| Sponsor | Ramesh Jain |
| Area | ECE |
Problem
Traditional approaches for indexing of data depend on the type of data. Current approaches for indexing multimedia create many silos (audio, video, stats, text, relational data, etc.) that do not talk to each other. This makes the access of information very tedious for users. A good example of this are some of the news and sports websites, like CNN, where information exists in multimedia form in many different sections in multiple silos.
In this project, different approaches for indexing different types of data will be studied with the goal to find a unifying scheme that will break these silos and unify the data from different sources in different media. A desirable outcome of the study will be to understand the current limitations of data indexing. It is expected that effort equivalent to reading about 8-10 research papers will be required. The findings of the research will be presented in a short white paper. Student will work closely with the sponsor in selecting the papers and synthesizing their findings.