Simpler is More: Efficient Top-K Nearest Neighbors Search on Large Road Networks

Yiqi Wang,Long Yuan,Wenjie Zhang,Xuemin Lin,Zi Chen,Qing Liu

2024-08-10

Abstract:Top-k Nearest Neighbors (kNN) problem on road network has numerous applications on location-based services. As direct search using the Dijkstra's algorithm results in a large search space, a plethora of complex-index-based approaches have been proposed to speedup the query processing. However, even with the current state-of-the-art approach, long query processing delays persist, along with significant space overhead and prohibitively long indexing time. In this paper, we depart from the complex index designs prevalent in existing literature and propose a simple index named KNN-Index. With KNN-Index, we can answer a kNN query optimally and progressively with small and size-bounded index. To improve the index construction performance, we propose a bidirectional construction algorithm which can effectively share the common computation during the construction. Theoretical analysis and experimental results on real road networks demonstrate the superiority of KNN-Index over the state-of-the-art approach in query processing performance, index size, and index construction efficiency.

Databases

What problem does this paper attempt to address?

### Problems Addressed by the Paper

The paper aims to address the problem of Top-𝑘 Nearest Neighbor (𝑘NN) search in large-scale road networks. Specifically, the paper focuses on the following points:

1. **Long Query Latency**: Existing complex indexing methods, although accelerating query processing, still suffer from long query latency, especially when the road network is large.
2. **Large Index Space Occupation**: Existing methods require a large amount of storage space to maintain the index, which is a significant limitation in practical applications.
3. **Long Index Construction Time**: The time required to construct these complex indexes is very long, making these methods less feasible in practical applications.

### Solution

To overcome the above issues, the paper proposes a new simple index design method called KNN-Index. The main features of this method include:

1. **Simple Index Structure**: KNN-Index only records the top 𝑘 nearest neighbors for each vertex, thereby significantly reducing the index space occupation.
2. **Efficient Index Construction Algorithm**: By introducing a bidirectional construction algorithm, the computation during the construction process is effectively shared, significantly improving the efficiency of index construction.
3. **Progressive Query Processing**: KNN-Index supports progressive query processing, allowing results to be output gradually within limited latency, enabling users to obtain useful information before the query is fully completed.

### Main Contributions

1. **New Index Design Paradigm**: A simple and effective 𝑘NN index design is proposed, addressing the issues of long query latency, large index space, and long index construction time caused by existing complex indexes.
2. **Efficient Index Construction and Maintenance Algorithms**: New index construction algorithms are designed to effectively share shortest distance computations, and maintenance algorithms are proposed to handle object insertion and deletion.
3. **Extensive Experimental Validation**: Extensive experiments were conducted on real road network datasets, and the results show that compared to the state-of-the-art TEN-Index, KNN-Index significantly improves performance in terms of index space, query time, and index construction time.

### Experimental Results

- **Index Space**: The index space of KNN-Index is two orders of magnitude smaller than that of TEN-Index.
- **Query Time**: The query time of KNN-Index is two orders of magnitude faster than that of TEN-Index.
- **Index Construction Time**: The index construction time of KNN-Index is two orders of magnitude faster than that of TEN-Index.

### Conclusion

By proposing KNN-Index, the paper successfully addresses the key issues of query latency, index space, and index construction time in Top-𝑘 Nearest Neighbor search in large-scale road networks, providing an efficient and feasible solution for practical applications.