Schedule

All times are local Dublin (UTC/GMT +1 hour)

Time Programmed Session
09:00 - 09:10 Opening
09:10 - 10:30 Oral presentation session 1
10:30 - 11:00 Coffee Break
11:00 - 12:00 Shared Task Session
12:00 - 13:00 Invited Talk: Dirk Hovy - Mind the Gaps and Normal Accidents
13:00 - 14:00 Lunch Break
14:00 - 15:00 Invited Talk: Rada Mihalcea - Beneath the Tip of the Iceberg: Current Challenges and New Directions in Sentiment Analysis Research
15:00 - 15:30 Coffee break
15:30 - 16:15 In-person poster session
16:15 - 17:15 Oral presentation session 2
17:15 - 18:00 Virtual poster session

Oral presentation session 1

  • 9:10 - 9:30 - Assessment of Massively Multilingual Sentiment Classifiers. Krzysztof Rajda, Lukasz Augustyniak, Piotr Gramacki, Marcin Gruza, Szymon Woźniak, Tomasz Jan Kajdanowicz. In-person
  • 9:30 - 9:50 - English-Malay Word Embeddings Alignment for Cross-lingual Emotion Classification with Hierarchical Attention Network. Ying Hao Lim, Jasy Suet Yan Liew. Online
  • 9:50 - 10:10 - Uncertainty Regularized Multi-Task Learning. Kourosh Meshgi, Maryam Sadat Mirzaei, Satoshi Sekine.Online
  • 10:10 - 10:30 - Improving Social Meaning Detection with Pragmatic Masking and Surrogate Fine-Tuning. Chiyu Zhang, Muhammad Abdul-Mageed. Online

Shared Task

  • Shared Task Overview Valentin Barriere, Shabnam Tafreshi, João Sedoc, Sawsan Alqahtani
  • Empathy and Distress Prediction using Transformer Multi-output Regression and Emotion Analysis with an Ensemble of Supervised and Zero-Shot Learning Models. Flor Miriam Plaza del Arco,Jaime Collado-Montañez, L. Alfonso Ureña, María-Teresa Martín-Valdivia.
  • Continuing Pre-trained Model with Multiple Training Strategies for Emotional Classification. Bin Li, Yixuan Weng, qiya song, Bin Sun, Shutao Li.
  • Prompt-based Pre-trained Model for Personality and Interpersonal Reactivity Prediction. Bin Li, Yixuan Weng, qiya song, Fuyan Ma, Bin Sun, Shutao Li.

Invited Talk: Dirk Hovy

Mind the Gaps and Normal Accidents

NLP is now stable enough to be used in production systems, and will soon become even more pervasive. However, even today’s systems are already highly complex and unpredictable. As they become more ubiquitous, different algorithms will interact with each other directly leading to tightly coupled systems whose capacity to cause harm we will be unable to predict.

In his book Normal Accidents, the sociologist Charles Perrow proposed a framework to analyze technologies and their risks according to their complexity and the interdependence of their components. He showed that accidents were nigh on unavoidable due to those two features.

We apply Perrow’s framework to NLP and argue that under the current paradigm, “normal accidents” are built into the system and only a matter of time, and that some issues in current NLP practice that aid this development: the early adoption of methods without sufficient understanding or analysis; the preference for computational methods regardless of risks associated with their limitations; the dangers of unexplainable methods.

If these issues are not addressed, we risk a loss of reproducibility, reputability, and subsequently public trust in our field. Together, these factors can help us better understand the risks of complex NLP systems, and to make our systems safer and more reliable.

Bio

Dirk Hovy is an associate professor of Computer Science in the department of marketing at Bocconi University and director of the Data and Marketing Insights research unit at the Bocconi Center for Data Science and Analytics. He did his PhD at the University of Southern California in Los Angeles, working as research assistant at the Information Sciences Institute.

His research focuses on computational social science and he is interested in what language can tell us about society, and what computers can tell us about language, as well as ethics in NLP and demographic biases.

Invited Talk: Rada Mihalcea

Beneath the Tip of the Iceberg: Current Challenges and New Directions in Sentiment Analysis Research

Sentiment analysis as a field has come a long way since it was first introduced as a task nearly 20 years ago. It has widespread commercial applications in various domains like marketing, risk management, market research, and politics, to name a few. Given its saturation in specific subtasks – such as sentiment polarity classification – and datasets, there is an underlying perception that this field has reached its maturity. In this article, we discuss this perception by pointing out the shortcomings and under-explored, yet key aspects of this field that are necessary to attain true sentiment understanding. We analyze the significant leaps responsible for its current relevance. Further, we attempt to chart a possible course for this field that covers many overlooked and unanswered questions.

Joint work with Soujanya Poria, Devamanyu Hazarika, Navonil Majumder

Bio

Rada Mihalcea is the Janice M. Jenkins Collegiate Professor of Computer Science and Engineering at the University of Michigan and the Director of the Michigan Artificial Intelligence Lab. Her research interests are in computational linguistics, with a focus on lexical semantics, multilingual natural language processing, and computational social sciences. She serves or has served on the editorial boards of the Journals of Computational Linguistics, Language Resources and Evaluations, Natural Language Engineering, Journal of Artificial Intelligence Research, IEEE Transactions on Affective Computing, and Transactions of the Association for Computational Linguistics. She was a program co-chair for EMNLP 2009 and ACL 2011, and a general chair for NAACL 2015 and *SEM 2019. She currently serves as ACL Past President. She is the recipient of a Presidential Early Career Award for Scientists and Engineers awarded by President Obama (2009), an ACM Fellow (2019) and a AAAI Fellow (2021). In 2013, she was made an honorary citizen of her hometown of Cluj-Napoca, Romania.

In-person poster session

  • On the Complementarity of Images and Text for the Expression of Emotions in Social Media. Anna Khlyzova, Carina Silberer, Roman Klinger.
  • Multiplex Anti-Asian Sentiment before and during the Pandemic: Introducing New Datasets from Twitter Mining. Hao Lin, Pradeep Kumar Nalluri, Lantian Li, Yifan Sun, Yongjun Zhang.
  • SentEMO: A Multilingual Adaptive Platform for Aspect-based Sentiment and Emotion Analysis. Ellen De Geyndt, Orphee De Clercq, Cynthia Van Hee, Els Lefever, Pranaydeep Singh, Olivier Parent, Veronique Hoste.
  • Irony Detection for Dutch: a Venture into the Implicit. Aaron Maladry, Els Lefever, Cynthia Van Hee, Veronique Hoste.
  • Items from Psychometric Tests as Training Data for Personality Profiling Models of Twitter Users. Anne Kreuter, Kai Sassenberg, Roman Klinger.
  • Polite Task-oriented Dialog Agents: To Generate or to Rewrite? Diogo Silva, David Semedo, Joao Magalhaes.

Oral presentation session 2

  • 16:15 - 16:35 - Distinguishing In-Groups and Onlookers by Language Use. Joshua R Minot, Milo Z Trujillo, Samuel F Rosenblatt, Guillermo de Anda-Jáuregui, Emily Moog, Allison M. Roth, Briane Paul Samson, Laurent Hébert-Dufresne. In-person
  • 16:35 - 16:55 - “splink” is happy and “phrouth” is scary: Emotion Intensity Analysis for Nonsense Words. Valentino Sabbatino, Enrica Troiano, Antje Schweitzer, Roman Klinger. In-person
  • 16:55 - 17:15 - Can Emotion Carriers Explain Automatic Sentiment Prediction? A Study on Personal Narratives Seyed Mahed Mousavi, Gabriel Roccabruna, Aniruddha Tammewar, Steve Azzolin, Giuseppe Riccardi. In-person

Virtual poster session

  • On the Complementarity of Images and Text for the Expression of Emotions in Social Media. Anna Khlyzova, Carina Silberer, Roman Klinger.
  • Multiplex Anti-Asian Sentiment before and during the Pandemic: Introducing New Datasets from Twitter Mining. Hao Lin, Pradeep Kumar Nalluri, Lantian Li, Yifan Sun, Yongjun Zhang.
  • SentEMO: A Multilingual Adaptive Platform for Aspect-based Sentiment and Emotion Analysis. Ellen De Geyndt, Orphee De Clercq, Cynthia Van Hee, Els Lefever, Pranaydeep Singh, Olivier Parent, Veronique Hoste.
  • Irony Detection for Dutch: a Venture into the Implicit. Aaron Maladry, Els Lefever, Cynthia Van Hee, Veronique Hoste.
  • Items from Psychometric Tests as Training Data for Personality Profiling Models of Twitter Users. Anne Kreuter, Kai Sassenberg, Roman Klinger.
  • Domain-Aware Contrastive Knowledge Transfer for Multi-domain Imbalanced Data. Zixuan Ke, Mohammad Kachuee, Sungjin Lee.
  • Infusing Knowledge from Wikipedia to Enhance Stance Detection. Zihao He, Negar Mokhberian, Kristina Lerman.
  • Understanding BERT’s Mood: The Role of Contextual-Embeddings as User-Representations for Depression Assessment. Matthew Matero, Albert Hung, H. Schwartz.
  • Emotion Analysis of Writers and Readers of Japanese Tweets on Vaccinations. Patrick John Co Ramos, Kiki Ferawati, Kongmeng Liew, Eiji Aramaki, Shoko Wakamiya.
  • Opinion-based Relational Pivoting for Cross-domain Aspect Term Extraction. Ayal Klein, Oren Pereg, Daniel Korat, Vasudev Lal, Moshe Wasserblat, Ido Dagan.
  • Pushing on Personality Detection from Verbal Behavior: A Transformer Meets Text Contours of Psycholinguistic Features. Elma Kerz, Yu Qiao, Sourabh Zanwar, Daniel Wiechmann.
  • XLM-EMO: Multilingual Emotion Prediction in Social Media Text. Federico Bianchi, Debora Nozza, Dirk Hovy.
  • Evaluating Content Features and Classification Methods for Helpfulness Prediction of Online Reviews: Establishing a Benchmark for Portuguese. Rogério Figueredo Sousa, Thiago A. S. Pardo.
  • Tagging Without Rewriting: A Probabilistic Model for Unpaired Sentiment and Style Transfer. Shuo Yang.
  • NLPOP: a Dataset for Popularity Prediction of Promoted NLP Research on Twitter. Leo Obadić, Martin Tutek, Jan Snajder.
  • Polite Task-oriented Dialog Agents: To Generate or to Rewrite? Diogo Silva, David Semedo, Joao Magalhaes.
  • Leveraging Emotion-Specific features to improve Transformer performance for Emotion Classification. Shaily Desai, Atharva M Kshirsagar, Aditi Sidnerlikar, Nikhil Vinod Khodake, Manisha Vinayak Marathe.
  • Transformer-based Architecture for Empathy Prediction and Emotion Classification. Himil Vasava, Pramegh Prashant Uikey, Gaurav Ramesh Wasnik, Raksha Sharma.
  • Team IITP-AINLPML at WASSA 2022: Empathy Detection, Emotion Classification and Personality Detection. Soumitra Ghosh, Dhirendra Kumar Maurya, Asif Ekbal, Pushpak Bhattacharyya.
  • SURREY-CTS-NLP at WASSA2022: An Experiment of Discourse and Sentiment Analysis for the Prediction of Empathy, Distress and Emotion. Shenbin Qian, Constantin Orasan, Diptesh Kanojia, Hadeel Saadany, Félix do Carmo.
  • Transformer based ensemble for emotion detection. Aditya Kane, Shantanu Patankar, Sahil Khose, Neeraja Kirtane.
  • IUCL at WASSA 2022 Shared Task: A Text-only Approach to Empathy and Emotion Detection. Yue Chen, Yingnan Ju, Sandra Kübler.
  • An Ensemble Approach to Detect Emotions at an Essay Level. Himanshu Maheshwari, Vasudeva Varma.