Croatia

Recent studies have suggested that neural language models learn and store a large amount of facts and commonsense knowledge from training data. The ability of language models to restore such knowledge is often evaluated via zero-shot cloze-style QA tasks. However, such evaluations rely only on prediction accuracy without punishing the systems for their mistakes, e.g., simply guessing or hallucinating likely answers. Selective prediction is a more informative evaluation framework that takes the confidence of predictions into account. Under the selective prediction setting, a model is evaluated not only by the number of correct predictions, but also by the ability to filter out dubious predictions by estimating the confidence of individual predictions. Such confidence-aware evaluation is crucial for determining whether to
trust zero-shot predictions of language models. In this paper, we apply the selective prediction setting to an existing benchmark, LAMA probe, and conduct extensive experiments with recent neural language models and different confidence functions. We empirically show that our Selective-LAMA evaluation is more robust to the effect of simple guesses than the conventional accuracy-based evaluation. Our
evaluation reveals the importance of the choice of confidence functions by showing that simply relying on token probabilities is not always the best choice. Further analysis shows that various confidence functions exhibit different preferences over predicted tokens for a given context.

EACL 2023

Selective-LAMA: Selective Prediction for Confidence-Aware Evaluation of Language Models

selective prediction

pre-trained language models

evaluation

### Welcome to the 17th 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 had selected the Ukrainian community for the 16th EACL, which took place online due to the COVID pandemic. Unfortunately, the ongoing war made the organisation in Kyiv impossible. Considering the importance of physical interaction among researchers, especially after the restrictions imposed by the COVID pandemic, in addition to an online mode, the conference will be held in Dubrovnik, Croatia, from 2 to 6 of May, 2023. The original aim of strengthening the connection with the Ukrainian community will not change as our program will feature a dedicated session and a workshop to highlight work on Ukrainian language technologies. 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.

You need to register for the conference in order to access this site. Please visit https://2023.eacl.org/registration for more information.

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.

workshop paper

Contrastive Language--Image Pre-training (CLIP) has shown remarkable success in learning with cross-modal supervision from extensive amounts of image--text pairs collected online. Thus far, the effectiveness of CLIP has been investigated primarily in general-domain multimodal problems. In this work, we evaluate the effectiveness of CLIP for the task of Medical Visual Question Answering (MedVQA). We present PubMedCLIP, a fine-tuned version of CLIP for the medical domain based on PubMed articles. Our experiments conducted on two MedVQA benchmark datasets illustrate that PubMedCLIP achieves superior results improving the overall accuracy up to 3% in comparison to the state-of-the-art Model-Agnostic Meta-Learning (MAML) networks pre-trained only on visual data. 
The PubMedCLIP model with different back-ends, the source code for pre-training them and reproducing our MedVQA pipeline is publicly available at https://github.com/sarahESL/PubMedCLIP.

PubMedCLIP: How Much Does CLIP Benefit Visual Question Answering in the Medical Domain?

Open Information Extraction (OIE) is the task of extracting tuples of the form (subject, predicate, object), without any knowledge of the type and lexical form of the predicate, the subject, or the object. In this work, we focus on improving OIE quality by exploiting domain knowledge about the subject and object. 
More precisely, knowing that the subjects and objects in sentences are often named entities, we explore how to inject constraints in the extraction through constrained inference and constraint-aware training. 
Our work leverages the state-of-the-art OpenIE6 platform, which we adapt to our setting. Through a carefully constructed training dataset and constrained training, we obtain a $29.17\%$ F1-score improvement in the CaRB metric and a $24.37\%$ F1-score improvement in the WIRe57 metric. Our technique has important applications -- one of them is investigative journalism, where automatically extracting conflict-of-interest between scientists and funding organizations helps understand the type of relations companies engage with the scientists. Our code and data are available at \url{https://github.com/prajnaupadhyay/openie-with-entities}

Open Information Extraction with Entity Focused Constraints

Quotes are critical for establishing credibility in news articles. A direct quote enclosed in quotation marks has a strong visual appeal and is a sign of a reliable citation. Unfortunately, this journalistic practice is not strictly followed, and a quote in the headline is often “contextomized." Such a quote uses words out of context in a way that alters the speaker’s intention so that there is no semantically matching quote in the body text. We present QuoteCSE, a contrastive learning framework that represents the embedding of news quotes based on domain-driven positive and negative samples to identify such an editorial strategy. The dataset and code are available at https://github.com/ssu-humane/
contextomized-quote-contrastive.

Detecting Contextomized Quotes in News Headlines by Contrastive Learning

To mitigate the impact of the scarcity of labelled data on fact-checking systems, we focus on few-shot claim verification. Despite recent work on few-shot classification by proposing advanced language models, there is a dearth of research in data annotation prioritisation that improves the selection of the few shots to be labelled for optimal model performance. We propose Active PETs, a novel weighted approach that utilises an ensemble of Pattern Exploiting Training (PET) models based on various language models, to actively select unlabelled data as candidates for annotation. Using Active PETs for few-shot data selection shows consistent improvement over the baseline methods, on two technical fact-checking datasets and using six different pretrained language models. We show further improvement with Active PETs-o, which further integrates an oversampling strategy. Our approach enables effective selection of instances to be labelled where unlabelled data is abundant but resources for labelling are limited, leading to consistently improved few-shot claim verification performance.

Active PETs: Active Data Annotation Prioritisation for Few-Shot Claim Verification with Pattern Exploiting Training

Recent work has shown that an answer verification step introduced in Transformer-based answer selection models can significantly improve the state of the art in Question Answering. This step is performed by aggregating the embeddings of top $k$ answer candidates to support the verification of a target answer. Although the approach is intuitive and sound, it still shows two limitations: (i) the supporting candidates are ranked only according to the relevancy with the question and not with the answer, and (ii) the support provided by the other answer candidates is suboptimal as these are retrieved independently of the target answer. 
In this paper, we address both drawbacks by proposing (i) a double reranking model, which, for each target answer, selects the best support; and (ii) a second neural retrieval stage designed to encode question and answer pair as the query, which finds more specific verification information. The results on well-known datasets for Answer Sentence Selection show significant improvement over the state of the art.

Double Retrieval and Ranking for Accurate Question Answering

Poem generation with language models requires the modeling of rhyming patterns. We propose a novel solution for learning to rhyme, based on synthetic data generated with a rule-based rhyming algorithm. The algorithm and an evaluation metric use a phonetic dictionary and the definitions of perfect and assonant rhymes. We fine-tune a GPT-2 English model with 124M parameters on 142~MB of natural poems and find that this model generates consecutive rhymes infrequently (11%). We then fine-tune the model on 6 MB of synthetic quatrains with consecutive rhymes (AABB) and obtain nearly 60% of rhyming lines in samples generated by the model. Alternating rhymes (ABAB) are more difficult to model because of longer-range dependencies, but they are still learnable from synthetic data, reaching 45% of rhyming lines in generated samples.

GPoeT: a Language Model Trained for Rhyme Generation on Synthetic Data

Scientific writing is assumed to have become
more informationally dense over time (Halliday,
1988; Biber and Gray, 2016). Given that
scientific writing is intended for communication
between experts, we hypothesize a tendency
towards optimizing language use by striving
for a balance between highly informative
content and a conventionalized style of writing.
We study this by means of fractal analysis, asking
whether the degree of informativity has become
more persistent with predictable patterns
of gradual changes between high vs. low informational
content, indicating a trend towards
an optimal code for scientific communication.
Specifically, surprisal is used to measure informativity
and the Hurst exponent is used as a
long-term dependence measure for fractality
analysis, quantifying the degree of persistence
of informativity in scientific texts.

Fractality of informativity in 300 years of English scientific writing

Emotion regulation is a crucial element in dealing with emotional events and has positive effects on mental health. This paper aims to provide a more comprehensive understanding of emotional events by introducing a new French corpus of emotional narratives collected using a questionnaire for emotion regulation. We follow the theoretical framework of the Component Process Model which considers emotions as dynamic processes composed of four interrelated components (behavior, feeling, thinking and territory). Each narrative is related to a discrete emotion and is structured based on all emotion components by the writers. We study the interaction of components and their impact on emotion classification with machine learning methods and pre-trained language models. Our results show that each component improves prediction performance, and that the best results are achieved by jointly considering all components. Our results also show the effectiveness of pre-trained language models in predicting discrete emotion from certain components, which reveal differences in how emotion components are expressed.

Emotion Recognition based on Psychological Components in Guided Narratives for Emotion Regulation

This presentation describes the process of interlinking the Neulateinische Wortliste, a a lexical resource consisting of a list of more than 20,000 Neo-Latin words, with other resources for Latin. The resources are made interoperable thanks to their linking to the LiLa Knowledge Base, which applies Linguistic Linked Open Data practices and data categories to describe and publish on the Web both textual and lexical resources for the Latin language

Linking the Neulateinische Wortliste to the anonymous Knowledge Base of Interoperable Resources for Latin

Towards computational systems capable of dealing with complex and general linguistic phenomena, it is essential to understand figurative language, which verbal humor is an instance of. This paper reports state-of-the-art results for Humor Recognition in Portuguese, specifically, an F1-score of 99.6% with a BERTbased classifier. However, following the surprising high performance in such a challenging task, we further analyzed what was actually learned by the classifiers. Our main conclusions were that classifiers based on contentfeatures achieve the best performance, but rely mostly on stylistic aspects of the text, not necessarily related to humor, such as punctuation and question words. On the other hand, for humor-related features, we identified some important aspects, such as the presence of named entities, ambiguity and incongruity.

Downloads

Next from EACL 2023

PubMedCLIP: How Much Does CLIP Benefit Visual Question Answering in the Medical Domain?

Similar lecture

Selective-LAMA: Selective Prediction for Confidence-Aware Evaluation of Language Models