Welcome to ICON 2021!

Due to COVID-19 Pandemic, ICON 2021 will be held virtually

Latest Updates



Keynote Talks

Prof. Josef van Genabith

Title of the talk: Phrases and Self-Supervision in Neural Machine Translation     

ABSRTACT: In the talk I cover parts of two of the research strands of our Multilinguality and Languages Technology (MLT) lab at DFKI: improving transformer models and addressing scarcity or lack of bitext training data. Transformer models are the workhorses of modern language technologies. In the talk I show how source phrases can be successfully integrated in the transformer model (Xu et al. ACL 2020). This holds promise, as phrases may improve translation quality and the treatment of non-local phenomena, but is challenging as a naïve integration of phrases (as explicit strings and their embeddings) would explode the vocabulary size. I show how phrases can be computed “on the fly” and integrated into the transformer model, with significant improvements in translation quality (BLEU) and the treatment of long-distance phenomena. Given enough parallel bitext training data, neural machine translation can produce close to human professional translation performance. Unfortunately, parallel data is scarce or unavailable for most language pairs. I present Self-Supervised NMT (SSNMT: Ruiter et al. ACL 2019, EMNLP 2020, MTSummit 2021) that jointly learns (i) translation and (ii) finding its own supervision signal at the same time in a virtuous loop, using comparable (not parallel) training data. The talk is based on joint work with our students and colleagues: Dana Ruiter, Dietrich Klakow, Josef van Genabith, Cristina España-Bonet: Integrating Unsupervised Data Generation into Self-Supervised Neural Machine Translation for Low-Resource Languages. MTSummit (1) 2021: 76-91 Hongfei Xu, Josef van Genabith, Deyi Xiong, Qiuhui Liu, Jingyi Zhang: Learning Source Phrase Representations for Neural Machine Translation. ACL 2020: 386-396 Dana Ruiter, Josef van Genabith, Cristina España-Bonet: Self-Induced Curriculum Learning in Self-Supervised Neural Machine Translation. EMNLP (1) 2020: 2560-2571 Dana Ruiter, Cristina España-Bonet, Josef van Genabith: Self-Supervised Neural Machine Translation. ACL (1) 2019: 1828-1834.

SHORT BIO: Prof. Josef van Genabith is one of the Scientific Directors of the DFKI, the German Research Centre for Artificial Intelligence, where he heads the Multilingual Language Technologies (MLT) Lab. He is Full Professor at Saarland University where he holds the Chair of Translation Oriented Language Technologies. He was the founding Director of CNGL, the Centre for Next Generation Localisation (now ADAPT), Director of the National Center for Language Technology (NCLT), and an Associate Professor, Senior Lecturer and Lecturer in the School of Computing at Dublin City University (DCU), Ireland. He worked as a postdoctoral researcher at IMS, University of Stuttgart, Germany, and obtained an MA and a PhD from the University of Essex, U.K. His first degree is in Electronic Engineering and English from RWTH Aachen, Germany.

Prof. Philip Resnik

Title of the talk: Mental Health as an Application Area for Natural Language Processing: Prospects and Challenges     

ABSRTACT: The cost of mental illness is staggering. In purely economic terms, the World Economic Forum estimates that the cumulative worldwide cost of mental illness from 2011 through 2030 is expected to be US$16T, more than cancer, chronic respiratory diseases, and diabetes combined. The human cost is even worse. Globally, more than 264 million people are affected by depression. Over a million people die by suicide each year; that's one every 40 seconds on average, nearly 2% of worldwide deaths. For countries with lower or middle income, 75-85% of individuals with a mental health disorder go untreated. Compounding these existing problems is an “echo pandemic” emerging in the wake of COVID-19, as people have struggled with isolation, stress, and sustained disruptions of day to day life.
Language is a window into mental state. Indeed, what clinicians do when they are assessing and treating patients is largely a language analysis task -- a search for relevant, predictive signal. So, what might language technology be able to do to help? In this talk I will discuss mental health as an application area for natural language processing, taking a high level view and looking at both the promise and the challenges for NLP approaches. This will include work using predictive modeling, the predominant approach, but I will also advocate for a shift in emphasis from binary classification to prioritization under limited-resource constraints, and I will discuss secure data enclaves as a path forward for making technological collaborations possible with sensitive data.

SHORT BIO: Prof. Philip Resnik is Professor at University of Maryland in the Department of Linguistics and Institute for Advanced Computer Studies. He earned his bachelor's degree in Computer Science at Harvard and his PhD in Computer and Information Science at the University of Pennsylvania. Prior to joining UMD, he was an associate scientist at BBN, a graduate summer intern at IBM T.J. Watson Research Center (subsequently awarded an IBM Graduate Fellowship) while at UPenn, and a research scientist at Sun Microsystems Laboratories. Resnik was named a Fellow of the Association for Computational Linguistics in 2020. Outside his academic research, Resnik has been a technical co-founder of CodeRyte (NLP for electronic health records, acquired by 3M in 2012), and he is an advisor to FiscalNote (machine learning and analytics for government relations), SoloSegment (web site search and content optimization), Converseon (social strategy and analytics), and the Coleridge Initiatlve (nonprofit focused on effective use of data for public decision-making). Resnik's most recent research focus has been in computational social science, with an emphasis on connecting the signal available in people's language use with underlying mental state -- this has applications in computational political science, particularly in connection with ideology and framing, and in mental health, focusing on the ways that linguistic behavior may help to identify and monitor depression, suicidality, and schizophrenia. He has also been ramping up a new area of research in the computational cognitive neuroscience of language, developing models using brain imaging data with a focus on the role of contextual prediction in sentence understanding.

Prof. Rada Mihalcea

Title of the talk: Human-centered Natural Language Processing     

ABSTRACT: The typical approach in natural language processing is to use one-size-fits-all representations, obtained from training one model on very large text collections. While this approach is effective for those people whose language style is well represented in the data, it fails to account for variations between people. In this talk, I will challenge the one-size-fits-all assumption, and show that we can identify words that are used in significantly different ways by speakers from different culture, and we can effectively use information about the people behind the words to build better natural language processing models for different tasks, including word associations, language models, and humor generation.

SHORT 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 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.

Associate Prof. Louis-Philippe Morency

Title of the talk: Multimodal AI: Understanding Human Behaviors     

ABSRTACT: Human face-to-face communication is a little like a dance, in that participants continuously adjust their behaviors based on verbal and nonverbal cues from the social context. Today's computers and interactive devices are still lacking many of these human-like abilities to hold fluid and natural interactions. Leveraging recent advances in machine learning, audio-visual signal processing and computational linguistic, my research focuses on creating computational technologies able to analyze, recognize and predict human subtle communicative behaviors in social context. Central to this research effort is the introduction of new computational models able to learn the temporal and fine-grained latent dependencies across behaviors, modalities and interlocutors. In this talk, I will present some of our recent achievements in multimodal machine learning, addressing five core challenges: representation, alignment, fusion, translation and co-learning.

SHORT BIO: Louis-Philippe Morency is currently a tenure-track Faculty at CMU Language Technology Institute where he lead the Multimodal Communication and Machine Learning Laboratory (MultiComp Lab). He was previously Research Faculty at USC Computer Science Department. He received my Ph.D. in Computer Science from MIT Computer Science and Artificial Intelligence Laboratory. His research focuses on building the computational foundations to enable computers with the abilities to analyze, recognize and predict subtle human communicative behaviors during social interactions. Central to this research effort is the technical challenge of multimodal machine learning: mathematical foundation to study heterogeneous multimodal data and the contingency often found between modalities. This multi-disciplinary research topic overlaps the fields of multimodal interaction, social psychology, computer vision, machine learning and artificial intelligence, and has many applications in areas as diverse as medicine, robotics and education.

Call for Papers

The eighteenth International Conference on Natural Language Processing (ICON-2021) will be held at NIT Silchar, India during December 16-19, 2021. The ICON Conference series is a forum for promoting interaction among researchers in the field of Natural Language Processing (NLP) and Computational linguistics (CL) in India and abroad. The main conference is on December 17-18, 2021. This will be preceded by one day of pre-conference tutorials / workshops on December 16, 2021 and one day of post conference shared tasks / tools / demos on December 19, 2021.

Papers in ICON proceedings will be indexed in ACL Anthology. ACL Anthology is a digital archive of research papers in Computational Linguistics for major international conferences under Association for Computational Linguistics (ACL), which is one of the most well-known associations for NLP and CL. The previous proceedings of ICON 2014, ICON 2015, ICON 2016, ICON 2017, ICON 2019 and ICON 2020 can be found in ACL Anthology.


Papers are invited on substantial, original and unpublished research on all aspects of Natural Language Processing, with a particular focus on South Asian languages and other less resourced languages, issues, and applications relevant to South Asia. However, other languages of the world are not excluded. The areas of interest include, but are not limited to:

Call for Tutorials / Workshops

Proposals are invited for pre-conference tutorials/workshops. Tutorials/Workshops can be of half-day or full-day duration. The proposal should be presented in the form of an extended abstract (1-2 pages) as per the ICON 2021 template (ACL template). This should contain a topical outline of the content, description of the proposers and their qualifications relating to the tutorial content.
Proposals for Tutorial/Workshop can be submitted at this link. Send tutorial/workshop proposals to the ICON-2021 Secretariat by email to icon2021nitsilchar@gmail.com. For further information, please refer to the Conference URL or contact the ICON-2021 Secretariat.

Call for Doctoral Consortium

The ICON organising Committee pleased to call for papers for the 3rd Doctoral Consortium. This event extends an opportunity for doctoral candidates to present and discuss their research with a panel of experts. The discussion would include a feedback on the evolution and progress of their research. It also helps them to identify the roadmap and additional studies, which could help refine the shape of their doctoral thesis. The doctoral consortium will be a one-day or half day event being organised on December 16, 2021, as part of the ICON-2021 conference at the National Institute of Technology Silchar (NIT Silchar). The applicants are required to submit a two-page extended abstract of their PhD research work. Submit your abstracts at this link . The shortlisted candidates would be invited to the consortium where they are required to present a summary of their research. Each candidate will be given 30 minutes for the presentation, which will be followed by a discussion of 15 minutes, led by a panel of experts. Prospective doctoral students from language technologies related disciplines are invited to apply. The selection of participants will be based on the submitted abstracts.


The invitation is open to all participants of ICON 2021. The applicants are required to submit an extended two-page abstract on their ongoing doctoral research. The submission can be extended to a maximum of two pages including all text, figures and tables, plus an additional third page exclusively for references. The submissions must follow the ICON template provided in the author's kit.The abstracts may incorporate published and in-progress work from the authors. Submissions are expected to present a fair picture of the research undertaken towards the thesis. Participants are advised to refrain from submitting a shorter version of their conference papers. Submissions must have the participant as the sole author. Acknowledgements to their supervisors, supporting agencies/bodies, and contributors to thework, can be made in a separate section.

The submission must highlight the following: The motivation of the research; Key issues identified/addressed; Major contributions; Methodologies, Experiments; Discussion of results; Future plans and Roadmap for the thesis.

Call for Shared Tasks /Tools / Demos

Proposals are invited for post-conference shared tasks / tools / demos. Shared Tasks /Tools / Demos can be of half-day or full-day duration. The proposal should be presented in the form of an extended abstract (1-2 pages) as per the ICON 2021 template (ACL template). This should contain a topical outline of the content, description of the proposers and their qualifications relating to the shared tasks / tools / demos content.
Proposals for Shared Tasks /Tools / Demos can be submitted at this link. Send shared task / tool / demo proposals to the ICON-2021 Secretariat by email to icon2021nitsilchar@gmail.com. For further information, please refer to the Conference URL or contact the ICON-2021 Secretariat.

Important Dates


Paper Submission Deadline

October 15, 2021October 22, 2021

Paper Acceptance Notification

November 15, 2021November 25, 2021

Paper Camera Ready Paper Submission

December 5, 2021

Doctoral Consortium Deadline

October 15, 2021

Paper Acceptance Notification (Doctoral Consortium)

November 15, 2021

Workshop Proposal Submission

September 15, 2021

Workshop Acceptance Notification

September 30, 2021

Tutorial Proposal Submission

October 10, 2021

Tutorial Acceptance Notification

October 30, 2021

Shared Task /Tool/ Demo Proposal Submission

September 15, 2021

Shared Task /Tool/ Demo Acceptance Notification

September 30, 2021


December 16-19, 2021

Paper Submission Information

Long Papers

Long paper submissions must describe substantial, original, completed and unpublished work.
Long papers may consist of up to 8 pages of content, plus unlimited references. Final versions of long papers will be given one additional page of content (up to 9 pages) plus any no of pages for the references.

Short Papers

Short paper submissions must describe original and unpublished work. Please note that a short paper is not a shortened long paper. Instead short papers should have a point that can be made in a few pages. Some kinds of short papers are:

A small, focused contribution

A negative result

An opinion piece

An interesting application nugget

Short papers may consist of up to 4 pages of content, plus unlimited references. Upon acceptance, short papers will be given 5 content pages in the proceedings.
Authors are encouraged to use this additional page to address reviewers' comments in their final versions.

Instructions for Double-Blind Review

As reviewing will be double blind, papers must not include authors' names and affiliations. Furthermore, self-references or links (such as github) that reveal the author's identity, e.g., "We previously showed (Smith, 1991) .." must be avoided. Instead, use citations such as "Smith previously showed (Smith, 1991) ..." Papers that do not conform to these requirements will be rejected without review.
Papers should not refer, for further detail, to documents that are not available to the reviewers. For example, do not omit or redact important citation information to preserve anonymity. Instead, use third person or named reference to this work, as described above ("Smith showed" rather than "we showed").
Papers may be accompanied by a resource (software and/or data) described in the paper, but these resources should be anonymized as well.


The author list for submissions should include all (and only) individuals who made substantial contributions to the work presented. Each author listed on a submission to ICON 2021 will be notified of submissions, revisions and the final decision. No changes to the order or composition of authorship may be made to submissions to ICON 2021 after the paper submission deadline.

Paper Submission and Templates

Submission is electronic, using the Softconf START conference management system. The submission site is now available at https://www.softconf.com/icon2021/papers/

The deadline for submission of both long and short papers is 22 October, 2021 (GMT -12).

Both long and short papers must follow the ACL Author Guidelines

Style sheets (Latex, Word) are available here: https://acl2020.org/downloads/acl2020-templates.zip

The Overleaf template is also available here: https://www.overleaf.com/latex/templates/acl-2020-proceedings-template/zsrkcwjptpcd

Please do not modify these style files, or use templates designed for other conferences. Submissions that do not conform to the required styles, including paper size, margin width, and font size restrictions, will be rejected without review.