The case for an open standard platform for World-Wide Social Media Analysis
- Details
- Aangemaakt op vrijdag, 15 juli 2011 14:38
- Geschreven door Arthur
Starting with the Usenet systems in 1979, social media has evolved into a medium that influences the lives of more and more people each day, in both their professional and personal context. People tell each other about their views, people learn from each other, and some views become truths. Insights in these developing opinions provide many opportunities for companies but also for governments and for researchers. Meanwhile, borders hardly exist anymore: companies and institutes join forces, people move, researchers collaborate; the world has become flat (Friedman).
Inspired by the potential benefits of understanding the similarities, differences, and development of opinions between countries, one of our dreams is to create a World-Wide, language-independent, opinion mining platform.
Such a platform will help companies, governments, and researchers to build applications to get a better grip on developing trends, the opinion towards their products, art, or legislation, and show insights in inter-cultural differences.
Combining insights across countries on a World-Wide scale, brings up many challenges that are until now underexposed. The relation between natural language and affective information has been studied widely. Sentiment analysis, social network analysis, and personalized access to social media have gained interest in the past years. Still, few systems understand multiple languages and contexts. On the other hand many existing language resources have difficulties dealing with the brevity and informal language use of social media content.
Starting from resources and collections like SentiWordNet, local sentiment dictionaries, TREC’s blog and Microblog tracks, we should aim to facilitate the open exchange and re-use of social media analysis components.
