Canada

In this paper, we highlight our approach for the &quot;WASSA 2023 Empathy and Emotion Shared Task&quot;. We present our approach for track 3 of the shared task which aims to identify emotions from text. Each sample in the dataset has one or more labels making it a multi-label classification task. We compared multiple transformer-based models by fine-tuning them for multi-label classification. Oversampling was used to overcome the class imbalance in the dataset. Ensembling techniques were used to improve the performance of the system. We obtained a macro F1-score of 0.5649 using XLNet on the test dataset in the official phase and secured rank 6 on the official leaderboard. During the post-competition phase, a threshold-based voting mechanism was performed on three models (Longformer, BERT, BigBird) that yielded the highest overall macro F1-score of 0.6605.

ACL 2023

Converge at WASSA 2023 Empathy, Emotion and Personality Shared Task: A Transformer-based Approach for Multi-Label Emotion Classification

mutli-label text classification

ensemble

transformers

### 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.

In this paper, we highlight our approach for the "WASSA 2023 Empathy and Emotion Shared Task". We present our approach for track 3 of the shared task which aims to identify emotions from text. Each sample in the dataset has one or more labels making it a multi-label classification task. We compared multiple transformer-based models by fine-tuning them for multi-label classification. Oversampling was used to overcome the class imbalance in the dataset. Ensembling techniques were used to improve the performance of the system. We obtained a macro F1-score of 0.5649 using XLNet on the test dataset in the official phase and secured rank 6 on the official leaderboard. During the post-competition phase, a threshold-based voting mechanism was performed on three models (Longformer, BERT, BigBird) that yielded the highest overall macro F1-score of 0.6605.

workshop paper

This paper presents our approach for the WASSA 2023 Empathy, Emotion and Personality Shared Task. Empathy and distress are human feelings that are implicitly expressed in natural discourses. Empathy and distress detection are crucial challenges in Natural Language Processing that can aid our understanding of conversations. The provided dataset consists of several long-text examples in the English language, with each example associated with a numeric score for empathy and distress. We experiment with several BERT-based models as a part of our approach. We also try various ensemble methods. Our final submission has a Pearson's r score of 0.346, placing us third in the empathy and distress detection subtask.

PICT-CLRL at WASSA 2023 Empathy, Emotion and Personality Shared Task: Empathy and Distress Detection using Ensembles of Transformer Models

This paper describes the approach that we used to take part in the multi-label multi- class emotion classification as Track 3 of the WASSA 2023 Empathy, Emotion and Personality Shared Task at ACL 2023. The overall goal of this track is to build models that can predict 8 classes (7 emotions + neutral) based on short English essays written in response to news article that talked about events perceived as harmful to people. We used OpenAI generative pretrained transformers with full-scale APIs for the emotion prediction task by fine-tuning a GPT-3 model and doing prompt engineering for zero-shot / few-shot learning with ChatGPT and GPT-4 models based on multiple experiments on the dev set. The most efficient method was fine-tuning a GPT-3 model which allowed us to beat our baseline character-based XGBoost Classifier and rank 2nd among all other participants by achieving a macro F1 score of 0.65 and a micro F1 score of 0.7 on the final blind test set.


Team Bias Busters at WASSA 2023 Empathy, Emotion and Personality Shared Task: Emotion Detection with Generative Pretrained Transformers

This presentation introduces the participation of team HIT-SCIR to the WASSA 2023 Shared Task on Empathy Detection and Emotion Classification and Personality Detection in Interactions. We focus on three tracks: Track 1 (Empathy and Emotion Prediction in Conversations, CONV), Track 2 (Empathy Prediction, EMP) and Track 3 (Emotion Classification, EMO), and designed three different models to address them separately. For Track 1, we designed a direct finetuning DeBERTa model for three regression tasks at the utterance-level. For Track 2, we designed a multi-task learning RoBERTa model for two regression tasks at the essay-level. For Track 3, we designed a RoBERTa model with data augmentation for the classification task at the essay-level. Finally, our team ranked 1st in the Track 1 (CONV), 5th in the Track 2 (EMP) and 3rd in the Track 3 (EMO) in the evaluation phase.

HIT-SCIR at WASSA 2023: Empathy and Emotion Analysis at the Utterance-Level and the Essay-Level

This Shared Task, “Emotion Classification from essays written in reaction to news articles,” is about predicting emotions from essay-level texts. The target labels consist of thirty-one categories of emotions, including individual and mixed sets of emotion categories, making it a multiclass classification problem. Our system based on transformer architecture (BERT-Base) scored the tenth rank out of 13 reported teams. Our experiments conclude that although FLAIR are powerful word representations built to capture out-of-vocabulary words, they are less effective than contextual embeddings when used for small and imbalanced datasets.

VISU at WASSA 2023 Shared Task: Detecting Emotions in Reaction to News Stories Using Transformers and Stacked Embeddings

Code-mixing refers to the phenomenon of using two or more languages interchangeably within a speech or discourse context. This practice is particularly prevalent on social media platforms, and determining the embedded affects in a code-mixed sentence remains as a challenging problem. In this submission we describe our system for WASSA 2023 Shared Task on Emotion Detection in English-Urdu code-mixed text. In our system we implement a multiclass emotion detection model with label space of 11 emotions. Samples are code-mixed English-Urdu text, where Urdu is written in romanised form. Our submission is limited to one of the subtasks - Multi Class classification and we leverage transformer-based Multilingual Large Language Models (MLLMs), XLM-RoBERTa and Indic-BERT. We fine-tune
MLLMs on the released data splits, with and without pre-processing steps (translation to english), for classifying texts into the appropriate emotion category. Our methods did not surpass the baseline, and our submission is ranked sixth
overall.

PrecogIIITH@WASSA2023: Emotion Detection for Urdu-English Code-mixed Text

In this era of digital communication and social media, texting and chatting among individuals occur mainly through code-mixed or Romanized versions of the native language prevalent in the region. The presence of Romanized and code-mixed language develops the need to build NLP systems in these domains to leverage the digital content for various use cases. This paper describes our contribution to the subtask MCEC of the shared task WASSA 2023:Shared Task on Multi-Label and Multi-Class Emotion Classification on Code-Mixed Text Messages. We explore how one can build sentence transformers models for low-resource languages using unsupervised data by leveraging contrastive learning techniques described in the SIMCSE paper and using the sentence transformer developed to build classification models using the SetFit approach. Additionally, we’ll publish our code and models on GitHub and HuggingFace, two open-source hosting services.

BpHigh at WASSA 2023: Using Contrastive Learning to build Sentence Transformer models for Multi-Class Emotion Classification in Code-mixed Urdu

This paper describes the approach that we utilized to participate in the shared task for multi-label and multi-class emotion classification organized as part of WASSA 2023 at ACL 2023. The objective was to build models that can predict 11 classes of emotions, or the lack thereof (neutral class) based on code-mixed Roman Urdu and English SMS text messages. We participated in Track 2 of this task - multi-class emotion classification (MCEC). We used generative pretrained transformers, namely ChatGPT because it has a commercially available full-scale API, for the emotion detection task by leveraging the prompt engineering and zero-shot / few-shot learning methodologies based on multiple experiments on the dev set. Although this was the first time we used a GPT model for the purpose, this approach allowed us to beat our own baseline character-based XGBClassifier, as well as the baseline model trained by the organizers (bert- base-multilingual-cased). We ranked 4th and achieved the macro F1 score of 0.7038 and the accuracy of 0.7313 on the blind test set.

Generative Pretrained Transformers for Emotion Detection in a Code-Switching Setting

Assessing the capacity of numerical understanding of vision-and-language models over images and texts is crucial for real vision-and-language applications, such as systems for automated medical image analysis. We provide a visual reasoning dataset focusing on numerical understanding in the medical domain. The experiments using our dataset show that current vision-and-language models fail to perform numerical inference in the medical domain. However, the data augmentation with only a small amount of our dataset improves the model performance, while maintaining the performance in the general domain.

Medical Visual Textual Entailment for Numerical Understanding of Vision-and-Language Models

Valuable datasets that contain sensitive information are not shared due to privacy and copyright concerns. This hinders progress in many areas and prevents the use of machine learning solutions to solve relevant tasks. One possible solution is sharing models that are trained on such datasets. However, this is also associated with potential privacy risks due to data extraction attacks. In this work, we propose a solution based on sharing parts of the model's parameters, and using a proxy dataset for complimentary knowledge transfer. Our experiments show encouraging results, and reduced risk to potential training data identification attacks. We present a viable solution to sharing knowledge with data-disadvantaged parties, that do not have the resources to produce high-quality data, with reduced privacy risks to the sharing parties. We make our code publicly available.

Privacy-Preserving Knowledge Transfer through Partial Parameter Sharing

Alzheimer's Disease (AD) is a neurodegenerative disorder that affects cognitive abilities and memory, especially in older adults. One of the challenges of AD is that it can be difficult to diagnose in its early stages. However, recent research has shown that changes in language, including speech decline and difficulty in processing information, can be important indicators of AD and may help with early detection. Hence, the speech narratives of the patients can be useful in diagnosing the early stages of Alzheimer's disease. While the previous works have presented the potential of using speech narratives to diagnose AD in high-resource languages, this work explores the possibility of using a low-resourced language, i.e., Hindi language, to diagnose AD. In this paper, we present a dataset specifically for analyzing AD in the Hindi language, along with experimental results using various state-of-the-art algorithms to assess the diagnostic potential of speech narratives in Hindi. Our analysis suggests that speech narratives in the Hindi language have the potential to aid in the diagnosis of AD. Our dataset and code are made publicly available at https://github.com/rkritesh210/DementiaBankHindi. 


Downloads

Next from ACL 2023

PICT-CLRL at WASSA 2023 Empathy, Emotion and Personality Shared Task: Empathy and Distress Detection using Ensembles of Transformer Models

Similar lecture

AdityaPatkar at WASSA 2023 Empathy, Emotion, and Personality Shared Task: RoBERTa-Based Emotion Classification of Essays, Improving Performance on Imbalanced Data

Stay up to date with the latest Underline news!

PRESENTATIONS

CONFERENCES

COMPANY

RESOURCES