Read more
In this monograph, we study the problem of high-dimensional indexing and systematically introduce two efficient index structures: one for range queries and the other for similarity queries. Extensive experiments and comparison studies are conducted to demonstrate the superiority of the proposed indexing methods.
Many new database applications, such as multimedia databases or stock price information systems, transform important features or properties of data objects into high-dimensional points. Searching for objects based on these features is thus a search of points in this feature space. To support efficient retrieval in such high-dimensional databases, indexes are required to prune the search space. Indexes for low-dimensional databases are well studied, whereas most of these application specific indexes are not scaleable with the number of dimensions, and they are not designed to support similarity searches and high-dimensional joins.
List of contents
High-Dimensional Indexing.- Indexing the Edges - A Simple and Yet Efficient Approach to High-Dimensional Range Search.- Performance Study of Window Queries.- Indexing the Relative Distance - An Efficient Approach to KNN Search.- Similarity Range and Approximate KNN Searches with iMinMax.- Performance Study of Similarity Queries.- Conclusions.
Summary
In this monograph, we study the problem of high-dimensional indexing and systematically introduce two efficient index structures: one for range queries and the other for similarity queries. Extensive experiments and comparison studies are conducted to demonstrate the superiority of the proposed indexing methods.
Many new database applications, such as multimedia databases or stock price information systems, transform important features or properties of data objects into high-dimensional points. Searching for objects based on these features is thus a search of points in this feature space. To support efficient retrieval in such high-dimensional databases, indexes are required to prune the search space. Indexes for low-dimensional databases are well studied, whereas most of these application specific indexes are not scaleable with the number of dimensions, and they are not designed to support similarity searches and high-dimensional joins.
