# Workshop Description

For the past fifteen years, the workshops in the TextGraphs series have published and promoted the synergy between the field of Graph Theory (GT) and Natural Language Processing (NLP). The mix between the two started small, with graph theoretical frameworks providing efficient and elegant solutions for NLP applications. Graph-based solutions initially focused on single-document part-of-speech tagging, word sense disambiguation, and semantic role labeling, and became progressively larger to include ontology learning and information extraction from large text collections. Nowadays, graph-based solutions also target on Web-scale applications such as information propagation in social networks, rumor proliferation, e-reputation, multiple entity detection, language dynamics learning, and future events prediction, to name a few.

The fifteenth edition of the TextGraphs workshop aims to extend the focus on graph-based representations for (1) large-scale knowledge bases and reasoning about them and (2) graph-based and graph-supported machine learning and deep learning methods.

# Important Dates

- March 22, 2021: Workshop Papers Due Date
- April 15, 2021: Notification of Acceptance
- April 26, 2021: Camera-ready Papers Due
- June 11, 2021: Workshop Date

# Shared Task

We are organizing a shared task before the workshop!

Many-hop multi-hop inference is challenging because there are often multiple ways of assembling a good explanation for a given question. This 2021 instantiation of the shared task focuses on the theme of determining relevance versus completeness in large multi-hop explanations. To this end, this year we include a very large dataset of approximately 250,000 expert-annotated relevancy ratings for facts ranked highly by baseline language models from previous years (e.g. BERT, RoBERTa).

Submissions using a variety of methods (graph-based or otherwise) are encouraged. Submissions that evaluate how well existing models designed on 2-hop multihop question answering datasets (e.g. HotPotQA, QASC, etc) perform at many-fact multi-hop explanation regeneration are welcome.

More information about the task held in TextGraphs-15 can be found here:

* https://competitions.codalab.org/competitions/29228 (Overview and Submission)
* https://competitions.codalab.org/forums/25924/ (Forums)
* https://github.com/cognitiveailab/tg2021task (Instructions and Baseline)

We welcome papers on the workshop topics even if you do not participate in the shared task.

# Workshop Topics

TextGraphs-15 invites submissions on (but not limited to) the following topics:

* Graph-based and graph-supported machine learning methods:

- Graph embeddings and their combinations with text embeddings
- Graph-based and graph-supported deep learning (e.g., graph-based recurrent and recursive networks)
- Probabilistic graphical models and structure learning methods
- Exploration of capabilities and limitations of graph-based methods being applied to neural networks
- Investigation of aspects of neural networks that are (not) susceptible to graph-based analysis

* Graph-based methods for Information Retrieval, Information Extraction and Text Mining:

- Graph-based methods for word sense disambiguation
- Graph-based representations for ontology learning,
- Graph-based strategies for semantic relation identification
- Encoding semantic distances in graphs
- Graph-based techniques for text summarization, simplification, and paraphrasing
- Graph-based techniques for document navigation and visualization
- Reranking with graphs

* New graph-based methods for NLP applications:

- Random walk methods in graphs
- Semi-supervised graph-based methods
- Dynamic graph representations
- Graph kernels

* Graph-based methods for applications on social networks

- Rumor proliferation
- E-reputation
- Multiple identity detection
- Language dynamics studies
- Surveillance systems

* Graph-based methods for NLP and the Semantic Web:

- Representation learning methods for knowledge graphs (i.e., knowledge graph embedding)
- Using graph-based methods to populate ontologies using textual data
- Inducing knowledge of ontologies into NLP applications using graphs
- Merging ontologies with graph-based methods using NLP techniques

# Submission

We invite submissions of up to eight (8) pages maximum, plus bibliography for long papers and four (4) pages, plus bibliography, for short papers.

The NAACL 2021 templates must be used; these are provided in LaTeX and also Microsoft Word format. Submissions will only be accepted in PDF format. Deviations from the provided templates will result in rejections without review.

Download the Word and LaTeX templates here: https://2021.naacl.org/calls/style-and-formatting/

Submit papers by the end of the deadline day (time zone is UTC-12) via our Softconf.

# Organizers

- Alexander Panchenko, Skoltech, Russia
- Fragkiskos D. Malliaros, Paris-Saclay University, CentraleSupelec, Inria, France
- Varvara Logacheva, Skoltech, Russia
- Abhik Jana, University of Hamburg, Germany
- Dmitry Ustalov, Yandex, Russia
- Peter Jansen, University of Arizona, USA

# Contact

Please direct all questions and inquiries to our official e-mail address (textgraphsOC@gmail.com) or contact any of the organizers via their individual emails.

Join us on Facebook: https://www.facebook.com/groups/900711756665369/

Follow us on Twitter: https://twitter.com/textgraphs

Join us on LinkedIn: https://www.linkedin.com/groups/4882867