An important part of our information-gathering behavior has always been to find out what other people think. With the growing availability and popularity of opinion-rich resources such as online review sites and personal blogs, new opportunities and challenges arise as people now can, and do, actively use information technologies to seek out and understand the opinions of others. The sudden eruption of activity in the area of opinion mining and sentiment analysis, which deals with the computational treatment of opinion, sentiment, and subjectivity in text, has thus occurred at least in part as a direct response to the surge of interest in new systems that deal directly with opinions as a first-class object. This survey covers techniques and approaches that promise to directly enable opinion-oriented information-seeking systems. Our focus is on methods that seek to address the new challenges raised by sentiment-aware applications, as compared to those that are already present in more traditional fact-based analysis. We include materialon summarization of evaluative text and on broader issues regarding privacy, manipulation, and economic impact that the development of opinion-oriented information-access services gives rise to. To facilitate future work, a discussion of available resources, benchmark datasets, and evaluation campaigns is also provided. 1

This paper presents five empirical tests of the popular modeling abstraction that bidders in ascending online auctions bid “as if ” they were in a sealed bid auction. The tests rely on observations of the magnitudes and timings of top two proxy bids, with the different tests stemming from different regularity assumptions about the underlying joint distribution of signals and timings. We apply the tests to data from three eBay markets- MP3 players, DVDs and used cars- and we reject the sealed-bid abstraction in all three datasets. This consistent rejection casts doubt on several existing theories of online-auction behavior. Moreover, we reject the sealed-bid abstraction even in carefully selected subsets of the data that one might consider more likely to conform to sealed bidding. Given these findings, demand-estimation using eBay data is more difficult than previously thought. In particular, our results suggest that the empirical strategies based on multiple order-statistics of the bidding distribution will not work. We thank Raphael Thomadsen, Jeremy Fox, Ali Hortacsu, Harry Paarsch, and seminar participants at Chicago GSB for comments and suggestions. We also thank eBay for providing the data, and Uri Simonsohn for cleaning up and sharing the movie dataset. The views expressed do not necessarily reflect the views of the FTC or any individual Commissioner. All errors are our own. Contact: Robert Zeithammer, UCLA Anderson School of Management, 110 Westwood