Sentiment Identification on Blogs

Researchers in the IDA section here at University of Glasgow were at the forefront of combining information retrieval and natural language processing techniques to identify opinionated blog posts. For instance, we led the creation of the TREC Blogs06 test collection, and the development of the corresponding TREC Blog track & Microblog test collections. We also published around the retrieval of posts expressing sentiment about a target. The datasets and research that we led in the TREC Blog track has been used by over 80 academic & industrial research groups worldwide, and led onto the popular TREC Microblog track initiative.

Related Research & Links