Malta

Simultaneous interpretation is an especially challenging form of translation because it requires converting speech from one language to another in real-time. Though prior work has relied on out-of-the-box machine translation metrics to evaluate interpretation data, we hypothesize that strategies common in high-quality human interpretations, such as summarization, may not be handled well by standard machine translation metrics. In this work, we examine both qualitatively and quantitatively four potential barriers to evaluation of interpretation: disfluency, summarization, paraphrasing, and segmentation. Our experiments reveal that, while some machine translation metrics correlate fairly well with human judgments of interpretation quality, much work is still needed to account for strategies of interpretation during evaluation. As a first step to address this, we develop a fine-tuned model for interpretation evaluation, and achieve better correlation with human judgments than the state-of-the-art machine translation metrics.

EACL 2024

Barriers to Effective Evaluation of Simultaneous Interpretation

translation

interpretation

evaluation

metrics

machine translation

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.

poster

We introduce MultiMUC, the first multilingual parallel corpus for template filling, comprising translations of the classic MUC-4 template filling benchmark into five languages: Arabic, Chinese, Farsi, Korean, and Russian. We obtain automatic translations from a strong multilingual machine translation system and manually project the original English annotations into each target language. For all languages, we also provide human translations for key portions of the dev and test splits. Finally, we present baselines on MultiMUC both with state-of-the-art template filling models for MUC-4 and with ChatGPT. We release MUC-4 and the supervised baselines to facilitate further work on document-level information extraction in multilingual settings.

MultiMUC: Multilingual Template Filling on MUC-4

The main objective of Knowledge Graph (KG) embeddings is to learn low-dimensional representations of entities and relations, enabling the prediction of missing facts. A significant challenge in achieving better KG embeddings lies in capturing relation patterns, including symmetry, antisymmetry, inversion, commutative composition, non-commutative composition, hierarchy, and multiplicity. This study introduces a novel model called 3H-TH (3D Rotation and Translation in Hyperbolic space) that captures these relation patterns simultaneously. In contrast, previous attempts have not achieved satisfactory performance across all the mentioned properties at the same time. The experimental results demonstrate that the new model outperforms existing state-of-the-art models in terms of accuracy, hierarchy property, and other relation patterns in low-dimensional space, meanwhile performing similarly in high-dimensional space.

3D Rotation and Translation for Hyperbolic Knowledge Graph Embedding

We propose attribute-aware multimodal entity linking, where the input consists of a mention described with a text paragraph and images, and the goal is to predict the corresponding target entity from a multimodal knowledge base (KB) where each entity is also accompanied by a text description, visual images, and a collection of attributes that present the meta-information of the entity in a structured format. 
To facilitate this research endeavor, we construct Ameli, encompassing a new multimodal entity linking benchmark dataset that contains 16,735 mentions described in text and associated with 30,472 images, and a multimodal knowledge base that covers 34,690 entities along with 177,873 entity images and 798,216 attributes. To establish baseline performance on Ameli, we experiment with several state-of-the-art architectures for multimodal entity linking and further propose a new approach that incorporates attributes of entities into disambiguation. Experimental results and extensive qualitative analysis demonstrate that extracting and understanding the attributes of mentions from their text descriptions and visual images play a vital role in multimodal entity linking. To the best of our knowledge, we are the first to integrate attributes in the multimodal entity linking task. The programs, model checkpoints, and the dataset are publicly available at https://github.com/VT-NLP/Ameli.

Ameli: Enhancing Multimodal Entity Linking with Fine-Grained Attributes

Recent evidence from cognitive science suggests that there exist two classes of cognitive representations within the spatial terms of a language, one represented geometrically (e.g., above, below) and the other functionally (e.g., on, in). It has been hypothesized that geometric terms are more constrained and are mastered relatively early in language learning, whereas functional terms are less constrained and are mastered over longer time periods (Landau, 2016). One consequence of this hypothesis is that these two classes should exhibit different cross-linguistic variability, which is supported by human elicitation studies. 

In this work we present to our knowledge the first corpus-based empirical test of this hypothesis. We develop a pipeline for extracting, isolating, and aligning spatial terms in basic locative constructions from parallel text. Using Shannon entropy to measure the variability of spatial term use across eight languages, we find supporting evidence that variability in functional terms differs significantly from that of geometric terms. We also perform latent variable modeling and find support for the division of spatial terms into geometric and functional classes.

Large-Scale Bitext Corpora Provide New Evidence for Cognitive Representations of Spatial Terms

The escalating debate on AI's capabilities warrants developing reliable metrics to assess machine "intelligence." Recently, many anecdotal examples were used to suggest that newer Large Language Models (LLMs) like ChatGPT and GPT-4 exhibit Neural Theory-of-Mind (N-ToM); however, prior work reached conflicting conclusions regarding those abilities. We investigate the extent of LLMs' N-ToM through an extensive evaluation of 6 tasks and find that while LLMs exhibit certain N-ToM abilities, this behavior is far from being robust. We further examine the factors impacting performance on N-ToM tasks and discover that LLMs struggle with adversarial examples, indicating reliance on shallow heuristics rather than robust ToM abilities. We caution against drawing conclusions from anecdotal examples, limited benchmark testing, and using human-designed psychological tests to evaluate models.

Clever Hans or Neural Theory of Mind? Stress Testing Social Reasoning in Large Language Models

Natural Language Processing (NLP) can advance psychotherapy research by scaling up therapy dialogue analysis as well as by allowing researchers to examine client-therapist interactions in detail. Previous studies have mainly either explored the clients' behavior or the therapists' intervention in dialogues. Yet, modelling conversations from both dialogue participants is crucial to understanding the therapeutic interaction. This study explores speaker contribution-based dialogue acts at the utterance-level; i.e, the therapist - Intervention Prediction (IP) and the client - Emotion Recognition (ER) in psychotherapy using a pan-theoretical schema. We perform experiments with fine-tuned language models and light-weight adapter solutions on a Hebrew dataset. We deploy the results from our ER model predictions in investigating the coherence between client self-reports on emotion and the utterance-level emotions. Our best adapters achieved on-par performance with fully fine-tuned models, at 0.64 and 0.66 micro F1 for IP and ER, respectively. In addition, our analysis identifies ambiguities within categorical clinical coding, which can be used to fine-tune the coding schema. Finally, our results indicate a positive correlation between client self-reports and utterance-level emotions.

Predicting Client Emotions and Therapist Interventions in Psychotherapy Dialogues

The propagation of offensive content through social media channels has garnered attention of the research community. Multiple works have proposed various semantically related yet subtle distinct categories of offensive speech. In this work, we explore meta-learning approaches to leverage the diversity of offensive speech corpora to enhance their reliable and efficient detection. We propose a joint embedding architecture that incorporates the input's label and definition for classification via Prototypical Network. Our model achieves at least 75% of the maximal F1-score while using less than 10% of the available training data across 4 datasets. Our experimental findings also provide a case study of training strategies valuable to combat resource scarcity.

“Define Your Terms” : Enhancing Efficient Offensive Speech Classification with Definition

The concept of {\em acquired spatial knowledge} is crucial in spatial cognitive research, particularly when it comes to communicating routes. However, NLP navigation studies often overlook the impact of acquired knowledge on textual descriptions. Current navigation studies concentrate on egocentric local descriptions (e.g., `it will be on your right') that require reasoning over the agent's local perception. These instructions are typically given in a sequence of steps, with each action-step explicitly mentioned and followed by a landmark that the agent can use to verify that they are on the correct path (e.g., `turn right and then you will see...'). 
In contrast, descriptions based on knowledge acquired through a map provide a complete view of the environment and capture its compositionality. These instructions typically contain allocentric relations, are non-sequential, with implicit actions and multiple spatial relations without any verification (e.g., `south of Central Park and a block north of a police station'). This paper introduces the Rendezvous (RVS) task and dataset, which includes 10,404 examples of English geospatial instructions for reaching a target location using map-knowledge. Our analysis reveals that RVS exhibits a richer use of spatial allocentric relations, and requires resolving more spatial relations simultaneously compared to previous text-based navigation benchmarks.

Where Do We Go From Here? Multi-scale Allocentric Relational Inferencefrom Natural Spatial Descriptions

With the rapid evolution of large language models (LLMs), new and hard-to-predict harmful capabilities are emerging. This requires developers to identify potential risks through the evaluation of ``dangerous capabilities'' in order to responsibly deploy LLMs. Here we aim to facilitate this process. In particular, we collect an open-source dataset to evaluate the safeguards in LLMs, to facilitate the deployment of safer open-source LLMs at a low cost. Our dataset is curated and filtered to consist only of instructions that responsible language models should not follow. 
We assess the responses of six popular LLMs to these instructions, and we find that simple BERT-style classifiers can achieve results that are comparable to GPT-4 on automatic safety evaluation. Our data and code are available at https://github.com/Libr-AI/do-not-answer

Do-Not-Answer: Evaluating Safeguards in LLMs

With the recent emergence of powerful visio-linguistic models comes the question of how fine-grained their multi-modal understanding is. This has lead to the release of several probing datasets. Results point towards models having trouble with prepositions and verbs, but being relatively robust when it comes to color.

To gauge how deep this understanding goes, we compile a comprehensive probing dataset to systematically test multi-modal alignment around color. We demonstrate how human perception influences descriptions of color and pay special attention to the extent to which this is reflected within the predictions of a visio-linguistic model. Probing a set of models with diverse properties with our benchmark confirms the superiority of models that do not rely on pre-extracted image features, and demonstrates that augmentation with too much noisy pre-training data can produce an inferior model. While the benchmark remains challenging for all models we test, the overall result pattern suggests well-founded alignment of color terms with hues. Analyses do however reveal uncertainty regarding the boundaries between neighboring color terms.

Downloads

Next from EACL 2024

MultiMUC: Multilingual Template Filling on MUC-4

Similar lecture

Barriers to Effective Evaluation of Simultaneous Interpretation