2009 | OriginalPaper | Buchkapitel
Identifying the Most Endangered Objects from Spatial Datasets
verfasst von : Hua Lu, Man Lung Yiu
Erschienen in: Scientific and Statistical Database Management
Verlag: Springer Berlin Heidelberg
Aktivieren Sie unsere intelligente Suche, um passende Fachinhalte oder Patente zu finden.
Wählen Sie Textabschnitte aus um mit Künstlicher Intelligenz passenden Patente zu finden. powered by
Markieren Sie Textabschnitte, um KI-gestützt weitere passende Inhalte zu finden. powered by
Real-life spatial objects are usually described by their geographic locations (e.g., longitude and latitude), and multiple quality attributes. Conventionally, spatial data are queried by two orthogonal aspects: spatial queries involve geographic locations only; skyline queries are used to retrieve those objects that are not dominated by others on all quality attributes. Specifically, an object
p
i
is said to dominate another object
p
j
if
p
i
is no worse than
p
j
on all quality attributes and better than
p
j
on at least one quality attribute. In this paper, we study a novel query that combines both aspects meaningfully. Given two spatial datasets
P
and
S
, and a neighborhood distance
δ
, the
most endangered object query
(MEO) returns the object
s
∈
S
such that within the distance
δ
from
s
, the number of objects in
P
that dominate
s
is maximized. MEO queries appropriately capture the needs that neither spatial queries nor skyline queries alone have addressed. They have various practical applications such as business planning, online war games, and wild animal protection. Nevertheless, the processing of MEO queries is challenging and it cannot be efficiently evaluated by existing solutions. Motivated by this, we propose several algorithms for processing MEO queries, which can be applied in different scenarios where different indexes are available on spatial datasets. Extensive experimental results on both synthetic and real datasets show that our proposed advanced spatial join solution achieves the best performance and it is scalable to large datasets.