Coarse and Fine Grained Sentiment Analysis of Online Text

Sentiment analysis - the automated extraction of expressions of positive and negative attitudes from text - has received a great amount of attention over the last ten years. Over the same period, via the widespread growth in the use of what we have come to call social media, there has been an explosion in the amount of publically available user generated text on the Web. This text has the potential of providing a source of real time, time tagged sentiments from people all over the globe.

The tools provided by statistical natural language processing and machine learning, along with exciting new scalable approaches to working with Big Data, make it possible to begin the work of extracting sentiment from the Web. We will discuss some of the challenges and approaches in doing this work. In particular, we will describe work we have done in annotating sentiment in blogs at the sentence and clausal level; in classifying subjectivity at the sentence level; and in identifying the targets, or topics, of sentiment at the clausal level.

Participate online on dimdim.