Knowledge graphs (KGs) have demonstrated remarkable effectiveness on various knowledge-intensive computational tasks in a range of domains, including journalism and science. Most work on the automatic construction of KGs has focused narrowly on extracting factual assertions from the source material, ignoring an important dimension--argumentation. Many texts that communicate factual knowledge couch that knowledge in argument, in language that asserts, interprets, and reasons. Arguments are an essential means for forming stances on controversial topics, arriving at successful decisions in the face of uncertainty, and establishing consensus on problems that require interpretation. An active research community has investigated computational models of textual argumentation. Still, many important questions lack concrete answers, including the best approaches to formalizing argumentation as a graph, deploying such graphs at scale, and integrating argumentative knowledge with factual knowledge typically found in KGs.
This inaugural workshop on Argumentative Knowledge Graphs (ArgKG) will explore the automatic construction of KGs that encode argumentative knowledge and structures as well as the incorporation of argumentation into factual knowledge graphs, drawing together researchers focusing on NLP, automatic KG construction, and computational analysis of argumentation.
The topics for submissions include but are not limited to:
We invite the submission of papers and abstracts describing previously unpublished research, including new methodology, datasets, evaluations, surveys, reproduced results, negative results, and visionary positions.
Papers are non-archival, with the flexibility to submit in any venue in addition to the ArgKB workshop. Accepted papers will be listed on the workshop schedule, and their full texts will be published in OpenReview. We invite papers of all types, including mature works, late-breaking results, work-in-progress, and position papers. Articles under review in other venues are also welcome if that is permissible. Papers should be restricted to 10 single-column pages, excluding references. Appendices should be put after references and submitted in one PDF document. Like main submissions, supplementary material must be anonymized.
We welcome submissions of extended abstracts (2 pages max) related to the research topics mentioned above. Submissions may include previously published results, late-breaking results, and work in progress. Extended abstracts can summarize existing work, work in progress, or a collection of works under a unified theme (e.g., a series of closely related papers that build on each other or tackle a common problem). Abstracts will be lightly reviewed to ensure that the topic is within the scope of the workshop. The abstracts are non-archival; participants are free to also submit their work for publication elsewhere. Figures and tables will count toward the length limit while references are excluded. Similar to full papers, extended abstracts should be anonymous during submission.
***All submission deadlines are 11.59 pm UTC -12h (“Anywhere on Earth”)***
Workshop registration is included in AKBC 2021 registration, along with access to the main conference (October 4-7).
Website Credits: https://uskb-workshop.github.io