First Workshop on ​Combating On​line Ho​st​ile Posts in ​Regional L​anguages dur​ing Emerge​ncy Si​tuation

Collocated with AAAI 2021

8th February, 2021



The increasing accessibility of the Internet has dramatically changed the way we consume information. The ease of social media usage not only encourages individuals to freely express their opinion (freedom of speech) but also provides content polluters with ecosystems to spread hostile posts (hate speech, fake news, cyberbullying, etc.). Such hostile activities are expected to increase manifold during emergencies such as the 2021 US presidential election, COVID-19 pandemic spreading. Most of such hostile posts are written in regional languages, and therefore can easily evade online surveillance engines, the majority of which are trained on the posts written in resource-rich languages such as English and Chinese. Therefore, regions such as Asia, Africa, South America, where low-resource ​regional languages are used for day-to-day communication, suffer due to the lack of tools, benchmark datasets and learning techniques. Other developing countries such as Italy, Spain, where the used languages (pseudo-low-resource) are not as equipped with sophisticated computational resources as English, might also be facing the same issues.
CONSTRAINT’21 will encourage researchers from interdisciplinary domains working on multilingual social media analytics to think beyond the conventional way of combating online hostile posts. The workshop will emphasise on three major points:

  1. Regional language: The offensive posts under inspection may be written in low-resource regional languages (e.g., Tamil, Urdu, Bangali, Polish, Czech, Lithuanian, etc.).
  2. Emergency situation: ​The proposed solutions should be able to tackle misinformation during emergency situations where due to the lack of enough historical data, learning models need to adopt additional intelligence to handle emerging and novel posts.
  3. Early detection: ​Since the effect of misinformation during emergency situations is highly detrimental for society (e.g., health-related misadvice during a pandemic may take human’s life), we encourage the solutions to be able to detect such hostile posts as early as possible after their appearance on social media.



  • Topics of Interests: We invite the submission of high-quality manuscripts reporting relevant research in the area of collecting, managing, mining, and understanding hostile data from social media platforms. Topics of interest include, but are not limited to:
    • Fake news detection in regional languages
    • Hate speech detection in regional languages
    • Evolution of fake news and hate speech
    • Analyzing user behavior for hostile post propagation
    • Real-world tool development for combating hostile posts
    • Psychological study of the spreaders of hostile posts
    • Hate speech normalization
    • Information extraction, ontology design and knowledge graph for combating hostile posts
    • Early detection for hostile posts
    • Design lightweight tools with less data for hostile post detection
    • Code-mixed and code-switched hostile post analysis
    • Open benchmark and dashboard related to regional hostile posts
    • Specific case studies and surveys related to hostile posts
    • Claim detection and verification related to misinformation
    • Fact-check worthiness of misinformation
    • Cross-region language analysis for hostile posts
    • Computational social science analysis for hostile posts
    • Network analysis for fake news spreading and evolution

  • Submission Instructions:
    • Regular papers: Regular papers (maximum 12 pages including references) should be prepared in English and follow the Springer CCIS template, downloadable from here. Papers will be published in Springer CCIS.
    • Previously rejected papers: If your paper has been rejected from EMNLP/NeurIPS or AAAI 2021 regular track, you can also submit the peer reviews, ratings and 1 page cover letter of the changes made to address those comments. Note that these things should be in a separate file and NOT appended to the submitted manuscript. These papers will be reviewed separately.
    • Extended abstracts: We also invite extended abstract submission. The length of extended abstracts is 2 pages, including figures and tables (references not included). Note that there will be no published proceedings for extended abstracts. Work in progress or accepted papers are encouraged.
    • All papers must be submitted via our EasyChair submission page. Regular papers will go through a double-blind peer-review process. Extended abstracts may be either single blind (i.e., reviewers are blind, authors have names on submission) or double blind (i.e., authors and reviewers are blind). Only manuscripts in PDF or Microsoft Word format will be accepted.

  • Important Dates:
    • Dec 5, 2020: Research papers and extended abstracts due at 11:59 PM UTC-12
    • Dec 20, 2020: Notification of papers due at 11:59 PM UTC-12
    • Dec 28, 2020: Camera ready submission due of accepted papers at 11:59 PM UTC-12
    • Feb 8, 2021: CONSTRAINT workshop


  • Task: We are planning to release two types of datasets:
    • COVID19 Fake news detection (English)
    • Hostility detection (Hindi)
    For more details and updates please visit

  • Important Dates:
    • October 1, 2020: Release of the training set
    • December 1, 2020: Release of the test set
    • December 10, 2020: Deadline for submitting the final results
    • December 12, 2020: Announcement of the results
    • December 25, 2020: System paper submission deadline (All teams are invited to submit a paper)


Selected workshop papers will be invited for an extension to be considered in a special issue of Neurocomputing journal (Impact factor: 4.438)


  • Ashwin Singh and Rudraroop Ray. Identifying Offensive Content in Social Media Post
  • Omar Sharif and Mohammed Moshiul Hoque. Identification and Classification of Textual Aggression in Social Media: Resource Creation and Evaluation
  • Anna Glazkova, Maksim Glazkov and Timofey Trifonov. g2tmn at Constraint@AAAI2021: Exploiting CT-BERT and Ensembling Learning for COVID-19 Fake News Detection
  • Ipek Baris and Zeyd Boukhers. ECOL: Early Detection of COVID Lies UsingContent, Prior Knowledge and Source Information
  • Tathagata Raha, Sayar Ghosh Roy, Ujwal Narayan, Zubair Abid and Vasudeva Varma. Task Adaptive Pretraining of Transformers for Hostility Detection
  • Parth Patwa, Shivam Sharma, Srinivas Pykl, Vineeth Guptha, Gitanjali Kumari, Md Shad Akhtar, Asif Ekbal, Amitava Das and Tanmoy Chakraborty. Fighting an Infodemic: COVID-19 Fake News Dataset
  • Yejin Bang, Etsuko Ishii, Samuel Cahyawijaya, Ziwei Ji and Pascale Fung. Model Generalization on COVID-19 Fake News Detection
  • Siyao Zhou, Rui Fu and Jie Li. Fake news and hostile post detection using an ensemble learning model
  • Ben Chen, Bin Chen, Dehong Gao, Qijin Chen, Xiaonan Meng, Chengfu Huo, Yang Zhou and Weijun Ren. Transfomrer-based Language Model Fine-tuning Methods for COVID-19 Fake News Detection
  • Arkadipta De, Venkatesh E, Kaushal Kumar Maurya and Maunendra Sankar Desarkar. Coarse and Fine-Grained Hostility Detection in Hindi Posts using Fine Tuned Multilingual Embeddings
  • Shreyash Arya and Hridoy Sankar Dutta. Revealing the Blackmarket Retweet Game: A Hybrid Approach
  • Mohammed Azhan and Mohammad Ahmad. LaDiff ULMFiT: A Layer Differentiated training approach for ULMFiT
  • Renyuan Liu and Xiaobing Zhou. Extracting latent information from datasets in The CONSTRAINT-2020 shared task on the hostile post detection
  • Boshko Koloski, Timen Stepišnik-Perdih and Blaž Škrlj. Identification of COVID-19 related Fake News via Neural Stacking
  • Anand Zutshi and Aman Raj. Tackling the infodemic : Analysis using Transformer based model
  • Varad Bhatnagar, Prince Kumar, Sairam Moghili and Pushpak Bhattacharyya. Divide and Conquer: An Ensemble Approach for Hostile Post Detection in Hindi
  • Xiangyang Li, Yu Xia, Long Xiang, Zheng Li and Sujian Li. Exploring Text-transformers in AAAI 2021 Shared Task: COVID-19 Fake News Detection in English
  • Siva Sai, Alfred William Jacob, Sakshi Kalra and Yashvardhan Sharma. Stacked embeddings and multiple fine-tuned XLM-RoBERTa models for Enhanced hostility identification
  • Apurva Wani, Isha Joshi, Snehal Khandve, Vedangi Wagh and Raviraj Joshi. Evaluating Deep Learning Approaches for COVID19 Fake News Detection
  • Parth Patwa, Mohit Bhardwaj, Vineeth Guptha, Gitanjali Kumari, Shivam Sharma, Srinivas PYKL, Amitava Das, Asif Ekbal, Shad Akhtar, Tanmoy Chakraborty. Overview of CONSTRAINT 2021 Shared Tasks: Detecting English COVID-19 Fake News and Hindi Hostile Posts
  • Sourya Dipta Das, Ayan Basak and Saikat Dutta. A Heuristic-driven Ensemble Framework for COVID-19 Fake News Detection
  • Ojasv Kamal, Adarsh Kumar and Tejas Vaidhya. Hostility Detection in Hindi leveraging Pre-Trained Language Models
  • Venktesh V, Akansha Gautam and Sarah Masud. Fake News Detection System using XLNet model with Topic Distributions: CONSTRAINT@AAAI2021 Shared Task



Workshop paper presentation (2)

Panel Discussion


Shared task overview

Workshop paper presentation (2)

Shared task papers (9)


Shared task papers (9)

Best paper announcement and vote of thanks




Subscribe to our mailing list      Click Here

Follow Us