Chord-Based Multi-Attribute Multi-Keyword Query and Hot-Set Cache

In the paper, a new model (MAK-Chord) is presented which is expanded from Chord. It generates fingerprints for each resource which include all the attribute keyword information and take into account the query frequency difference. It gives two different mappings between resources and nodes and effectively supports multi-attribute multi-keyword fuzzy-matching query. And it also includes a new load balancing mechanism (hot-set cache) which can balance the load due to hot files for its query and other range or similar query. The experiment results show MAK-Chord has high recall rate with low storage redundancy, and it can effectively balance load when user query load is unbalanced.