Dear colleague, we recently started a W3C Community Group on Linked Data
Models for Emotion and Sentiment Analysis
(http://www.w3.org/community/sentiment/). The group is chaired by Bjoern
Schuller (Technical University Munich; Imperial College London) and
Fernando Sanchez (Universidad Politecnica Madrid).
The Community Group is a forum to promote the use of Linked Data best
practices in emotion and sentiment analysis research, in particular on:
- The definition of Linked Data based vocabularies for emotion and
sentiment analysis, where already existing vocabularies and schema for
emotion and sentiment analysis can be taken as a starting point, e.g.
EmotionML (http://www.w3.org/TR/emotionml/), Marl
(http://marl.gi2mo.org/), Onyx
(http://www.gsi.dit.upm.es/ontologies/onyx/) among others.
- Aligning requirements for and reaching consensus on the meaning and
use of varying features for describing polarity values, e.g. 1...5,
-2/.../2, positive/neutral/negative, very-bad/bad/.../good/very-good
- Aligning and representing as Linked Data the different schemas for
emotion analysis, see also http://www.w3.org/TR/emotion-voc/
- Other topics of interest to be suggested by the Community Group
We will have a first Community Group telco on July 30, 11am CET (one
hour meeting). We will use Webex for this meeting.
Please register your interest and availability by sending an email to
[email protected]
Also, please consider registering as Community Group participant at
http://www.w3.org/community/sentiment/
Regards
for the CG chairs: Paul Buitelaar
Insight Centre for Data Analytics (formerly DERI)
National University of Ireland, Galway
http://www.paulbuitelaar.net/
http://eurosentiment.eu/
