Canada

Verbal deception has been studied in psychology, forensics, and computational linguistics for a variety of reasons, like understanding behaviour patterns, identifying false testimonies, and detecting deception in online communication. Varying motivations across research fields lead to differences in the domain choices to study and in the conceptualization of deception, making it hard to compare models and build robust deception detection systems for a given language. With this paper, we improve this situation by surveying available English deception datasets which include domains like social media reviews, court testimonials, opinion statements on specific topics, and deceptive dialogues from online strategy games. We consolidate these datasets into a single unified corpus. Based on this resource, we conduct a correlation analysis of linguistic cues of deception across datasets to understand the differences and perform cross-corpus modeling experiments which show that a cross domain generalization is challenging to achieve.

ACL 2023

UNIDECOR: A Unified Deception Corpus for Cross-Corpus Deception Detection

deception

domain

corpus

robustness

workshop paper

### Welcome to ACL 2023, the 61st Annual Meeting of the Association for Computational Linguistics! 
 The conference will be held in Toronto, Canada, July 9-14, 2023. 
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. 
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)**

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.

The scarcity of annotated data is a major impediment to natural language processing (NLP) research in Bengali, a language that is considered low-resource. In particular, the health and medical domains suffer from a severe paucity of annotated data. Thus, this study aims to introduce BanglaSocialHealth, an annotated social media health corpus that provides sentence-level annotations of four distinct types of expression modes, namely narrative (NAR), informative (INF), suggestive (SUG), and inquiring (INQ) modes in Bengali. We provide details regarding the annotation procedures and report various statistics, such as the median and mean length of words in different sentence modes. Additionally, we apply classical machine learning (CML) classifiers and transformer-based language models to classify sentence modes. We find that most of the statistical properties are similar in different types of sentence modes. To determine the sentence mode, the transformer-based M-BERT model provides slightly better efficacy than the CML classifiers. Our developed corpus and analysis represent a much-needed contribution to Bengali NLP research in medical and health domains and have the potential to facilitate a range of downstream tasks, including question-answering, misinformation detection, and information retrieval.

Discourse Mode Categorization of Bengali Social Media Health Text

Paraphrase generation, a.k.a. paraphrasing, is a common and important task in natural language processing. Emotional paraphrasing, which changes the emotion embodied in a piece of text while preserving its meaning, has many potential applications, including moderating online dialogues and preventing cyberbullying. We introduce a new task of fine-grained emotional paraphrasing along emotion gradients, that is, altering the emotional intensities of the paraphrases in fine-grained settings following smooth variations in affective dimensions while preserving the meaning of the original text. We reconstruct several widely used paraphrasing datasets by augmenting the input and target texts with their fine-grained emotion labels. Then, we propose a framework for emotion and sentiment guided paraphrasing by leveraging pre-trained language models for conditioned text generation. Extensive evaluation of the fine-tuned models suggests that including fine-grained emotion labels in the paraphrase task significantly improves the likelihood of obtaining high-quality paraphrases that reflect the desired emotions while achieving consistently better scores in paraphrase metrics such as BLEU, ROUGE, and METEOR.

Emotion and Sentiment Guided Paraphrasing

Work on emotion detection is often focused on textual data from i.e. Social Media. If multimodal data (i.e. speech) is analysed, the focus again is often placed on the transcription. 
This paper takes a closer look at how crucial acoustic information actually is for the recognition of emotions from multimodal data. To this end we use the IEMOCAP data, which is one of the larger data sets that provides transcriptions, audio recordings and manual emotion categorization. We build models for emotion classification using text-only, acoustics-only and combining both modalities in order to examine the influence of the various modalities on the final categorization. Our results indicate that using text-only models outperform acoustics-only models. But combining text-only and acoustic-only models improves the results. 
Additionally, we perform a qualitative analysis and find that a range of misclassifications are due to factors not related to the model, but to the data such as, recording quality, a challenging classification task and misclassifications that are unsurprising for humans.

Emotions in Spoken Language - Do we need acoustics?

The valence analysis of speakers' utterances or written posts helps to understand the activation and variations of the emotional state throughout the conversation. More recently, the concept of Emotion Carriers (EC) has been introduced to explain the emotion felt by the speaker and its manifestations. In this work, we investigate the natural inter-dependency of valence and ECs via a multi-task learning approach. We experiment with Pre-trained Language Models (PLM) for single-task, two-step, and joint settings for the valence and EC prediction tasks. We compare and evaluate the performance of generative (GPT-2) and discriminative (BERT) architectures in each setting. We observed that providing the ground truth label of one task improves the prediction performance of the models in the other task. We further observed that the discriminative model achieves the best trade-off of valence and EC prediction tasks in the joint prediction setting. As a result, we attain a single model that performs both tasks, thus, saving computation resources at training and inference times.

Understanding Emotion Valence is a Joint Deep Learning Task

We present the Verifee dataset: a multimodal dataset of news articles with fine-grained trustworthiness annotations. We bring a diverse set of researchers from social, media, and computer sciences aboard to study this interdisciplinary problem holistically and develop a detailed methodology that assesses the texts through the lens of editorial transparency, journalist conventions, and objective reporting while penalizing manipulative techniques. We collect over 10,000 annotated articles from 60 Czech online news sources. Each item is categorized into one of the 4 proposed classes on the credibility spectrum -- ranging from entirely trustworthy articles to deceptive ones -- and annotated of its manipulative attributes. We fine-tune prominent sequence-to-sequence language models for the trustworthiness classification task on our dataset and report the best F-1 score of 0.53. We open-source the dataset, annotation methodology, and annotators' instructions in full length at https://www.verifee.ai/research/ to enable easy build-up work.

Czech-ing the News: Article Trustworthiness Dataset for Czech

Manipulated news online is a growing problem which necessitates the use of automated systems to curtail its spread. We argue that while misinformation and disinformation detection have been studied, there has been a lack of investment in the important open challenge of detecting harmful agendas in news articles; identifying harmful agendas is critical to flag news campaigns with the greatest potential for real world harm. Moreover, due to real concerns around censorship, harmful agenda detectors must be interpretable to be effective. In this work, we propose this new task and release a dataset, \textsc{NewsAgendas}, of annotated news articles for agenda identification. We show how interpretable systems can be effective on this task and demonstrate that they can perform comparably to black-box models.

Towards Detecting Harmful Agendas in News Articles

Sentiment Analysis is an important task for
analysing online content across languages for
tasks such as content moderation and opinion
mining. Though a significant amount of resources
are available for Sentiment Analysis in
several Indian languages, there do not exist any
large-scale, open-access corpora for Gujarati.
Our paper presents and describes the Gujarati
Sentiment Analysis Corpus (GSAC), which has
been sourced from Twitter and manually annotated
by native speakers of the language. We
describe in detail our collection and annotation
processes and conduct extensive experiments
on our corpus to provide reliable baselines for
future work using our dataset.

GSAC: A Gujarati Sentiment Analysis Corpus from Twitter

While deep learning models have greatly improved the performance of many tasks related to sentiment analysis and classification, they are often criticized for being untrustworthy due to their black-box nature. As a result, numerous explainability techniques have been proposed to better understand the model predictions and
to improve the deep learning models. In this work, we introduce InfoBarometer, the first benchmark for examining interpretable methods related to sentiment analysis in the German automotive sector based on online news. Each news article in our dataset is annotated with respect to the overall sentiment (i.e., positive, negative and neutral), the target of the sentiment (focusing on innovation-related topics such as e.g. electromobility) and the rationales, i.e., textual explanations for the sentiment label that can be leveraged during both training and evaluation. For this research, we compare different state-of-the-art approaches to perform sentiment analysis and observe that even models that perform very well in classification do not score high on explainability metrics like model plausibility and faithfulness. We calculated
the polarity scores for the best method BERT and got a macro F1-score of 73.8. Moreover, we evaluated different interpretability algorithms (LIME, SHAP, Integrated Gradients, Saliency) based on explicitly marked rationales by human annotators quantitatively and qualitatively.
Our experiments demonstrate that the textual explanations often do not agree with human interpretations, and rarely help to justify the models decision. However, global features provide useful insights to help uncover spuriousfeatures in the model and biases within the dataset. We intend to make our dataset public for other researchers.

An XAI Dataset for Mining the Sentimental Mood of the German Automotive Industry

Opinion summarisation is a task that aims to condense the information presented in the source documents while retaining the core message and opinions.
A summary that only represents the majority opinions will leave the minority opinions unrepresented in the summary. In this paper, we use the stance towards a certain target as an opinion. We study bias in opinion summarisation from the perspective of opinion diversity, which measures whether the model generated summary can cover a diverse set of opinions. In addition, we examine opinion similarity, a measure of how closely related two opinions are in terms of their stance on a given topic, and its relationship with opinion diversity. Through the lens of stances towards a topic, we examine opinion diversity and similarity using three debatable topics under COVID-19. Experimental results on these topics revealed that a higher degree of similarity of opinions did not indicate good diversity or fairly cover the various opinions originally presented in the source documents. We found that BART and ChatGPT can better capture diverse opinions presented in the source documents.

Examining Bias in Opinion Summarisation through the Perspective of Opinion Diversity

Unsupervised text style transfer is a challenging task that aims to alter the stylistic attributes of a given text without affecting its original content. One of the methods to achieve this is controllable style transfer, which allows for the control of the degree of style transfer. However, an issue encountered with controllable style transfer is the instability of transferred text fluency when the degree of the style transfer changes. To address this problem, we propose a novel approach that incorporates additional syntax parsing information during style transfer. By leveraging the syntactic information, our model is guided to generate natural sentences that effectively reflect the desired style while maintaining fluency. Experimental results show that our method achieves robust performance and improved fluency compared to previous controllable style transfer methods.

UNIDECOR: A Unified Deception Corpus for Cross-Corpus Deception Detection

Downloads

Next from ACL 2023

Discourse Mode Categorization of Bengali Social Media Health Text

Stay up to date with the latest Underline news!

PRESENTATIONS

CONFERENCES

COMPANY

RESOURCES