Print

Print


*****  To join INSNA, visit http://www.insna.org  *****

Extreme Multilabel Classification for Social Media, 24th April 2018
In association with The Web Conference 2018, Lyon, France

https://urldefense.proofpoint.com/v2/url?u=https-3A__sites.google.com_view_xmlc_&d=DwIFaQ&c=pZJPUDQ3SB9JplYbifm4nt2lEVG5pWx2KikqINpWlZM&r=uXI5O6HThk1ULkPyaT6h2Ws3RKNKSY__GQ4DuS9UHhs&m=3E4U52oAdeiDlSQh8fhhaKr-CHXDVUFHAUou4uGGZm8&s=iBPHdA2tCnEtN6bpNMqQ86SEvI8qRhCLXwZvdxzNLQY&e= 

The Web Conference showcases state-of-the-art research in the fields of
information retrieval, machine learning, artificial intelligence and
computer science in general. The theme of this workshop is Extreme
Multilabel Classification (XMLC).

XMLC is a very active and rapidly growing research area that deals with the
problem of labeling an item with a set of tags out of an extremely large
number of potential tags. While the difficulty and the potential
applications of XMLC are well understood in the core machine learning
community, to the best of our knowledge, XMLC has not made inroads in the
field of Information Retrieval (IR) and related areas. The aim of this
workshop is to bring researchers from academia and industry to further
advance this very exciting field and come up with potential applications of
XMLC in new areas.

Authors are invited to submit long (8 pages) and short (4 pages) papers,
please clicks the following link for submission:

https://urldefense.proofpoint.com/v2/url?u=https-3A__easychair.org_conferences_-3Fconf-3Dwww2018satellites&d=DwIFaQ&c=pZJPUDQ3SB9JplYbifm4nt2lEVG5pWx2KikqINpWlZM&r=uXI5O6HThk1ULkPyaT6h2Ws3RKNKSY__GQ4DuS9UHhs&m=3E4U52oAdeiDlSQh8fhhaKr-CHXDVUFHAUou4uGGZm8&s=MFnHLUhcYYRB9M8OL60GMRjWbKYg_H0CadMlte7_PG8&e= 


Topics of interest include:

Given that the main aim of this workshop is to identify new application
areas for XMLC, we propose topics that are aligned with this goal along
with other topics in this area:

* New applications of XMLC: social media events, hashtags detection e.g.,
Twitter moments, e-commerce, multi-lingual XMLC

* Structured XMLC: knowledge graph/taxonomy, events as labels: temporally
structured events, spatially similar events

* Incremental inclusion of new labels and training data: zero shot
learning, pre- and post-training, active learning

* Multi-task multilabel learning: transfer learning, semi-supervised
learning

* Computational aspects of XMLC: log-time and log-space prediction, model
and computation parallelization

* Bayesian models for XMLC: generative models for XMLC, tackling label
polysemy, synonymy and correlations.

* Deep XMLC: neural models for XMLC

* Evaluation for XMLC: novel metrics for XMLC evaluation

* Feature extraction and feature engineering for XMLC



Important Dates

Submission Deadline: 10 February 2018
Acceptance Notification: 25 February 2018
Final Version Due: 4 March 2018
Workshop Date: 24 April 2018


Organizing Committee:

Akshay Soni, Yahoo Research, Sunnyvale
Robert Busa-Fekete, Yahoo Research, New York
Krzysztof Dembczyński, Poznan University of Technology
Aasish Pappu, Yahoo Research, New York

_____________________________________________________________________
SOCNET is a service of INSNA, the professional association for social
network researchers (http://www.insna.org). To unsubscribe, send
an email message to [log in to unmask] containing the line
UNSUBSCRIBE SOCNET in the body of the message.