The file wnscores_inquirer.csv was created using the algorithm of @inproceedings{Blair-Goldensohn-etal08, Address = {Beijing, China}, Author = {Blair-Goldensohn, Sasha and Hannan, Kerry and McDonald, Ryan and Neylon, Tyler and Reis, George A. and Reynar, Jeff}, Booktitle = {{WWW} Workshop on {NLP} in the Information Explosion Era (NLPIX)}, Title = {Building a Sentiment Summarizer for Local Service Reviews}, Year = {2008}} with the Harvard Inquirer words as its seeds (Positiv, Negativ, and unmarked for those categories), but limiting attention to the (word, tag) pairs from that database that are also in WordNet. For additional details on the mapping from Inquirer Tags to WordNet tags, as well as an overview of the Blair-Goldensohn et al. algorithm, see the 'Sentiment lexicons' handout from Linguist 287 / CS 424P: Extracting Social Meaning and Sentiment, Stanford, Fall 2010. ---Chris Potts, 2010-10-04