Existing key-value stores support web applications with high scalability and availability. However, the
limited data access via primary keys has restricted data retrieval from key-value stores. In this paper,
we have proposed a searchable key-value store with hierarchical data model, HierKV, to support fast
data retrieval and search with secondary attributes. HierKV accelerates data retrieval by attributes via
a hierarchical data model and index structure. A hierarchical TF-IDF ranking mechanism is proposed
for HierKV to properly reflect keyword matches in different hierarchies. Experiments have shown that
the proposed HierKV system can achieve good search performance.