Andrew G. West

Andrew G. West

... is a Sr. Research Scientist at Verisign in the Washington D.C. area. In my role with the Strategy & Analytics group I apply deep learning, machine learning, forecasting, and other data science techniques across our domain name business and associated marketing efforts. I publish my research and pursue intellectual property protection via my affiliation with Verisign Labs. Prior to Verisign, I completed my Ph.D. at UPenn in 2013 (MSE 2010) and received my B.Sc. at Washington & Lee in 2007.

My recent work has focused on: (1) Leveraging deep learning for generative and classification tasks, (2) modeling and forecasting many aspects of the domain name lifecycle, (3) applying big data techniques to Internet-scale quantification problems, and (4) designing crowd-sourced data collection for market measurement and competitive intelligence. Complementing this work are broader research interests that include reputation management, metadata analysis, email/Web 2.0 abuse, underground economies, behavioral profiling, and Internet virality. Though the majority of my insights are now internally facing, my research approach draws from a rigorous background of 35+ peer-reviewed publications and a desire to bring emergent techniques to bear on business problems.

Before my current role, my dissertation research investigated security in "open collaboration" applications. I examined how user-generated content and collaborative semantics change how abuses manifest and can be detected. This yielded tools still in popular use, notably STiki, which has been used to remove 1.1+ million damaging revisions from Wikipedia. Media outlets such as the Chronicle of Higher Education and Gizmodo have recognized my contributions, and I remain an active volunteer in the Wikipedia community.

Publications [C.V.] [BIBTEX] [@SELECTED_WORKS] [@GOOGLE-SCHOLAR] [@DBLP]

In the interest of brevity, this list contains only recent and prominent writings. My C.V. enumerates all writings, for which full texts are available on my SelectedWorks profile. All downloadable texts are author's versions, with the authoritative copy belonging to the cited venue/journal:

West, A.G (2017). Analyzing the Keystroke Dynamics of Web Identifiers. In WebSci '17: Proc. of the 9th Intl. ACM Web Science Conference. Troy, NY, USA. [FORTHCOMING]

Heilman, J.M., and West, A.G (2015). Wikipedia and Medicine: Quantifying Readership, Editors, and the Significance of Natural Language. Journal of Medical Internet Research, 17(3). [ABSTRACT + AUTHOR'S PDF | JOURNAL HTML]

West, A.G., and Mohaisen, A. (2014). Metadata-driven Threat Classification of Network Endpoints Appearing in Malware. In DIMVA '14: Proceedings of the 11th Conference on Detection of Intrusions and Malware & Vulnerability Assessment, LNCS 8550 (S. Dietrich ed.), pp. 152-171. Egham, UK. [ABSTRACT + PDF | SLIDES]

West, A.G., and Aviv, A.J. (2014). On the Privacy Concerns of URL Query Strings. In W2SP '14: Proceedings of the 8th Workshop on Web 2.0 Security and Privacy. San Jose, CA, USA. A journal extension: Measuring Privacy Disclosures in URL Query Strings appears in IEEE Internet Computing, 18(6): pp.52-59. [CONF. PDF | CONF. SLIDES | JOURNAL PRE-PRINT]

West, A.G. (2013). Damage Detection and Mitigation in Open Collaboration Applications. PhD thesis, University of Pennsylvania. [ABSTRACT + PDF | SLIDES]

West, A.G., Hayati, P., Potdar, V., and Lee, I. (2012). Spamming for Science: Active Measurement in Web 2.0 Abuse Research. In WECSR '12: Proceedings of the 3rd Workshop on Ethics in Computer Security Research, LNCS 7398 (J. Blythe, S. Dietrich, and L.J. Camp eds.), pp. 98-111. Kralendijk, Bonaire. [ABSTRACT + PDF | SLIDES]

West, A.G., and Lee, I. (2011). What Wikipedia Deletes: Characterizing Dangerous Collaborative Content. In WikiSym '11: Proc. of the 7th Intl. Symposium on Wikis and Open Collaboration, pp. 25-28. Mountain View, CA, USA. [ABSTRACT + PDF | SLIDES]

West, A.G., Chang, J., Venkatasubramanian, K., Sokolsky, O., and Lee, I. (2011). Link Spamming Wikipedia for Profit. In CEAS '11: Proc. of the 8th Annual Collaboration, Electronic Messaging, Anti-Abuse, and Spam Conference, pp. 152-161. Perth, Australia. (co-Best Paper Award). [ABSTRACT + PDF | SLIDES]

Adler, B.T., de Alfaro, L., Mola-Velasco, S.M., Rosso, P., and West, A.G. (2011). Wikipedia Vandalism Detection: Combining Natural Language,Metadata, and Reputation Features. In CICLing '11: Proc. of the 12th Intl. Conference on Intelligent Text Processing and Computational Linguistics. LNCS 6609, pp. 277-288. Tokyo, Japan. [ABSTRACT + PDF]

West, A.G., Aviv, A.J., Chang, J., & Lee, I. (2010). Spam Mitigation using Spatio-Temporal Reputations from Blacklist History. In ACSAC '10: Proceedings of the 26th Annual Computer Security Applications Conference, pp. 161-170. Austin, Texas, USA. (Preliminarily published as UPENN-MS-CIS-10-04). [ABSTRACT + PDF | SLIDES]

West, A.G., Kannan, S., & Lee, I. (2010). Detecting Wikipedia Vandalism via Spatio-Temporal Analysis of Revision Metadata. In EUROSEC '10: Proceedings of the Third European Workshop on System Security, pp. 22-28. Paris, France. (A preliminary version was published as UPENN-MS-CIS-10-05). [ABSTRACT + PDF | SLIDES]

Software & Datasets

STiki (Spatio-Temporal analysis over Wikipedia): A Wikipedia anti-vandalism tool consisting of a back-end engine for real-time revision processing and a user-facing GUI which presents probable-vandalism to users. An extension to this software includes the code/analysis necessary to detect link spam edits. Software. [WEBSITE]

WikiAudit: Heuristics to detect unconstructive wiki changes from a provided set of IPs; useful for network admins and investigations into editor bias. Software. [WEBSITE]

Reputation Management Simulator: A framework to evaluate reputation algorithm performance. Software. Last updated: 2009/10/25. [ZIP | DIAGRAM-PNG]

Personal Details

On a purely personal level, I have three fanatical interests in life:

[1] West Virginia University Athletics: I was born and raised in WV.
[2] Tom Petty & The Heartbreakers: Thirty years on and still rockin' hard.
[3] Maxwell [1, 2, 3]: Best friend and extremely intimidating 7-pound chihuahua.

Contact Information

Andrew G. West
Verisign Headquarters, 7241
12061 Bluemont Way
Reston, VA 20190

email: awest {at, @} verisign [dot] com (business)
email: west [dot] andrew [dot] g {at, @} gmail [dot] com (personal)