NSU Newsroom
SharkBytes
Horizons
This version of NSU News has been archived as of February 28, 2019. To search through archived articles, visit nova.edu/search. To access the new version of NSU News, visit news.nova.edu.
This version of SharkBytes has been archived as of February 28, 2019. To search through archived articles, visit nova.edu/search. To access the new version of SharkBytes, visit sharkbytes.nova.edu.
GSCIS Student Receives Amazon Web Services Research Grant
Gregory Reinert, doctoral student in computer information systems at the Graduate School of Computer and Information Sciences, recently received the Amazon Web Services research grant to support work on his dissertation.
Reinert’s proposed dissertation is entitled “Early Detection of Online Auction Opportunistic Sellers Through the Use of Negative-Positive Feedback.” When a buyer makes a purchase from a seller on eBay, the buyer has the option to leave feedback type (negative, positive or neutral) about the transaction plus a feedback comment. The proposed research will investigate if the presence of subtle buyer behavior in the form of negative-positive grammar constructs in the feedback comments for positive feedback types is a signal to other potential buyers that a seller is behaving fraudulently.
The traditional analysis technique is using multiple dedicated human raters. Each rater must review all the feedback comments. Even with a minimum of three raters, reviewing the estimated 400,000 to 700,000 feedback comments would take a minimum of 175 days at a cost of $37,000. Using an alternative technique called crowdsourcing, the raters’ work should be completed in an estimated 5-7 days at a cost of $3,500. The crowdsourcing service will be provided by using Amazon Mechanical Turk (www.mturk.com).