Sentiment Analysis is the newest buzzword these days owing to its immense potential to provide real-time feedback. Businesses, governments and individuals benefit from it every day. It is a growing field and of interest to researchers in the domains of Computer Science, Social Sciences and Marketing alike. It is, however, still in its nascent stages and improvements are required on many fronts. The objective of this thesis is to investigate the challenges in the field of Sentiment and Affect Analysis on Social Media.