Below is a list of projects that I have worked on, in reverse chronological order. My publications are also available.
Work in progress, starting from the spring of 1997, with the rest of the WHIPS team.
Work in progress. Under the guidance of my advisor, Prof. Jennifer Widom.
Garlic is a distributed, heterogeneous, multimedia information system. Query optimization in such systems is challenging because a single query can span several specialized sources with diverse search capabilities.
I worked on query processing in Garlic during the summer of 1996, at IBM Almaden Research Center, with Laura Haas, Donald Kossmann, and Edward Wimmers. We designed and implemented a query processor prototype that can be taught about the capabilities of new data sources, and produce high quality plans even in the presence of nonstandard data, strange methods, and unusual query processing capabilities.
l2m and m2l are automatic translators from LOREL (query language for LORE, with an OQL flavor) to MSL (internal language used by TSIMMIS, with a Datalog flavor), and vice versa.
I worked on l2m in the winter of 1995, and m2l in the winter of 1996. l2m is already linked into MOBIE to provide TSIMMIS users with a uniform language interface. m2l is also finished; it will enable TSIMMIS wrappers and mediators to use LORE databases for their local storage.
Cellsim is a simulation package that models the movements and behaviors of cells immersed in an incompressible fluid field. This project is led by Lance Davidson, under the supervision of Prof. George Oster in ESPM Department. The applications of cellsim include the simulation of epithelial cell rearrangements, embryogenesis, polymer movements, etc. Cellsim package also includes a parser for a simulation language, tools for building interesting inputs (such as Voronoi tessellation), and a visualization program for X windows system. Cellsim runs on both uniprocessor and multiprocessor architectures. The parallel version of cellsim is written in Split-C.
I worked on cellsim for a year since the summer of 1994. My work includes optimizing the parallel implementation, porting the newest sequential version, developing support software, and performance modeling on various parallel architectures, under the guidance of Prof. Katherine Yelick in CS Division, as a part of the Multipol project.