Canada

In the task of entity description generation, given a context and a specified entity, a model must describe that entity correctly and in a contextually-relevant way. In this task, as well as broader language generation tasks, the generation of a nonfactual description (factual error) versus an incongruous description (contextual error) is fundamentally different, yet often conflated. We develop an evaluation paradigm that enables us to disentangle these two types of errors in naturally occurring textual contexts. We find that factuality and congruity are often at odds, and that models specifically struggle with accurate descriptions of entities that are less familiar to people. This shortcoming of language models raises concerns around the trustworthiness of such models, since factual errors on less well-known entities are exactly those that a human reader will not recognize.

&lt;iframe src=&quot;https://app.sli.do/event/or9aypsi961h595NQtD7kk/embed/polls/0b93f68a-b7a3-4a84-976f-20bcf8ceffe8&quot; width=&quot;300&quot; height=&quot;400&quot;&gt;&lt;/iframe&gt;

ACL 2023

Factual or Contextual? Disentangling Error Types in Entity Description Generation

factual and contextual errors

referring expression generation

model analysis

### Welcome to ACL 2023, the 61st Annual Meeting of the Association for Computational Linguistics! 
<br>The conference will be held in Toronto, Canada, July 9-14, 2023.<br><br>
Following the succession of the recent conferences in our field, ACL 2023 will adopt a hybrid format.
While the impact of Covid has considerably diminished in terms of traveling, obtaining visas to Canada
entails a very long process. Moreover, the global economic conditions pose challenges for many individuals to travel to conferences. Recognizing these circumstances, we know many participants may not be
able to attend the conference in person. Therefore, we are committed to providing a great virtual platform
so everyone has the opportunity to interact with other participants and enjoy the conference. Based on the
current registered participants, approxiately 30% have chosen to attend the conference virtually. Whether
you join us in person or virtually, we sincerely hope everyone has a remarkable conference experience.<br>
This General Chair’s message is where I express my gratitude to the many individuals who have made
enormous contributions to the conference over the past year.

Read [**ACL 2023 General Chair's message**](https://docs.google.com/document/d/1WobYM7norbG4dI48s75HfJoD89qgX5a_F-6U8AteLSA/edit?usp=sharing/) in full.

##### **[Conference Handbook](https://2023.aclweb.org/downloads/acl2023-handbook.pdf)**
<br><br><br><br>

The Association for Computational Linguistics (ACL) is the premier international scientific and professional society for people working on computational problems involving human language, a field often referred to as either computational linguistics or natural language processing.

In the task of entity description generation, given a context and a specified entity, a model must describe that entity correctly and in a contextually-relevant way. In this task, as well as broader language generation tasks, the generation of a nonfactual description (factual error) versus an incongruous description (contextual error) is fundamentally different, yet often conflated. We develop an evaluation paradigm that enables us to disentangle these two types of errors in naturally occurring textual contexts. We find that factuality and congruity are often at odds, and that models specifically struggle with accurate descriptions of entities that are less familiar to people. This shortcoming of language models raises concerns around the trustworthiness of such models, since factual errors on less well-known entities are exactly those that a human reader will not recognize.

<iframe src="https://app.sli.do/event/or9aypsi961h595NQtD7kk/embed/polls/0b93f68a-b7a3-4a84-976f-20bcf8ceffe8" width="300" height="400"></iframe>

technical paper

Multi-vector retrieval methods combine the merits of sparse (e.g. BM25) and dense (e.g. DPR) retrievers and have achieved state-of-the-art performance on various retrieval tasks.These methods, however, are orders of magnitude slower and need much more space to store their indices compared to their single-vector counterparts.In this paper, we unify different multi-vector retrieval models from a token routing viewpoint and propose conditional token interaction via dynamic lexical routing, namely CITADEL, for efficient and effective multi-vector retrieval.CITADEL learns to route different token vectors to the predicted lexical keys such that a query token vector only interacts with document token vectors routed to the same key.This design significantly reduces the computation cost while maintaining high accuracy.Notably, CITADEL achieves the same or slightly better performance than the previous state of the art, ColBERT-v2, on both in-domain (MS MARCO) and out-of-domain (BEIR) evaluations, while being nearly 40 times faster. Source code and data are available at https://github.com/facebookresearch/dpr-scale/tree/citadel.

<iframe src="https://app.sli.do/event/oApzFnCrRo86e5qieE2QHQ/embed/polls/f2603a7c-9666-4a30-889a-9bd2d6278db3" width="300" height="400"></iframe>

CITADEL: Conditional Token Interaction via Dynamic Lexical Routing for Efficient and Effective Multi-Vector Retrieval

Audio-visual speech recognition (AVSR) attracts a surge of research interest recently by leveraging multimodal signals to understand human speech. Mainstream approaches addressing this task have developed sophisticated architectures and techniques for multi-modality fusion and representation learning. However, the natural heterogeneity of different modalities causes distribution gap between their representations, making it challenging to fuse them. In this paper, we aim to learn the shared representations across modalities to bridge their gap. Different from existing similar methods on other multimodal tasks like sentiment analysis, we focus on the temporal contextual dependencies considering the sequence-to-sequence task setting of AVSR. In particular, we propose an adversarial network to refine frame-level modality-invariant representations (MIR-GAN), which captures the commonality across modalities to ease the subsequent multimodal fusion process. Extensive experiments on public benchmarks LRS3 and LRS2 show that our approach outperforms the state-of-the-arts.

<iframe src="https://app.sli.do/event/jQes3hFZo3TYjGXEHm6w95/embed/polls/81c65359-e7d5-4915-8280-78c60a9d7152" width="300" height="400"></iframe>

MIR-GAN: Refining Frame-Level Modality-Invariant Representations with Adversarial Network for Audio-Visual Speech Recognition

Human-annotated labels and explanations are critical for training explainable NLP models. However, unlike human-annotated labels whose quality is easier to calibrate (e.g., with a majority vote), human-crafted free-form explanations can be quite subjective. Before blindly using them as ground truth to train ML models, a vital question needs to be asked: How do we evaluate a human-annotated explanation's quality? In this paper, we build on the view that the quality of a human-annotated explanation can be measured based on its helpfulness (or impairment) to the ML models' performance for the desired NLP tasks for which the annotations were collected. In comparison to the commonly used Simulatability score, we define a new metric that can take into consideration the helpfulness of an explanation for model performance at both fine-tuning and inference. With the help of a unified dataset format, we evaluated the proposed metric on five datasets (e.g., e-SNLI) against two model architectures (T5 and BART), and the results show that our proposed metric can objectively evaluate the quality of human-annotated explanations, while Simulatability falls short.

<iframe src="https://app.sli.do/event/532fnPQGAZvefKDC1fZsuL/embed/polls/891218f6-0270-4a94-82a2-c40792b4990d" width="300" height="400"></iframe>

Are Human Explanations Always Helpful? Towards Objective Evaluation of Human Natural Language Explanations

A common metric for evaluating Automatic Speech Recognition (ASR) is Word Error Rate (WER) which solely takes into account dis- crepancies at the word-level. Although use- ful, WER is not guaranteed to correlate well with human judgment or performance on down- stream tasks that use ASR. Meaningful assess- ment of ASR mistakes becomes even more im- portant in high-stake scenarios such as health- care. We propose 2 general measures to evaluate the severity of mistakes made by ASR systems, one based on sentiment analysis and another based on text embeddings. We evaluate these measures on simulated patient-doctor conversations using 5 ASR systems. Results show that these measures capture characteristics of ASR errors that WER does not. Furthermore, we train an ASR system incorporating severity and demonstrate the potential for using severity not only in the evaluation, but in the development of ASR. Advantages and limitations of this methodology are analyzed and discussed.

Evaluating and Improving Automatic Speech Recognition using Severity

Fact-checking of health-related claims has become necessary in this digital age, where any information posted online is easily available to everyone. The most effective way to verify such claims is by using evidences obtained from reliable sources of medical knowledge, such as PubMed. Recent advances in the field of NLP have helped automate such fact-checking tasks. In this work, we propose a domain-specific BERT-based model using a transfer learning approach for the task of predicting the veracity of claim-evidence pairs for the verification of health-related facts. We also improvise on a method to combine multiple evidences retrieved for a single claim, taking into consideration conflicting evidences as well. We also show how our model can be exploited when labelled data is available and how back-translation can be used to augment data when there is data scarcity.

Multiple Evidence Combination for Fact-Checking of Health-Related Information

As opposed to general English, many concepts in biomedical terminology have been designed in recent history by biomedical professionals with the goal of being precise and concise. This is often achieved by concatenating meaningful biomedical morphemes to create new semantic units. Nevertheless, most modern biomedical language models (LMs) are pre-trained using standard domain-specific tokenizers derived from large scale biomedical corpus statistics without explicitly leveraging the agglutinating nature of biomedical language. In this work, we first find that standard open-domain and biomedical tokenizers are largely unable to segment biomedical terms into meaningful components. Therefore, we hypothesize that using a tokenizer which segments biomedical terminology more accurately would enable biomedical LMs to improve their performance on downstream biomedical NLP tasks, especially ones which involve biomedical terms directly such as named entity recognition (NER) and entity linking. Surprisingly, we find that pre-training a biomedical LM using a more accurate biomedical tokenizer does not improve the entity representation quality of a language model as measured by several intrinsic and extrinsic measures such as masked language modeling prediction (MLM) accuracy as well as NER and entity linking performance. These quantitative findings, along with a case study which explores entity representation quality more directly, suggest that the biomedical pre-training process is quite robust to instances of sub-optimal tokenization.

Biomedical Language Models are Robust to Sub-optimal Tokenization

We systematically investigate lightweight strategies to adapt large language models (LLMs) for the task of radiology report summarization (RRS). Specifically, we focus on domain adaptation via pretraining (on natural language, biomedical text, or clinical text) and via discrete prompting or parameter-efficient fine-tuning. Our results consistently achieve best performance by maximally adapting to the task via pretraining on clinical text and fine-tuning on RRS examples. Importantly, this method fine-tunes a mere 0.32% of parameters throughout the model, in contrast to end-to-end fine-tuning (100% of parameters). Additionally, we study the effect of in-context examples and out-of-distribution (OOD) training before concluding with a radiologist reader study and qualitative analysis. Our findings highlight the importance of domain adaptation in RRS and provide valuable insights toward developing effective natural language processing solutions for clinical tasks.

Downloads

Next from ACL 2023

CITADEL: Conditional Token Interaction via Dynamic Lexical Routing for Efficient and Effective Multi-Vector Retrieval

Similar lecture

Understanding Factual Errors in Summarization: Errors, Summarizers, Datasets, Error Detectors