iCompass at WANLP 2022 Shared Task: ARBERT and MARBERT for Multilabel Propaganda Classification of Arabic Tweets

Propaganda content has seen massive spread in the biggest social media networks. Major global events such as Covid-19, presidential elections, and wars have all been infested with various propaganda techniques. In participation in the WANLP 2022 Shared Task(Alam et al., 2022), this paper provides a detailed overview of our machine learning system for propaganda techniques classification and its achieved results. The task was carried out using pre-trained transformer based models: ARBERT and MARBERT. The models were fine-tuned for the downstream task in hand: multilabel classification of Arabic tweets. According to the results, MARBERT and ARBERT attained 0.562 and 0.567 micro F1-score on the development set of subtask 1. The submitted model was MARBERT which attained a 0.597 micro F1-score and got the fifth rank. © 2022 Association for Computational Linguistics.

A Multilingual Approach to Identify and Classify Exceptional Measures Against COVID-19

The COVID-19 pandemic has witnessed the implementations of exceptional measures by governments across the world to counteract its impact. This work presents the initial results of an on-going project, EXCEPTIUS, aiming to automatically identify, classify and compare exceptional measures against COVID-19 across 32 countries in Europe. To this goal, we created a corpus of legal documents with sentence-level annotations of eight different classes of exceptional measures that are implemented across these countries. We evaluated multiple multi-label classifiers on a manually annotated corpus at sentence level. The XLM-RoBERTa model achieves highest performance on this multilingual multi-label classification task, with a macro-average F1 score of 59.8%. © 2021 Association for Computational Linguistics.

Detecting Fake News Conspiracies with Multitask and Prompt-Based Learning

In this paper, we present our participation to the MediaEval-2021 challenge on fake news detection about coronavirus related Tweets. It consists in three subtasks that can be seen as multi-labels classification problems we solved with transformer-based models. We show that each task can be solved independantly with mutiple monotasks models or jointly with an unique multitasks model. Moreover, we propose a prompt-based model that has been finetuned to generate classifications from a pre-trained model based on DistilGPT-2. Our experimental results show the multitask model to be the best to solve the three tasks. Copyright 2021 for this paper by its authors.

CAVES: A Dataset to facilitate Explainable Classification and Summarization of Concerns towards COVID Vaccines

Convincing people to get vaccinated against COVID-19 is a key societal challenge in the present times. As a first step towards this goal, many prior works have relied on social media analysis to understand the specific concerns that people have towards these vaccines, such as potential side-effects, ineffectiveness, political factors, and so on. Though there are datasets that broadly classify social media posts into Anti-vax and Pro-Vax labels, there is no dataset (to our knowledge) that labels social media posts according to the specific anti-vaccine concerns mentioned in the posts. In this paper, we have curated CAVES, the first large-scale dataset containing about 10k COVID-19 anti-vaccine tweets labelled into various specific anti-vaccine concerns in a multi-label setting. This is also the first multi-label classification dataset that provides explanations for each of the labels. Additionally, the dataset also provides class-wise summaries of all the tweets. We also perform preliminary experiments on the dataset and show that this is a very challenging dataset for multi-label explainable classification and tweet summarization, as is evident by the moderate scores achieved by some state-of-the-art models. © 2022 ACM.

Automatic Detection and classification of Correct placement of tubes on chest X-rays using deep learning with EfficientNet

In recent years, the rapid growth of data in healthcare has prompted a lot of interest in artificial intelligence (AI). Powerful AI algorithms are essential for extracting information from medical data and assisting clinicians in establishing quick and accurate diagnoses of a variety of ailments. In the current COVID-19 outbreak, critically ill patients were intubated and various medical tubes, including an endotracheal tube (ETT), were implanted to protect the airways. The Nasogastric tube (NGT) is used for feeding, whereas the Central Venous Catheter (CVC) is utilized for a variety of medical operations. The adoption of medical protocols by doctors to ensure proper tube installation is a major issue. Manual examination of CXR pictures takes time and frequently leads to misinterpretation. This research aims to create an Automated Medical Tube Detection System that can detect misplaced tubes from chest x-rays (CXR) using deep learning. As a result, using chest x-rays to detect poorly positioned tubes can save lives. On CXR the proposed CNN-based EfficientNet architecture efficiently detects and classifies incorrectly positioned tubes. After detailed experimentation, we were able to achieve 0.95 average area under the ROC curve (AUC). © 2022 IEEE.