Databases play an important role in our world today. But one of the greatest challenges in current database systems is the "Variety"of the data. Multi-model databases have emerged to address this challenge by supporting multiple data models against a single, integrated backend. Unfortunately, the existing principles and research ideas of multi-model data management are so scarce and far from perfect.
This project will tackle this challenge from the fundamental problems to practical techniques. We will propose novel solutions for pressing problems on multi-model data management, including unified data storage abstraction, unified querying and indexing techniques, relaxing consistency model and fine-grained multi-model isolation. As long term contribution, this project supports to build next-generation platform to solve the "Variety" challenge of big data.