4th-5th October 2019, London, UK
The mission
of the Conference on Truth and Trust Online (TTO) is to bring together all
parties working on automated approaches to augment manual efforts on improving
the truthfulness and trustworthiness of online communications.
The modern
media ecosystem in which the public seeks truthful information and trustworthy
sources is increasingly dominated by large-scale online social media and
communication platforms (e.g. Facebook, Twitter, YouTube) as well as messaging
apps, which operate alongside traditional news sources. However, online
platforms are facing significant challenges in this regard, as the majority of
the public mistrusts the content shared on them and only 50% trust any kind of
online news (Digital News Report 2018). The quest to improve the online media
ecosystem remains important and is in need of concerted research.
Manual
efforts cannot scale to the demand, and many proposed automated solutions for
content moderation and verification, including those presented in existing
academic venues (e.g. FEVER, RDSM, POC, etc.) are detached from real-world
needs faced by the users and the operators of these platforms.
Research in
this area is currently disseminated in two main avenues: 1) technical workshops
attached to conferences of different communities (e.g., WWW, KDD, ICWSM, ACL,
NIPS, etc.); 2) hybrid community gatherings such as MisinfoCon. The latter
brings journalists, social scientists, and technical entrepreneurs together,
but is more focused on sharing the latest findings instead of being a platform
for publishing original research. The former encourages the presentation of
original research, but suffers from fragmentation across different academic
disciplines. We believe it is time to bring the best from both avenues and
foster a technical community in this space focusing on publishing and
disseminating original research while also considering real world scenarios and
impact.
The Conference on Truth and Trust Online (TTO) aims to be an annual forum for
academia, industry, non-profit organizations and other stakeholders to meet and
discuss the problems facing social media platforms and technical solutions to
them. Key to this interaction is the openness on behalf of the media platforms
about their problems and the focus of the academic proposals to address them in
an impactful way.
The
intended outcomes of our conference are: a) for academics to learn about the
real problems that industry is facing and how their proposed solutions can be
more impactful; b) for industry to improve their product safety by
brainstorming collective actions together with other stakeholders; and c) for
the public to gain insights into how their concerns on social media safety are
being addressed.
We have the
following invited speakers (with more to be announced as confirmed):
Submissions
We invite
submissions on technical solutions for addressing current challenges facing
social media platforms on the following indicative list of topics:
- Misinformation
- Disinformation
- Hate speech
- Harassment/bullying
- Credibility scores
- Hyper-partisanship and bias
- Image/video verification
- Fake amplification
- Fake reviews
- Polarization and echo chambers
- Transparency in content and source moderation
- Privacy requirements
The focus
of the conference is on identifying what new problems and technical solutions
we need to work on, rather than incremental solutions to what we are doing
already. Therefore papers that only aim to raise awareness, or propose improved
methods on existing tasks/datasets, without showing how these could be used to
solve real world problems as the one indicated above are explicitly
discouraged.
We invite
three types of submissions: technical papers, talk proposals and dataset
proposals:
- Technical papers should contain novel, previously unpublished material related to the topics of the conference and should be limited to 8 pages with unlimited pages for references. The papers accepted will be presented either orally in single track sessions or as posters.
- Talk proposals should be 2 pages long and describe the content of a half-hour long talk giving a tutorial-style overview of academic work on a certain topic, or present hands-on experience from addressing the TTO challenges in the real world.
- Dataset proposals should be 2 pages long and propose the annotation of a dataset of publicly available data. A clear analysis of the copyrights, respect of privacy and data availability must be given. The size of the datasets to be built will depend on the difficulty of the task. The selected papers will be invited to refine their proposal to receive funding in a separate Request for Proposal.
Key dates
Deadline
for all submissions: 3rd of June 2019 (23:59, anywhere on earth)
Notification
of acceptance: early July
Registration
opens: end of June
Conference:
4th and 5th of October 2019, BMA House, London, UK
Organization
Program
Chairs: Maria Liakata (University of Warwick) and Andreas Vlachos (University of Cambridge)
Local
Organization: Guillaume Bouchard (Facebook), Simon Martin (Mumsnet), Marzieh
Saeidi (Facebook)
Publications
Chair: Kalina Bontcheva (University of Sheffield)
Publicity
Chair: Mevan Babakar (Full Fact), David Corney (Full Fact)
Area
Chairs: Emiliano De Cristofaro (University College London), Ferenc Huszar
(Twitter), Sebastian Riedel (Facebook and University College London), Cong Yu
(Google)
Sponsorship
Chairs: Arpit Mittal (Amazon), Miriam Redi (Wikimedia Foundation)
No hay comentarios:
Publicar un comentario