Publications
- We collect and analyze a set of Pinterest users, their content, and their interactions, and show that the follow mechanism on Pinterest is not well-utilized for content discovery.
Do I Need To Follow You?: Examining the Utility of The Pinterest Follow MechanismBluma Gelley, Ajita John
CSCW '15 Proceedings of the 18th ACM Conference on Computer Supported Cooperative Work & Social Computing, 2015 - Accept, decline, postpone: How newcomer productivity is reduced in English Wikipedia by pre-publication review.Jodi Schneider, Bluma Gelley, and Aaron Halfaker. In Proceedings of OpenSym 2014. pdf We discuss the pitfalls of pre-publication review in Wikipedia as exemplified by the optional Articles for Creation system. Using a dataset of AfC articles, we show that newcomer productivity is reduced by AfC, as compared to the regular article creation process.
- Automated Decision Support for Human Tasks in a Collaborative System: The Case of Deletion in Wikipedia. Bluma Gelley and Torsten Suel. In Proceedings of WikiSym 2013. pdf
We use a dataset of pages deleted from Wikipedia to create a classifier that identifies pages to be deleted with 98% accuracy. We plan to deploy this system on Wikipedia to ease the burden on humans of finding and deleting inappropriate pages. - What, Where, and When: Geotemporal Text Search in Twitter. S. Nepomnyachiy, B. Gelley, W. Jiang, T. Minkus, J. Friere. In Proceedings of GIR 2014. pdf
Fast, efficient geo-temporal indexing and query processing for Twitter. We exploit the structure of time-stamped data to dramatically shrink the temporal search space and use a shallow tree based on the spatial distribution of tweets to allow speedy search over the spatial and text dimensions. The system allows for low-cost dynamic index updates and adjusts the tree topology based on query distribution to optimize performance.
Works in Progress
- Modelling Answerers’ Social Personae Bluma Gelley, Ran Wolff, Dan Pelleg. 2015. In preparation for submission. We use a large dataset of Yahoo Answers users to validate whether sociological theories of dramaturgy and behavior apply to Yahoo Answers users.
- Investigating deletion in Wikipedia. Bluma Gelley. 2013. Invited to resubmit, Internet Computing. On Arxiv. We collect and analyze a dataset of deleted and retained pages from Wikipedia to determine whether the large volume of deletions on Wikipedia is legitimate or not. We conclude that the large number of deletions results from mostly uninformed users creating inappropriate pages, but that many deletions are carried out too hastily. We make recommendations for addressing both these problems. Wikipedia: A Survey. Bluma Gelley. 2012. In manuscript. pdf