Malta

Over 15 years ago, Ward and Birner (2006) suggested that non-canonical constructions in English can serve both to mark information status and to structure the information flow of discourse. One such construction is preposing, where a phrasal constituent appears to the left of its canonical position, typically sentence-initially. But computational work on discourse has, to date, ignored non-canonical syntax. We take account of non-canonical syntax by providing quantitative evidence relating NP/PP preposing to discourse relations. The evidence comes from an LLM mask-filling task that compares the predictions when a mask is inserted between the arguments of an implicit inter-sentential discourse relation -- first, when the right-hand argument (Arg2) starts with a preposed constituent, and again, when that constituent is in canonical (post-verbal) position. Results show that (1) the top-ranked mask-fillers in the preposed case agree more often with &quot;gold&quot; annotations in the Penn Discourse Treebank (Webber et al., 2019) than they do in the latter case, and (2) preposing in Arg2 can affect the distribution of discourse-relational senses.

EACL 2024

Syntactic Preposing and Discourse Relations

syntactic preposing

large language model

discourse relations

Welcome to the 18th Conference of the European Chapter of the Association for Computational Linguistics (EACL). Continuing its mission of expanding and involving the science community of all European countries, EACL has selected the Malta community for the 18th EACL. Considering the importance of physical interaction among researchers, the conference will be held at the Hotel Radisson Blu, St. Julians, in Malta, from 17 to 22 of March, 2024. The conference will also feature the possibility to participate virtually. As the flagship European conference in the field of computational linguistics, EACL welcomes European and international researchers covering a broad spectrum of research areas that are concerned with computational approaches to natural language. 
#### [Conference Handbook](https://drive.google.com/file/d/13Wn38q0yev6U3RTgnf4Tn5T9GuAuZ-xW/view?usp=sharing) *downloadable*

To access the EACL 2024 event page you need to register [**here**](https://acl.swoogo.com/EACL2024). 

Please register!

As the flagship European conference in the field of computational linguistics, EACL welcomes European and international researchers covering a broad spectrum of research areas that are concerned with computational approaches to natural language.

Over 15 years ago, Ward and Birner (2006) suggested that non-canonical constructions in English can serve both to mark information status and to structure the information flow of discourse. One such construction is preposing, where a phrasal constituent appears to the left of its canonical position, typically sentence-initially. But computational work on discourse has, to date, ignored non-canonical syntax. We take account of non-canonical syntax by providing quantitative evidence relating NP/PP preposing to discourse relations. The evidence comes from an LLM mask-filling task that compares the predictions when a mask is inserted between the arguments of an implicit inter-sentential discourse relation -- first, when the right-hand argument (Arg2) starts with a preposed constituent, and again, when that constituent is in canonical (post-verbal) position. Results show that (1) the top-ranked mask-fillers in the preposed case agree more often with "gold" annotations in the Penn Discourse Treebank (Webber et al., 2019) than they do in the latter case, and (2) preposing in Arg2 can affect the distribution of discourse-relational senses.

workshop paper

Recent language models have significantly boosted conversational AI by enabling fast and cost-effective response generation in dialogue systems. However, dialogue systems based on neural generative approaches often lack truthfulness, reliability, and the ability to analyze the dialogue flow needed for smooth and consistent conversations with users. To address these issues, we introduce GroundHog, a modified BART architecture, to capture long multi-grained inputs gathered from various factual and linguistic sources, such as Abstract Meaning Representation, discourse relations, sentiment, and grounding information. For experiments, we present an automatically collected dataset from Reddit that includes multi-party conversations devoted to movies and TV series. The evaluation encompasses both automatic evaluation metrics and human evaluation. The obtained results demonstrate that using several linguistic inputs has the potential to enhance dialogue consistency, meaningfulness, and overall generation quality, even for automatically annotated data. We also provide an analysis that highlights the importance of individual linguistic features in interpreting the observed enhancements.

GroundHog: Dialogue Generation using Multi-Grained Linguistic Input

Discourse analysis plays a crucial role in Natural Language Processing, with discourse relation prediction arguably being the most difficult task in discourse parsing.
Previous studies have generally focused on explicit or implicit discourse relation classification in monologues, leaving dialogue an under-explored domain.
Facing the data scarcity issue, we propose to leverage self-training strategies based on a Transformer backbone. Moreover, we design the first semi-supervised pipeline that sequentially predicts discourse structures and relations.
Using 50 examples, our relation prediction module achieves 58.4 in accuracy on the STAC corpus, close to supervised state-of-the-art.
Full parsing results show notable improvements compared to the supervised models both in-domain (gaming) and cross-domain (technical chat), with better stability.

Discourse Relation Prediction and Discourse Parsing in Dialogues with Minimal Supervision

Topics play an important role in the global organisation of a conversation as what is currently discussed constrains the possible contributions of the participant. Understanding the way topics are organised in interaction would provide insight on the structure of dialogue beyond the sequence of utterances. However, studying this high-level structure is a complex task that we try to approach by first segmenting dialogues into smaller topically coherent sets of utterances. Understanding the interactions between these segments would then enable us to propose a model of topic organisation at a dialogue level. In this paper we work with open-domain conversations and try to reach a comparable level of accuracy as recent machine learning based topic segmentation models but with a formal approach. The features we identify as meaningful for this task help us understand better the topical structure of a conversation.

With a Little Help from my (Linguistic) Friends: Topic segmentation of multi-party casual conversations

Existing discourse corpora are annotated based on different frameworks, which show significant dissimilarities in definitions of arguments and relations and structural constraints. Despite differences at surface level, these discourse models share some assumptions, such as the role of discourse relations. Studies that explore
correlations between discourse relation inventories are hindered by different segmentation criteria, and expert knowledge and manual examination are typically needed. In this paper, we introduce a fully automatic approach to address the challenges. Specifically, we extend the label-anchored contrastive learning method proposed by Zhang et al. (2022) to learn label embeddings during a classification task. These embeddings are then utilized to map discourse relations from different frameworks.

Learning Label Embeddings for Mapping Discourse Relations of Different Annotation Frameworks (Extended Abstract)

Document-level translation models are usually evaluated using general metrics such as BLEU, which are not informative about the benefits of context. Current work on context-aware evaluation, such as contrastive methods, only measure translation accuracy on words that need context for disambiguation. Such measures cannot reveal whether the translation model uses the correct supporting context. We propose to complement accuracy-based evaluation with measures of context utilization. We find that perturbation-based analysis (comparing models' performance when provided with correct versus random context) is an effective measure of overall context utilization. For a finer-grained phenomenon-specific evaluation, we propose to measure how much the supporting context contributes to handling context-dependent discourse phenomena. We show that automatically-annotated supporting context gives similar conclusions to human-annotated context and can be used as alternative for cases where human annotations are not available. Finally, we highlight the importance of using discourse-rich datasets when assessing context utilization.

On Measuring Context Utilization in Document-Level MT Systems

Foundational large language models (LLMs) can be instruction-tuned to perform open-domain question answering, facilitating applications like chat assistants. While such efforts are often carried out in a single language, we empirically analyze cost-efficient strategies for multilingual scenarios. Our study employs the Alpaca dataset and machine translations of it to form multilingual data, which is then used to tune LLMs through either low-rank adaptation or full-parameter training. Under a controlled computation budget, comparisons show that multilingual tuning is on par or better than tuning a model for each language. Furthermore, multilingual tuning with downsampled data can be as powerful and more robust. Our findings serve as a guide for expanding language support through instruction tuning.

Monolingual or Multilingual Instruction Tuning: Which Makes a Better Alpaca

The knowledge encapsulated in a model is the core factor determining its final performance on downstream tasks. Much research in NLP has focused on efficient methods for storing and adapting different types of knowledge, e.g., in dedicated modularized structures, and on how to effectively combine these modules, e.g., via parameter averaging at test time. However, given the many possible options in composing knowledge, a thorough understanding of the mechanisms involved is missing, and hence it remains unclear which strategies to utilize. In this work, we address this research gap by proposing a novel framework for zero-shot module composition, which encompasses existing and some novel variations for selecting, weighting, and combining parameter modules under a single unified notion. Focusing on the scenario of domain knowledge and adapter layers, our framework provides a systematic unification of concepts, allowing us to conduct the first comprehensive benchmarking study on various zero-shot knowledge composition strategies. In particular, we test two module combination methods (parameter averaging, output ensembling), and five selection and weighting strategies (uniform, and based on entropy, domain prior, TF-IDF, and semantic similarity) for their effectiveness and efficiency on 21 training and 10 evaluation domains across three models. Our results highlight the efficacy of ensembling, but also hint at the power of simple though often-ignored weighting methods. We further conduct various in-depth analyses, that, for instance, allow us to understand the role of weighting vs. top-k selection, and we show that, to a certain extent, the performance of adapter composition can even be predicted.

What the Weight?! A Unified Framework for Zero-Shot Knowledge Composition

State-of-the-art conversational AI systems raise concerns due to their potential risks of generating unsafe, toxic, unethical, or dangerous content. Previous works have developed datasets to teach conversational agents the appropriate social paradigms to respond effectively to specifically designed hazardous content. However, models trained on these adversarial datasets still struggle to recognize subtle unsafe situations that appear naturally in conversations or introduce an inappropriate response in a casual context. To understand the extent of this problem, we study prosociality in both adversarial and casual dialog contexts and audit the response quality of general-purpose language models in terms of propensity to produce unsafe content. We propose a dual-step fine-tuning process to address these issues using a socially aware n-pair contrastive loss. Subsequently, we train a base model that integrates prosocial behavior by leveraging datasets like Moral Integrity Corpus (MIC) and ProsocialDialog. Experimental results on several dialog datasets demonstrate the effectiveness of our approach in generating socially appropriate responses.

Improving Dialog Safety using Socially Aware Contrastive Learning

In the realm of dialogue systems, user simulation techniques have emerged as a game-changer, redefining the evaluation and enhancement of task-oriented dialogue (TOD) systems. These methods are crucial for replicating real user interactions, enabling applications like synthetic data augmentation, error detection, and robust evaluation. However, existing approaches often rely on rigid rule-based methods or on annotated data. This paper introduces DAUS, a Domain-Aware User Simulator. Leveraging large language models, we fine-tune DAUS on real examples of task-oriented dialogues. Results on two relevant benchmarks showcase significant improvements in terms of user goal fulfillment. Notably, we have observed that fine-tuning enhances the simulator's coherence with user goals, effectively mitigating hallucinations—a major source of inconsistencies in simulator responses.

Reliable LLM-based User Simulator for Task-Oriented Dialogue Systems

This paper introduces a novel approach to form-filling and dialogue system evaluation by leveraging Large Language Models (LLMs). The proposed method establishes a setup wherein multiple modules collaborate on addressing the form-filling task. The dialogue system is constructed on top of LLMs, focusing on defining specific roles for individual modules. We show that using multiple independent sub-modules working cooperatively on this task can improve performance and handle the typical constraints of using LLMs, such as context limitations. The study involves testing the modular setup on four selected forms of varying topics and lengths, employing commercial and open-access LLMs. The experimental results demonstrate that the modular setup consistently outperforms the baseline, showcasing the effectiveness of this approach. Furthermore, our findings reveal that open-access models perform comparably to commercial models for the specified task.

Downloads

Next from EACL 2024

GroundHog: Dialogue Generation using Multi-Grained Linguistic Input

Similar lecture

Syntactic Preposing and Discourse Relations

.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 EACL 2024

GroundHog: Dialogue Generation using Multi-Grained Linguistic Input

Similar lecture

Syntactic Preposing and Discourse Relations

Downloads