SHILAD SEN

Minneapolis, MN •    shilad [A-T] gmail [D-T] com  •   http://www.shilad.com

 

EDUCATION

 

Expected June ‘08

University of Minnesota , Minneapolis, MN – (GPA: 3.80/4.00)

Doctoral candidate in Computer Science

 

June '03 - June '04

Tufts University, Medford, MA – (GPA: 4.00/4.00)

Post-Baccalaureate Studies, Computer Science

 

June '99

Northwestern University, Evanston, IL – (GPA: 3.52/4.00)

B.A. Mathematics / B.M. Saxophone Performance

 

RESEARCH

 

Oct ’04 – Present

Research Assistant, GroupLens Research Group, University of Minnesota, Minneapolis, MN

·        Social Tagging Systems (e.g. http://flickr.com, http://del.icio.us):

o       Analyzing factors influencing vocabulary development in social tagging systems.

o       Building human and machine-learning powered systems to classify individual tags as good or bad.

o       Modeling user tagging behavior in order to understand impact of user interface design decisions.

·        Recommender Systems:

o       Implemented and evaluated a wide range of recommender algorithms using multiple datasets

·        Assisted in creation of http://wikilens.org, a member-maintained recommender community.

 

June '06 - Present

Research Intern, Thomson Legal Research and Development Group, Eagan, MN

·        Tasked with improving recommendation system for over 10 million documents on westlaw.com .

·        Designed and evaluated hybrid metadata/clickstream clustering algorithms using offline usage data.

·        Currently preparing for live, online evaluation of recommendation algorithm on http://westlaw.com.

 

Summer, '05

Research Intern, IBM Research Collaborative User Experience Group, Cambridge, MA

·        Improved alert notification systems in Activity Explorer, a project supporting ad-hoc group collaboration.

·        Served as lead author of CSCW 2006 research paper based on research.

·        Created general alert filtering system in employing rules-based and predictive software alert filtering.

·        Conducted user-study analyzing notification behavior of 35 users using Activity Explorer.

·        Filed six patents based on research results.

·        Reviewed existing notification system research.

 

PUBLICATIONS

 

 

·        Sen, Lam, Cosley, Al-Rashid, Frankowski, Harper, Osterhouse, Reidl, “tagging, communities, vocabulary, evolution.”  CSCW 2006.  Winner of Best Paper (1%).

·        Sen, Geyer, Muller, Moore, Brownholtz, Wilcox, “FeedMe: A Collaborative Alert Filtering System.”  CSCW 2006. (22%).

·        Frankowski, Cosley, Sen, Terveen, Riedl, “You Are What You Say: Privacy Risks of Public Mentions.”  SIGIR 2006 (19%).

·        Shafer, Sen, Frankowski, Herlocker, “Collaborative Filtering Recommender Systems” chapter to appear in “Adaptive Web-Based Systems” textbook.  Scheduled for publication in early 2007.

 

EXPERIENCE

 

Sep. '98 – Oct '04

Engineer and Lead Engineer, Sourcelight Technologies, Evanston, IL

·        Promoted to lead engineer after one year

·        Led research and development for Sourcelight's Recommendation Engines

o       Created and oversaw maintenance of hosted movie recommendation service for clients such as Blockbuster Inc., Hollywood Video, and Comcast Cable.

o       Created Bayesian regression models for movie prediction in C/C++ that derive an optimal subset of regression equations

o       Designed a scalable, redundant system architecture in C/C++ and python that supports features such as automatic failover and on-the-fly addition of new servers.  Demonstrated performance improvements of up to 20X over original system.

o       Developed clustering models for music prediction built from nearly half a billion user rating

 

Sep. '98 – Oct '04

Engineer and Lead Engineer, Sourcelight Technologies, Evanston, IL (continued)

·        Implemented projects including:

o       Transitioning a licensed recommendation engine to a J2EE, hosted ASP environment

o       Creating an intelligent enterprise media management system employing rules-based scheduling

o       Developing a kiosk-based movie application using Python, PHP, and Mozilla deployed to Blockbuster, Inc.

·        Created and maintained a movie database entry system used by over a dozen content editors using  Python, Perl and Java

 

July ’03 – Jan '05

(part time)

Independent Consultant, Harvard School of Public Health – Cohands Project, Boston, MA

·        Restructured Sybase database entry system to ensure access to critical statistical metrics

·        Performed maintenance and development for a Cold Fusion-based medical subject tracking system.

 

Sept. '98 – Sep. '99

Software Engineer, Sourcelight Technologies, Evanston, IL

·        Directed installation and maintenance for a movie kiosk program with Blockbuster, Inc.

·        Created and maintained a movie database entry system used by over a dozen content editors using  Python, Perl and Java

·        Led design and development for several movie websites written in Perl, Python and Java.

·        Promoted to lead engineer after one year

 

Summer, '98

Information Technology Development Intern, Swiss Bank Corporation, Chicago, IL

·        Overhauled website written in Perl used for clearing and settling SBC's North American trades

·        Standardized configuration and management of website, improving ease of deployment and enabling portability between server environments

·        Improved usability and speed of website and supporting Java applications

·        Offered full time position upon completion of internship.

 

PROJECTS

 

Mar '01 – April '03

Fast Python XML-RPC Library, http://www.sourceforge.net/projects/py-xmlrpc

·        Created cross platform implementation of XML-RPC protocol, a remote procedure call mechanism that serializes object using XML and sends them using  HTTP

·        Demonstrated speed improvements of up to a factor of 20 over similar libraries

·        Cultivated widespread adoption of library through its incorporation in at least two major Linux distributions (Debian and GenToo) and over 9000 downloads

·        Implemented second generation of library (now called ehs) with more efficient and modular C API.  See http://cvs.sourceforge.net/viewcvs.py/py-xmlrpc/ehs/

 

Sept '98 - June '99

Computer Jazz Improvisation System, http://www.shilad.com/jazz

·        Created software system in Java that allows users to input chords, harmonic devices, and songs

·        Derived and implemented expert-based rules to generate improvisational solos using these constructs

 

April ‘06

Co-producer and saxophonist, “Arctic Fury,” http://www.snowblindmusic.com

·        CD from Snowblind, a jazz quintet

 

 

Also see http://www.shilad.com/projects

 

HONORS

 

 

Best Paper Winner, CSCW Conference, 2006

 

First Place, Sourcelight Technologies Movie Prediction Modeling Contest (2001)

 

Second Place, North American Saxophone Alliance Jazz Competition (1999)

Lead Tenor Saxophone, University of Minnesota Jazz Ensemble (2005-2006)

 

National Merit Scholarship Finalist and Recipient

 

SKILLS

 

 

Programming Languages:  C/C++, Java, Python,  Perl, SQL, Tcl/Tk, VB, Shell, LISP, ML, Groovy

Operating Systems: Linux, Sun/Solaris, Windows

Technologies: SOAP, XML-RPC, HTTP, TCP/IP, J2EE, XML, Apache, Tomcat, JBoss, SWT, Hibernate

Tools: Standard unix tools (vi, make, sed, awk, etc.) Eclipse, JBuilder, Webstudio, Visual Studio, ant

Databases: Postgresql, MySQL, Sybase, Oracle

 

INTERESTS

 

 

Playing jazz saxophone, squash, traveling, building things, basketball