A road network embedding technique for k-nearest neighbor search in moving object databases

Shahabi, Cyrus; Kolahdouzan, Mohammad R.; Sharifzadeh, Mehdi

doi:10.1145/585163.585167

Cited by 83 publications

(94 citation statements)

References 1 publication

(1 reference statement)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…As a variant of NN queries, Continuous Nearest Neighbor queries (CNN) [1,13,24,15,16] report the kNN results continuously while the user is moving along a path. This type of queries aims to find the split points on the query path where an update of the kNN is required, and thus to avoid unnecessary recomputation.…”

Section: Related Workmentioning

confidence: 99%

Finding regions of interest using location based social media

et al. 2016

View full text Add to dashboard Cite

The discovery of regions of interest in city groups is increasingly important in recent years. In this light, we propose and investigate a novel problem called Region Discovery query (RD query) that finds regions of interest with respect to a user's current geographic location. Given a set of spatial objects O and a query location q, if a circular region ω is with high spatial-object density and is spatially close to q, it is returned by the query and is recommended to users. This type of query can bring significant benefit to users in many useful applications such as trip planning and region recommendation. The RD query faces a big challenge: how to prune the search space in the spatial and density domains. To overcome the challenge and process the RD query efficiently, we propose a novel collaboration search method and we define a pair of bounds to prune the search space effectively. The performance of the RD query is studied by extensive experiments on real and synthetic spatial data.

show abstract

Section: Related Workmentioning

confidence: 99%

Finding regions of interest using location based social media

et al. 2016

View full text Add to dashboard Cite

show abstract

“…One important class is the road network kNN where the space is constrained due to an underlying graph structure. Shahabi et al [5] introduce an embedding technique to transfer the road network to a constraint-free high dimensional space. Papadias et al [6] introduce techniques for network kNN queries by integrating network and Euclidean information and capturing pragmatic constraints.…”

Section: Knn Query Processing Techniquesmentioning

confidence: 99%

Indexing land surface for efficient kNN query

2008

Self Cite

View full text Add to dashboard Cite

The class of k Nearest Neighbor (kNN) queries is frequently used in geospatial applications. Many studies focus on processing kNN in Euclidean and road network spaces. Meanwhile, with the recent advances in remote sensory devices that can acquire detailed elevation data, the new geospatial applications heavily operate on this third dimension, i.e., land surface. Hence, for the field of databases to stay relevant, it should be able to efficiently process spatial queries given this constrained third dimension. However, online processing of the surface k Nearest Neighbor (skNN) queries is quite challenging due to the huge size of land surface models which renders any accurate distance computation on the surface extremely slow. In this paper, for the first time, we propose an index structure on land surface that enables exact and fast responses to skNN queries. Two complementary indexing schemes, namely Tight Surface Index (TSI) and Loose Surface Index (LSI), are constructed and stored collectively on a single novel data structure called Surface Index R-tree (SIR-tree). With those indexes, we can process skNN query efficiently by localizing the search and minimizing the invocation of the costly surface distance computation and hence incurring low I/O and computation costs. Our algorithm does not need to know the value of k a priori and can incrementally expand the search region using SIR-tree and report the query result progressively. It also reports the exact shortest surface paths to the query results. We show through experiments with real world data sets that our algorithm has better performance than the competitors in both efficiency and accuracy.

show abstract

“…Nearest neighbor (NN) search on road networks is an emerging research topic in recent years [11,6,9,8]. It is closely related to the single-source shortest path problem, which has been studied since Dijkstra [4].…”

Section: Related Workmentioning

confidence: 99%

“…Shahabi et al applied graph embedding techniques to kNN search on road networks [11]. They transformed a road network to a high-dimensional Euclidean space where traditional NN search algorithms can be applied.…”

Section: Related Workmentioning

confidence: 99%

“…Recently, the research focus is brought to spatial network databases (SNDB) where objects are restricted to move on predefined roads [11,6,9,8]. In SNDB, a road network is modeled as a graph G (< V, E >), where a vertex (node) denotes a road junction and an edge denotes the road between two junctions; and the weight of the edge denotes the network distance.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Fast Nearest Neighbor Search on Road Networks

Lee

2006

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Abstract. Nearest neighbor (NN) queries have been extended from Euclidean spaces to road networks. Existing approaches are either based on Dijkstra-like network expansion or NN/distance precomputation. The former may cause an explosive number of node accesses for sparse datasets because all nodes closer than the NN to the query must be visited. The latter, e.g., the Voronoi Network Nearest Neighbor (V N 3 ) approach, can handle sparse datasets but is inappropriate for medium and dense datasets due to its high precomputation and storage overhead. In this paper, we propose a new approach that indexes the network topology based on a novel network reduction technique. It simplifies the network by replacing the graph topology with a set of interconnected tree-based structures called SPIE's. An nd index is developed for each SPIE and our new (k)NN search algorithms on an SPIE follow a predetermined tree path to avoid costly network expansion. By mathematical analysis and experimental results, our new approach is shown to be efficient and robust for various network topologies and data distributions.

show abstract

A road network embedding technique for k-nearest neighbor search in moving object databases

Cited by 83 publications

References 1 publication

Finding regions of interest using location based social media

Finding regions of interest using location based social media

Indexing land surface for efficient kNN query

Fast Nearest Neighbor Search on Road Networks

Contact Info

Product

Resources

About