Kalina Bontcheva, University of Sheffield


Title: Natural Language Processing for Social Media: Are We There Yet?


Abstract: 

Social media poses three major computational challenges, dubbed by Gartner 
the 3Vs of big data: volume, velocity, and variety. NLP methods, in 
particular, face further difficulties arising from the short, noisy, and 
strongly contextualised nature of social media. As a result, NLP methods 
generally tend to perform significantly worse on social media, than on 
longer, cleaner texts, such as news. 

After a short introduction to the challenges of processing social media, 
the talk will cover key NLP algorithms (corpus annotation, linguistic 
pre-processing, information extraction and opinion mining) adapted to 
processing social content, discuss available evaluation datasets and 
outline remaining challenges. 

Since the lack of human-annotated NLP corpora of social media content is 
one of the key challenges, the talk will cover also crowdsourcing 
approaches used to collect training and evaluation data (including 
paid-for crowdsourcing with CrowdFlower and its combination with 
expert-sourcing). I will also discuss briefly practical and ethical 
considerations, arising from gathering and mining social media content.