PPP-Codes: Similarity Search Index
| Authors | |
|---|---|
| Year of publication | 2013 |
| MU Faculty or unit | |
| Web | PPP-Codes homepage |
| Description | Many current applications need to organize data with respect to mutual similarity between data objects (for instance biometric systems). A typical general strategy to retrieve the most similar objects to a given example is to access and then refine a candidate set of objects; the overall search costs (and search time) then typically correlate with the candidate set size. The PPP-Codes index provides a generic approach that combines several independent indexes by aggregating their candidate sets in such a way that the resulting candidate set can be one or two orders of magnitude smaller (while keeping the answer quality). This achievement comes at the expense of higher computational costs of the ranking algorithm but our experiments on various datasets indicate that the overall gain can be significant, especially for data types with large objects or expensive similarity function such as biometric systems. |
| Related projects: |