Canada

A key component of modern conversational systems is the Dialogue State Tracker (or DST), which models a user&#39;s goals and needs. Toward building more robust and reliable DSTs, we introduce a prompt-based learning approach to automatically generate effective adversarial examples to probe DST models. Two key characteristics of this approach are: (i) it only needs the output of the DST with no need for model parameters, and (ii) it can learn to generate natural language utterances that can target any DST. Through experiments over state-of-the-art DSTs, the proposed framework leads to the greatest reduction in accuracy and the best attack success rate while maintaining good fluency and a low perturbation ratio. We also show how much the generated adversarial examples can bolster a DST through adversarial training. These results indicate the strength of prompt-based attacks on DSTs and leave open avenues for continued refinement.

ACL 2023

PromptAttack: Probing Dialogue State Trackers with Adversarial Prompts

dialogue state trackers

prompt tuning

adversarial attacks

## Welcome!
"I am excited to welcome you to this year’s edition of the Conference on Empirical Methods in Natural Language Processing! Importantly, it marks the 30th edition of EMNLP. With over 8,000 submissions, more than 3,000 accepted papers, and thousands of attendees, we have come a long way from that first
workshop, which had 14 accepted papers. As the field looks ahead, Suzhou is the fitting location for celebrating this milestone: rooted in a long literary tradition, yet modern and forward-looking, and home to a large share of the NLP community."<br>

*Message from the General Chair, Dirk Hovy*





<br>

EMNLP 2025

Celebrate 30 Years of EMNLP! 
EMNLP 2025 will be held in Suzhou, China from November 5th to November 9th, 2025.

A key component of modern conversational systems is the Dialogue State Tracker (or DST), which models a user's goals and needs. Toward building more robust and reliable DSTs, we introduce a prompt-based learning approach to automatically generate effective adversarial examples to probe DST models. Two key characteristics of this approach are: (i) it only needs the output of the DST with no need for model parameters, and (ii) it can learn to generate natural language utterances that can target any DST. Through experiments over state-of-the-art DSTs, the proposed framework leads to the greatest reduction in accuracy and the best attack success rate while maintaining good fluency and a low perturbation ratio. We also show how much the generated adversarial examples can bolster a DST through adversarial training. These results indicate the strength of prompt-based attacks on DSTs and leave open avenues for continued refinement.

poster

### 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.

Pre-trained large language models (LMs) struggle to perform logical reasoning reliably despite advances in scale and compositionality. In this work, we tackle this challenge through the lens of symbolic programming. We propose DSR-LM, a Differentiable Symbolic Reasoning framework where pre-trained LMs govern the perception of factual knowledge, and a symbolic module performs deductive reasoning. In contrast to works that rely on hand-crafted logic rules, our differentiable symbolic reasoning framework efficiently learns weighted rules and applies semantic loss to further improve LMs. DSR-LM is scalable, interpretable, and allows easy integration of prior knowledge, thereby supporting extensive symbolic programming to robustly derive a logical conclusion. The results of our experiments suggest that DSR-LM improves the logical reasoning abilities of pre-trained language models, resulting in a significant increase in accuracy of over 20% on deductive reasoning benchmarks. Furthermore, DSR-LM outperforms a variety of competitive baselines when faced with systematic changes in sequence length.

<iframe src="https://app.sli.do/event/qQ87ZDnDYN6cfmCQWHFaok/embed/polls/07800ea9-b6a2-4c4d-ab27-e543d960e1ba" width="300" height="400"></iframe>

Improved Logical Reasoning of Language Models via Differentiable Symbolic Programming

NLP has achieved great progress in the past decade through the use of neural models and large labeled datasets. The dependence on abundant data prevents NLP models from being applied to low-resource or novel tasks where significant time, money, or expertise is required to label massive amounts of textual data. Recently, data augmentation  methods have been explored as a means of improving data efficiency in NLP. To date, there has been no systematic empirical overview of data augmentation for NLP in the limited labeled data setting, making it difficult to understand which methods work in which . In this paper, we provide an empirical survey of recent progress on data augmentation for NLP in the limited labeled data setting, summarizing the landscape of methods (including token-level augmentations, sentence-level augmentations, adversarial augmentations and hidden-space augmentations) and carrying out experiments on 11 datasets covering topics/news classification, inference tasks, paraphrasing tasks, and single-sentence tasks. Based on the results, we draw several conclusions to help practitioners choose appropriate augmentations in different and discuss the current challenges and future directions for limited data learning in NLP.

<iframe src="https://app.sli.do/event/uyG7EHDP5trs2LHubv2Gj8/embed/polls/d65f8690-43b4-42d9-aea1-34b29d934a8a" width="300" height="400"></iframe>

An Empirical Survey of Data Augmentation for Limited Data Learning in NLP

Models and metrics of linguistic code-switching (CS) have almost exclusively worked with word-level units. However, any two words are not equally likely CS points in bilingual speech. In addition, other-language single-word items and alternating-language multi-word items have distinct properties. Adapting these familiar metrics to the Intonation Unit (IU), we capture a shared tendency for CS to occur across rather than within prosodic boundaries. This constraint is distorted when single- and multi-word other-language items are merged. Individual differences according to language distribution and CS rates are independent, visualized in the number and breadth of language bands in transcripts of bilingual speech. These results are important to consider in future development of code-switched datasets for NLP tasks, as the IU token and exclusion/inclusion of single-word items highly impacts the CS represented in the input text.

Aligning Code-Switching Metrics with Bilingual Behavior

Neural text generation models have achieved remarkable success in carrying on short open-domain conversations. However, their performance degrades significantly in the long term, especially in their ability to ask coherent questions. A significant issue is the generation of redundant questions where the answer has already been provided by the user. We adapt and evaluate different methods, including negative training, decoding, and classification, to mitigate the redundancy problem. We also propose a simple yet effective method for generating training data without the need for crowdsourcing human-human or human-bot conversations. Experiments with the BlenderBot model show that our combined method significantly reduces the rate of redundant questions from 27.2% to 8.7%, while improving the quality of the original model. The code, dataset, and trained models can be found at our repository.

I already said that! Degenerating redundant questions in open-domain dialogue systems.

We propose a weak supervision pipeline SWEET for extracting person names, which combines simple antirules with multiple large language models fine-tuned on public NER benchmark datasets as well as a domain-specific dataset labelled with ChatGPT. The proposed training process addresses the major challenge of lack of training data in this domain, by effectively aggregating multiple weak signals. The proposed method SWEET outperforms the previous supervised state-of-the-art method for this task by 10% F1 score and better generalizes to the benchmark datasets.

SWEET: Weakly Supervised Person Name Extraction for Fighting Human Trafficking

Recent advances in deep learning, and especially the invention of encoder-decoder architectures, has significantly improved the performance of abstractive summarization systems. While the majority of research has focused on written documents, we have observed an increasing interest in the summarization of dialogues and multi-party conversation over the past few years. A system that could reliably transform the audio or transcript of a human conversation into an abridged version that homes in on the most important points of the discussion would be valuable in a wide variety of real-world contexts, from business meetings to medical consultations to customer service calls. This paper focuses on abstractive summarization for multi-party meetings, providing a  survey of the challenges, datasets and systems relevant to this task and a discussion of promising directions for future study.

Abstractive Meeting Summarization: A Survey

Localizing a semantic parser to support new languages requires effective cross-lingual generalization.  Recent work has found success with machine-translation or zero-shot methods although these approaches can struggle to model how native speakers ask questions. We consider how to effectively leverage minimal annotated examples in new languages for few-shot cross-lingual semantic parsing. We introduce a first-order meta-learning algorithm to train a semantic parser with maximal sample efficiency during cross-lingual transfer. Our algorithm uses high-resource languages to train the parser and simultaneously optimizes for cross-lingual generalization for lower-resource languages. Results across six languages on ATIS demonstrate that our combination of generalization steps yields accurate semantic parsers sampling ≤10% of source training data in each new language. Our approach also trains a competitive model on Spider using English with generalization to Chinese similarly sampling ≤10% of training data.

Meta-Learning a Cross-lingual Manifold for Semantic Parsing

How do different tokenizers perform on downstream tasks in scriptio continua languages?: A case study in Japanese

<iframe src="https://app.sli.do/event/uyG7EHDP5trs2LHubv2Gj8/embed/polls/13d959f9-994d-411c-9bc3-5c07458a544d" width="300" height="400"></iframe>

How do different tokenizers perform on downstream tasks in scriptio continua languages?: A case study in Japanese

Recent studies in natural language processing (NLP) have focused on modern languages and achieved state-of-the-art results in many tasks. Meanwhile, little attention has been paid to ancient texts and related tasks. Classical Chinese first came to Japan approximately 2,000 years ago. It was gradually adapted to a Japanese form called Kanbun-Kundoku (Kanbun) in Japanese reading and translating methods, which has significantly impacted Japanese literature. However, compared to the rich resources of ancient texts in mainland China, Kanbun resources remain scarce in Japan. To solve this problem, we construct the first Classical-Chinese-to-Kanbun dataset in the world. Furthermore, we introduce two tasks, character reordering and machine translation, both of which play a significant role in Kanbun comprehension. We also test the current language models on these tasks and discuss the best evaluation method by comparing the results with human scores. We release our code and dataset on GitHub.

Kanbun-LM: Reading and Translating Classical Chinese in Japanese Methods by Language Models

While embedding-based methods have been dominant in language clustering for multilingual tasks, clustering based on linguistic features has not yet been explored much, as it remains baselines (Tan et al., 2019; Shaffer, 2021). This study investigates whether and how theoretical linguistics improves language clustering for multilingual named entity recognition (NER). We propose two types of language groupings: one based on morpho-syntactic features in a nominal domain and one based on a head parameter. Our NER experiments show that the proposed methods largely outperform a state-of-the-art embedding-based model, suggesting that theoretical linguistics plays a significant role in multilingual learning tasks.

Downloads

Next from ACL 2023

Improved Logical Reasoning of Language Models via Differentiable Symbolic Programming

Similar lecture

Beyond Single-User Dialogue: Assessing Multi-User Dialogue State Tracking Capabilities of Large Language Models

.css-70qvj9{display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;-webkit-align-items:center;-webkit-box-align:center;-ms-flex-align:center;align-items:center;}Downloads

Next from ACL 2023

Improved Logical Reasoning of Language Models via Differentiable Symbolic Programming

Similar lecture

Beyond Single-User Dialogue: Assessing Multi-User Dialogue State Tracking Capabilities of Large Language Models

Downloads