Malta

Homophobic/Transphobic (H/T) content includes hate speeches, discriminatory texts directed at LGBT individuals. Furthermore, social media text data is often code-mixed, which is difficult to analyze, detect H/T content. To address this issue effectively, in this paper, we - team MUCS, describe the ML and TL based models submitted to &quot;Homophobia/Transphobia Detection in social me- dia comments:LT-EDI@EACL 2024&quot; shared task at European Chapter of the Association for Computational Linguistics (EACL) 2024. Three distinct models: i) Homo_Ensemble- a Machine Learning (ML) model trained with syllable TF-IDF of n= (1, 3), ii) Homo_TL - model makes use of oversampling technique used to fine-tune the Transfer Learning (TL) based pretrained Multilingual Bidirectional Encoder Representations from Transformers (mBERT) models, and iii) Homo_mvlearn - a multiview (mvlearn) model which ultizes the probability scores of two models. Among the proposed models Homo_Ensemble obtained macro F1 score of 0.958 securing 4th rank for Telugu, Homo_TL obtained macro F1 scores of 0.493, 0.537, 0.458, 0.948, 0.958 securing 2nd , 2nd , 1st , 1st , and 4th ranks for English, Marathi, Hindi, Kannada and Gujarathi languages respectively. The proposed Homo_mvlearn model obtained macro F1 scores of 0.860, 0.870, and 0.532 se- curing 2nd , 6th , and 2nd ranks for Tamil, Malayalam, and Spanish languages respectively. Tulu Language task consists of Few-Shot Learning (FSL) problem. For the same, proposed two FSL models comprised of Sentence Transformer (ST) model with BERT pre-trained variant models namely, indic_sbert, kannada_sbert. Among these ST model with kannada_sbert ob- tained macro F1 score of 0.620 securing 2nd rank.

EACL 2024

MUCS@LT-EDI-2024: Learning Approaches to Empower Homophobic/Transphobic Comment Identification

h/t content

machine learning

few-shot learning

transfer learning

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.

Homophobic/Transphobic (H/T) content includes hate speeches, discriminatory texts directed at LGBT individuals. Furthermore, social media text data is often code-mixed, which is difficult to analyze, detect H/T content. To address this issue effectively, in this paper, we - team MUCS, describe the ML and TL based models submitted to "Homophobia/Transphobia Detection in social me- dia comments:LT-EDI@EACL 2024" shared task at European Chapter of the Association for Computational Linguistics (EACL) 2024. Three distinct models: i) Homo_Ensemble- a Machine Learning (ML) model trained with syllable TF-IDF of n= (1, 3), ii) Homo_TL - model makes use of oversampling technique used to fine-tune the Transfer Learning (TL) based pretrained Multilingual Bidirectional Encoder Representations from Transformers (mBERT) models, and iii) Homo_mvlearn - a multiview (mvlearn) model which ultizes the probability scores of two models. Among the proposed models Homo_Ensemble obtained macro F1 score of 0.958 securing 4th rank for Telugu, Homo_TL obtained macro F1 scores of 0.493, 0.537, 0.458, 0.948, 0.958 securing 2nd , 2nd , 1st , 1st , and 4th ranks for English, Marathi, Hindi, Kannada and Gujarathi languages respectively. The proposed Homo_mvlearn model obtained macro F1 scores of 0.860, 0.870, and 0.532 se- curing 2nd , 6th , and 2nd ranks for Tamil, Malayalam, and Spanish languages respectively. Tulu Language task consists of Few-Shot Learning (FSL) problem. For the same, proposed two FSL models comprised of Sentence Transformer (ST) model with BERT pre-trained variant models namely, indic_sbert, kannada_sbert. Among these ST model with kannada_sbert ob- tained macro F1 score of 0.620 securing 2nd rank.

poster

The results of the Shared Task on Speech Recognition for Vulnerable Individuals in Tamil (LT-EDI-2024) are discussed in this paper. The goal is to create an automated system for Tamil voice recognition. The older population that speaks Tamil is the source of the dataset used in this task. The proposed ASR system is designed with pre-trained model akashsivanandan/wav2vec2-large-xls-r300m-tamil-colab-final. The Tamil common speech dataset is utilized to fine-tune the pretrained model that powers our system. The suggested system receives the test data that was released from the task; transcriptions are then created for the test samples and delivered to the task. Word Error Rate (WER) is the evaluation statistic used to assess the provided result based on the task. Our Proposed system attained a WER of 29.297%.

ASR TAMIL SSN@ LT-EDI-2024: Automatic Speech Recognition system for Elderly People

Misogynistic memes are a category of memes which contain disrespectful language targeting women on social media platforms. Hence, detecting such memes is necessary in order to maintain a healthy social media environment. To address the challenges of detecting misogynistic memes, "Multitask Meme classification - Unraveling Misogynistic and Trolls in Online Memes: LT-EDI@EACL 2024" shared task organized at European Chapter of the Association for Computational Linguistics (EACL) 2024, invites researchers to develop models to detect misogynistic memes in Tamil and Malayalam. The shared task has two subtasks and in this paper, we - team MUCS, describe the learning models submitted to Task 1 - Identification of Misogynistic Memes in Tamil and Malayalam. As memes represent multi-modal data of image and text, three models: i)~Bidirectional Encoder Representations from Transformers (BERT)+Residual Network (ResNet)-50, ii) Multilingual Representations for Indian Languages (MuRIL)+ResNet-50, and iii) multilingual BERT (mBERT)+ResNet-50, are proposed based on joint representation of text and image, for detecting misogynistic memes in Tamil and Malayalam. Among the proposed models, mBERT+ResNet-50 and MuRIL+ ResNet-50 models obtained macro F1 scores of 0.73 and 0.87 for Tamil and Malayalam datasets respectively securing 1st rank for both the datasets in the shared task.

MUCS@LT-EDI-2024: Exploring Joint Representation for Memes Classification

This paper provides a comprehensive summary of the "Homophobia and Transphobia Detection in Social Media Comments" shared task, which was held at the LT-EDI@EACL 2024. The objective of this task was to develop systems capable of identifying instances of homophobia and transphobia within social media comments. This challenge was extended across ten languages: English, Tamil, Malayalam, Telugu, Kannada, Gujarati, Hindi, Marathi, Spanish, and Tulu. Each comment in the dataset was annotated into three categories. The shared task attracted significant interest, with over 60 teams participating through the CodaLab platform. The submission of prediction from the participants was evaluated with the macro F1 score.

Overview of Third Shared Task on Homophobia and Transphobia Detection in Social Media Comments

The overview of the shared task on speech recognition for vulnerable individuals in Tamil (LT-EDI-2024) is described in this paper. The work comes with a Tamil dataset that was gath- ered from elderly individuals who identify as male, female, or transgender. The audio sam- ples were taken in public places such as mar- ketplaces, vegetable shops, hospitals, etc. The training phase and the testing phase are when the dataset is made available. The task required of the participants was to handle audio signals using various models and techniques, and then turn in their results as transcriptions of the pro- vided test samples. The participant’s results were assessed using WER (Word Error Rate). The transformer-based approach was employed by the participants to achieve automatic voice recognition. This overview paper discusses the findings and various pre-trained transformer- based models that the participants employed

Overview of the Third Shared Task on Speech Recognition for Vulnerable Individuals in Tamil

This paper offers a detailed overview of the first shared task on "Multitask Meme Classification - Unraveling Misogynistic and Trolls in Online Memes," organized as part of the LT-EDI@EACL 2024 conference. The task was set to discern and classify misogynistic content and troll memes within online platforms, focusing specifically on memes in Tamil and Malayalam languages. A total of 52 teams registered for the competition, with four submitting systems for the Tamil meme classification task and three for the Malayalam task. The outcomes of this shared task are significant, providing insights into the current state of misogynistic content in digital memes and highlighting the effectiveness of various computational approaches in identifying such detrimental content. The top-performing model got a macro F1 score of 0.73 in Tamil and 0.87 in Malayalam.

Overview of Shared Task on Multitask Meme classification - Unraveling Misogynistic and Trolls in Online Memes

This paper presents an overview of the shared task on "Caste/Immigration Hate Speech Detection." The shared task is organized as part of LT-EDI@EACL 2024. The system must classify whether the text is caste/immigration hate speech or not. This shared task is conducted in the Tamil language. Fifty-one teams participated in the task in total. Out of that, 15 teams submitted their findings to the task. As far as we know, this is the first shared task that happened on caste/immigration hate speech detection in texts. In this paper, we analyzed and presented all the submissions submitted by the participants. This article includes an in-depth investigation and presentation of the submissions made by all participants.

Overview of Shared Task on Caste/Immigration Hate Speech Detection

Missed recognition of named entities while de-identifying clinical narratives poses a critical challenge in protecting patient-sensitive health information. Mitigating name recognition errors is essential to minimize risk of patient re-identification. In this paper, we emphasize the need for stratified sampling and enhanced contextual considerations concerning Name Tokens using a fine-tuned Longformer BERT model for clinical text de-identifcation. We introduce a Hidden in Plain Sight (HIPS) Markov-based replacement technique for names to mask name recognition misses, revealing a significant reduction in name leakage rates. Our experimental results underscore the impact on addressing name recognition challenges in BERT-based de-identification systems for heightened privacy protection in electronic health records.

Handling Name Errors of a BERT-Based De-Identification System: Insights from Stratified Sampling and Markov-based Pseudonymization

Since medical text cannot be shared easily due to privacy concerns, synthetic data bears much potential for natural language processing applications. In the context of social media and user-generated messages about drug intake and adverse drug effects, this work presents different methods to examine the authenticity of synthetic text. We conclude that the generated tweets are untraceable and show enough authenticity from the medical point of view to be used as a replacement for a real Twitter corpus. However, original data might still be the preferred choice as they contain much more diversity.

Assessing authenticity and anonymity of synthetic user-generated content in the medical domain

In this paper we evaluate two annotation approaches for automatic detection and labelling of personal information in legal texts in relation to the ambiguity of the labels and the homogeneity of the annotations. For this purpose, we built a corpus of 44 case reports from the European Court of Human Rights in Spanish language and we annotated it following two different annotation approaches: automatic projection of the annotations of an existing English corpus, and manual annotation with our reinterpretation of their guidelines. Moreover, we employ Flair on a Named Entity Recognition task to compare its performance in the two annotation schemes.

Automatic Detection and Labelling of Personal Data in Case Reports from the ECHR in Spanish: Evaluation of Two Different Annotation Approaches

Since the announcement of the GDPR, the pseudonymization of legal documents has become a high-priority task in many legal organizations. This means that for making public a document, it is necessary to redact the identity of certain entities, such as witnesses. In this work, we present the first results obtained by PSILENCE, a pseudonymization tool created for redacting semi-automatically international arbitration documents in English. PSILENCE has been built using a Named Entity Recognition (NER) system, along with a Coreference Resolution system. These systems allow us to find the people that we need to redact in a clustered way, but also to propose the same pseudonym throughout one document. This last aspect makes it easier to read and comprehend a redacted legal document. Different experiments were done on four different datasets, one of which was legal, and the results are promising, reaching a Macro F-score of up to 0.72 on the legal dataset.

Downloads

Next from EACL 2024

ASR TAMIL SSN@ LT-EDI-2024: Automatic Speech Recognition system for Elderly People

Similar lecture

Quartet@LT-EDI 2024: Support Vector Machine based approach for Homophobia/Transphobia Detection in social media comments

Stay up to date with the latest Underline news!

PRODUCT

COMPANY

RESOURCES