Lavanya Ram
*** ***** *** ******, ********* #*, Ames, IA 50010
Ph: +1-515-***-**** Web: www.cs.iastate.edu/~lram
E-mail:abqfha@r.postjobfree.com
Objective
Seeking a full-time/co-op position.
Education
Master of Science in Computer Science (MS in CS), Iowa State University, Iowa
Anticipated May 2010, CGPA: 3.56/4.0
Bachelor of Technology in Computer Science and Engineering (B.Tech. CSE), SRM University, India
May 2008, CGPA 9.432/10, Recipient of Best Performance Award in the year 2008
Technical skills
Programming : Core Java, C, C++, PL/SQL, Python, Unix Scripting, Spring, Hibernate.
Databases : Oracle 9i, Microsoft Access, MySQL.
Statistical Tools : SAS, SPSS 14.0, MATLAB (Neural Network Toolbox), Weka 3.
Automated testing tools : Selenium RC, Selenium Grid, JUnit, Fitnesse.
Operating Systems : Windows 2000/XP/Vista, Linux, OpenSolaris, Mac OS X.
Servers : Apache Tomcat, Glassfish Server, Jetty.
Web development : JSF, HTML, XML, Microsoft FrontPage.
Version Control System : SVN.
IDEs : NetBeans IDE, Eclipse IDE, Selenium IDE.
Miscellaneous : Apache Ant, OpenGrok, Nutch, Apache Lucene, Solr, LuSQL, Luntbuild, Luke.
Professional Experience
May 2009 present.
Software Engineering Intern - Source Allies, Inc., Des Moines, Iowa
Automated testing - Vendor Portal Selenium RC, Grid
Domain: Automated web testing. Language : Java.
Wrote Selenium tests for automated testing of the Vendor portal web application using Selenium RC.
Automated testing saved many hours of costly manual testing by the testing and QA team and also employed
agile development techniques to find errors earlier in the development cycle.
Worked as a team to set up Selenium Grid and modified tests to run on Selenium Grid for distributed testing.
Configured Fitnesse to automatically detect installed browsers and run the Fitnesse tests on each of them.
CAS for web applications
Domain: Single sign-on authentication
Set up a single-sign on authentication system for the company s web applications.
Created new keystore with self-signed web page in trusted list and imported it into the JVM trust store.
Project helped staff to remotely access the web applications and also prevented unauthorized access.
Searching the company s source code Lucene, OpenGrok
Domain: Information retrieval, Open source search engine. Language: Java.
Set up and configured OpenGrok to index and search the company s source code.
Set up project features comparable to that of the Google Search Appliance thereby saving the company heavy
costs of buying the appliance. OpenGrok can also cross-reference code while the Google appliance cannot.
Searching the company s intranet Lucene, Nutch and Solr
Domain: Information retrieval, Open source search engine. Language: Java.
Set up Nutch and Solr to index company s intranet, local files, MySql databases and files in samba shares.
Set up crawling using the lower level internals of Nutch.
Designed and developed a Nutch plug-in to index meta tags and return recommended search results.
Designed and developed a plug-in to search web pages by the e-mail id of their author.
Corrected a bug in the index-category plug-in to be able to index and search within specific collections of urls.
Modified the search ranking preference using Dismax Request Handler.
Classroom Management System Spring/Hibernate
Domain: Aspect oriented programming, Web app development. Language: Java, Hibernate, Spring.
Developed a console application with O/R mapping DAOs using Spring framework and Hibernate API for
data access.
Designed the database structure and constructed ER diagrams.
Wrote JUnit tests to test the application.
User Interface for Fitnesse JSF
Domain: Web application development. Language: Java Server Faces
Developed a better user interface for Fitnesse.
The UI was more user friendly to QA staff for creating tests and suites in Fitnesse than editing wiki pages.
Research Experience
Research Assistant, Artificial Intelligence Laboratory, Iowa State University (Dec 08 - Present)
Domain: Bayesian Networks, Machine learning. Language: C++ in Linux, Weka.
Working under Dr. Jin Tian on the project Learning the k-best globally optimal Bayesian network structures
with respect to the Bayesian Dirichlet Equivalent (BDE) score.
Effect of in-degree restrictions on the performance and accuracy of Bayesian learners.
Description : The aim of the project is to make Bayesian network learning more efficient (which is known to be
NP-Hard). The algorithm formulated by Silander was modified to find the top-k Bayesian network structures when
there is an a priori restriction on the maximum in degree any node can have and make efficiency comparisons of the
two algorithms. This project will help statisticians to feasibly learn the optimal Bayesian network structure for datasets
with large number of variables (>33).
Project Details
Combating hoarding by honey pots in web search: An Improvement to the PageRank algorithm.
Domain: Web Search, web mining and Information retrieval. Language: Java.
Description : An attempt to improve the PageRank algorithm by developing a semi-automatic method to rank web
pages such that spam pages linked to by link farms and honey pots are ensured to get a lower score than the other
good pages. A set of seed pages selected from the web graph by a heuristic are initially evaluated by an expert. Once
the expert classifies the seed pages, the link structure of the web is analyzed to discover other good and bad pages.
Implementation of a Decision tree based learning agent.
Domain: Classifier Algorithms. Language: Java.
Description : Implemented an efficient algorithm for learning Decision trees from a data set. The data set is split
into training & test sets and the tree is learned using the training set. Validation is then done using the test set.
Neural Data Mining for Credit Card Fraud Detection A Predictive Modeling Approach.
Domain: Machine learning using Neural networks. Language & tools: Python, SPSS.
Description : Combines advanced data mining techniques and neural network algorithms to obtain a high fraud
coverage combined with a low false alarm rate. A RBF network was used to develop a set of rules to recognize and
predict certain conditions. It works best at recognizing, predicting and controlling patterns in payment reviews.
Languages known: English, Hindi, Tamil, German (beginner level).
Achievements
Received the Bharat Petroleum Scholarship Award from Bharat Petroleum Corporation Ltd., India in 2008.
Among the top 2% of the students in the 3rd National IT Aptitude test conducted by NIIT.
Recipient of the Certificate of Merit from Central Board of Secondary Education, India.
Awarded Distinction in the Third National Science Olympiad. (2001)
Responsibilities/Special Interests
Volunteered (Non Profit initiative) for a camp to help visually challenged students use computers using
JAWS at IIT Madras in 2008.
Organizer of an Industrial tour for about sixty students. (Sep 2006)
Organizer for National level Technical Symposia in 2007 and 2008 at SRM University.
Chief Editor of the CSE Department newsletter at SRM University.
References: Available upon request.