Patricia Roggero

Faster Proximity Searching with the Distal SAT

In this paper we present the Distal Spatial Approximation Tree (DiSAT), an algorithmic improvement of SAT. Our improvement increases the discarding power of the SAT by selecting distal nodes instead of the proximal nodes proposed in the original paper. Our approach is parameter free and it was the most competitive in an extensive benchmarking, from two to forty times faster than the SAT, and faster than the List of Clusters (LC) which is considered the state of the art for main memory, linear sized indexes in the model of distance computations.

Faster proximity searching with the distal SAT

Abstract Searching by proximity has been a source of puzzling behaviors and counter-intuitive findings for well established algorithmic design rules. One example is a linked list; it is the worst data structure for exact searching, and one of the most competitive for proximity searching. Common sense also dictates that an online data structure is less competitive than the full-knowledge, static version. A counter example in proximity searching is the static Spatial Approximation Tree ( SAT ), which is slower than its dynamic version ( DSAT ). In this paper we show that changing only the insertion policy of the SAT , leaving every other aspect of the data structure untouched, can produce a systematically faster index. We call the index Distal Spatial Approximation Tree ( DiSAT ). We found that even a random insertion policy produce a faster version of the SAT , which explains why the DSAT is faster than SAT . In brief, the SAT is improved by selecting distal, instead of proximal, nodes. This is the exact opposite of the insertion policy proposed in the original paper, and can be used in main or secondary memory versions of the index. We tested our approach with representatives of the state of the art in exact proximity searching. As it happens often in experimental setups, there are no absolute winners in all the aspects tested. Our data structure has no parameters to tune-up and a small memory footprint. In addition it can be constructed quickly. Our approach is among the most competitive, those outperforming DiSAT achieve this at the expense of larger memory usage or an impractical construction time.

Reaching near neighbors with far and random proxies

Proximity searching is an algorithmic abstraction covering a large number of applications in areas such as machine learning, statistics, multimedia information retrieval, computer vision and pattern recognition, to name a few. The algorithmic problem consist in preprocessing a set of objects to quickly find the objects near a given query.

Delayed Insertion Strategies in Dynamic Metric Indexes

Dynamic data structures are sensitive to insertion order, particularly tree-based data structures. In this paper we present a buffering heuristic allowing delayed root selection (when enough data has arrived to have valid statistics) useful for hierarchical indexes. Initially, when less than

List of Clustered Permutations in Secondary Memory for Proximity Searching

M

An Efficient Alternative for Deletions in Dynamic Spatial Approximation Trees

objects have been inserted queries are answered from the buffer itself using an online-friendly algorithm which can be simulated by AESA (Approximating and Eliminating Search Algorithm) or can be implemented with the dynamic data structure being optimized. When the buffer is full the tree root can be selected in a more informed way using the distances between the

New deletion method for dynamic spatial approximation trees

M

Indexación y recuperación de información multimedia

objects in the buffer. Buffering has an additional usage, multiple routing strategies can be designed depending on statistics of the query. A complete picture of the technique includes doing a recursive best-root selection with much more parameters. We focus on the Dynamic Spatial Approximation Tree ({\em DSAT}) investigating the improvement obtained in the first level of the tree (the root and its children). Notice that if the buffering strategy is repeated recursively we can obtain a boosting on the performance when the data structure reaches a stable state. For this reason even a very small improvement in performance is significant. We present a systematic improvement in the query complexity for several real time, publicly available data sets from the SISAP repository with our buffering strategies.