our proposed approach achieved 98.91% recall rate on i2b2 dataset. This performance is comparable to current state-of-the-art models for unstructured clinical text de-identification.
Cardiac ultrasound imaging is used to diagnose various heart diseases. Common analysis pipelines involve manual processing of the video frames by expert clinicians. This suffers from intra- and inter-observer variability. We propose a novel approach to ultrasound video analysis using a transformer architecture based on a Residual Auto-Encoder Network and a BERT model adapted for token classification. This enables videos of any length to be processed. We apply our model to the task of End-Systolic (ES) and End-Diastolic (ED) frame detection and the automated computation of the left ventricular ejection fraction. We achieve an average frame distance of 3.36 frames for the ES and 7.17 frames for the ED on videos of arbitrary length. Our end-to-end learnable approach can estimate the ejection fraction with a MAE of 5.95 and $R^2$ of 0.52 in 0.15s per video, showing that segmentation is not the only way to predict ejection fraction. Code and models are available at https://github.com/HReynaud/UVT.
Accurately extracting structured content from PDFs is a critical first step for NLP over scientific papers. Recent work has improved extraction accuracy by incorporating elementary layout information, e.g., each token's 2D position on the page, into language model pretraining. We introduce new methods that explicitly model VIsual LAyout (VILA) groups, i.e., text lines or text blocks, to further improve performance. In our I-VILA approach, we show that simply inserting special tokens denoting layout group boundaries into model inputs can lead to a 1.9% Macro F1 improvement in token classification. In the H-VILA approach, we show that hierarchical encoding of layout-groups can result in up-to 47% inference time reduction with less than 0.8% Macro F1 loss. Unlike prior layout-aware approaches, our methods do not require expensive additional pretraining, only fine-tuning, which we show can reduce training cost by up to 95%. Experiments are conducted on a newly curated evaluation suite, S2-VLUE, that unifies existing automatically-labeled datasets and includes a new dataset of manual annotations covering diverse papers from 19 scientific disciplines. Pre-trained weights, benchmark datasets, and source code are available at https://github.com/allenai/VILA.
Toxicity detection of text has been a popular NLP task in the recent years. In SemEval-2021 Task-5 Toxic Spans Detection, the focus is on detecting toxic spans within passages. Most state-of-the-art span detection approaches employ various techniques, each of which can be broadly classified into Token Classification or Span Prediction approaches. In our paper, we explore simple versions of both of these approaches and their performance on the task. Specifically, we use BERT-based models -- BERT, RoBERTa, and SpanBERT for both approaches. We also combine these approaches and modify them to bring improvements for Toxic Spans prediction. To this end, we investigate results on four hybrid approaches -- Multi-Span, Span+Token, LSTM-CRF, and a combination of predicted offsets using union/intersection. Additionally, we perform a thorough ablative analysis and analyze our observed results. Our best submission -- a combination of SpanBERT Span Predictor and RoBERTa Token Classifier predictions -- achieves an F1 score of 0.6753 on the test set. Our best post-eval F1 score is 0.6895 on intersection of predicted offsets from top-3 RoBERTa Token Classification checkpoints. These approaches improve the performance by 3% on average than those of the shared baseline models -- RNNSL and SpaCy NER.
Transformer-based language models, more specifically BERT-based architectures have achieved state-of-the-art performance in many downstream tasks. However, for a relatively low-resource language such as Thai, the choices of models are limited to training a BERT-based model based on a much smaller dataset or finetuning multi-lingual models, both of which yield suboptimal downstream performance. Moreover, large-scale multi-lingual pretraining does not take into account language-specific features for Thai. To overcome these limitations, we pretrain a language model based on RoBERTa-base architecture on a large, deduplicated, cleaned training set (78GB in total size), curated from diverse domains of social media posts, news articles and other publicly available datasets. We apply text processing rules that are specific to Thai most importantly preserving spaces, which are important chunk and sentence boundaries in Thai before subword tokenization. We also experiment with word-level, syllable-level and SentencePiece tokenization with a smaller dataset to explore the effects on tokenization on downstream performance. Our model wangchanberta-base-att-spm-uncased trained on the 78.5GB dataset outperforms strong baselines (NBSVM, CRF and ULMFit) and multi-lingual models (XLMR and mBERT) on both sequence classification and token classification tasks in human-annotated, mono-lingual contexts.
Natural language processing (NLP) techniques have become mainstream in the recent decade. Most of these advances are attributed to the processing of a single language. More recently, with the extensive growth of social media platforms focus has shifted to code-mixed text. The code-mixed text comprises text written in more than one language. People naturally tend to combine local language with global languages like English. To process such texts, current NLP techniques are not sufficient. As a first step, the text is processed to identify the language of the words in the text. In this work, we focus on language identification in code-mixed sentences for Hindi-English mixed text. The task of language identification is formulated as a token classification task. In the supervised setting, each word in the sentence has an associated language label. We evaluate different deep learning models and input representation combinations for this task. Mainly, character, sub-word, and word embeddings are considered in combination with CNN and LSTM based models. We show that sub-word representation along with the LSTM model gives the best results. In general sub-word representations perform significantly better than other input representations. We report the best accuracy of 94.52% using a single layer LSTM model on the standard SAIL ICON 2017 test set.
Language models based on the Transformer architecture have achieved state-of-the-art performance on a wide range of NLP tasks such as text classification, question-answering, and token classification. However, this performance is usually tested and reported on high-resource languages, like English, French, Spanish, and German. Indian languages, on the other hand, are underrepresented in such benchmarks. Despite some Indian languages being included in training multilingual Transformer models, they have not been the primary focus of such work. In order to evaluate the performance on Indian languages specifically, we analyze these language models through extensive experiments on multiple downstream tasks in Hindi, Bengali, and Telugu language. Here, we compare the efficacy of fine-tuning model parameters of pre-trained models against that of training a language model from scratch. Moreover, we empirically argue against the strict dependency between the dataset size and model performance, but rather encourage task-specific model and method selection. We achieve state-of-the-art performance on Hindi and Bengali languages for text classification task. Finally, we present effective strategies for handling the modeling of Indian languages and we release our model checkpoints for the community : https://huggingface.co/neuralspace-reverie.
State-of-the-art Neural Machine Translation (NMT) models struggle with generating low-frequency tokens, tackling which remains a major challenge. The analysis of long-tailed phenomena in the context of structured prediction tasks is further hindered by the added complexities of search during inference. In this work, we quantitatively characterize such long-tailed phenomena at two levels of abstraction, namely, token classification and sequence generation. We propose a new loss function, the Anti-Focal loss, to better adapt model training to the structural dependencies of conditional text generation by incorporating the inductive biases of beam search in the training process. We show the efficacy of the proposed technique on a number of Machine Translation (MT) datasets, demonstrating that it leads to significant gains over cross-entropy across different language pairs, especially on the generation of low-frequency words. We have released the code to reproduce our results.
We can consider Counterfactuals as belonging in the domain of Discourse structure and semantics, A core area in Natural Language Understanding and in this paper, we introduce an approach to resolving counterfactual detection as well as the indexing of the antecedents and consequents of Counterfactual statements. While Transfer learning is already being applied to several NLP tasks, It has the characteristics to excel in a novel number of tasks. We show that detecting Counterfactuals is a straightforward Binary Classification Task that can be implemented with minimal adaptation on already existing model Architectures, thanks to a well annotated training data set,and we introduce a new end to end pipeline to process antecedents and consequents as an entity recognition task, thus adapting them into Token Classification.