Murali Krishna, Somanur Ramasamy is a Masters student in Computer Science at Stony Brook University joined in Fall 2014.
Smart Scaling in Stateful Servers
- Most of the previous work in scaling has focused on application tier and very less work had been done on stateful nodes’ scaling.
- Even though scaling in stateful tier gives high energy efficiency, it is difficult to implement.
- Planning to explore the options in scaling distributed databases.
- Analyze the performance degradation in case of node failure in distributed databases (Cassandra).
- Explore the scaling opportunities by configuring parameters like replication factor under different types of load
- Extend the possible improvements in database tier with caching tier optimizations
BioMedical Text Mining
- The aim of the project is to answer few interesting questions from BioMed text corpus using Dynamic Topic Modeling.
- The possible questions include,
- What are the recent topics with the most research output?
- What are the topics that are most prominent in the entire corpus and how have they evolved over time?
- What topics/areas have a certain author been working on?
- What topics have evolved the least? or which research has evoked least interest in recent times?
- Which authors were working together on a topic during a particular time?
- Which articles were the most influential to a particular field?
- What topic was most researched on during a particular time?
- Which geographic location has the most research output for a particular topic?